Ludger O.

Figura
Arthur A. Teixeira

Food
Physics
Physical Properties - Measurement and
Applications
Second Edition
Food Physics
Ludger O. Figura • Arthur A. Teixeira

Food Physics
Physical Properties - Measurement
and Applications

Second Edition
Ludger O. Figura Arthur A. Teixeira
Food Engineering University of Florida
Hochschule Osnabrück, University Gainesville, FL, USA
of Applied Sciences
Osnabrück, Germany

ISBN 978-3-031-27397-1 ISBN 978-3-031-27398-8 (eBook)

https://doi.org/10.1007/978-3-031-27398-8

# The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland
AG 2007, 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

The demands on food production are increasing constantly. Food shall be fresh and
healthy, contain only few additives, be stable and safe. Foods shall be produced in a
way that is sustainable and ensures the needs of a growing world population.
In order to meet these challenges, cooperation between science and craftsmanship
is necessary to develop efficient, energy-saving, and low-waste processes that fit into
individual geographical situations.
In addition, there is a need for efficient forms of learning that enable us to extract,
understand, and scientifically evaluate the required content from the wealth of
digitally available information so that we can contribute to that task throughout
our working life.
The new edition of Food Physics wants to make a contribution to this. The core
statements of more than 1000 current publications have been included, each chapter
begins with simple physics and, with the help of examples, introduces practical
measurement technology and food characterization. At the end of each chapter, there
is a list of technical applications that can be used for in-depth study or for finding
ideas for your own investigations in projects, bachelor and master theses. The
practitioner will probably start a chapter with these applications and then scroll
forward if necessary.
The chapter on-line sensors and the sections on electromagnetic and optical
properties have been expanded, and a separate chapter has been devoted to the
texture of food. For efficient learning, didactic elements were integrated into the
book, in the E-book you can continue reading directly in the cited literature with the
help of links. However, you still have to read and to study by yourself.
With this in mind, we wish all readers the beneficial experience of developing
new understanding and using this together with own ideas for a sustainable global
food production.

Quakenbrück, Germany Ludger O. Figura

Spring 2023 Arthur A. Teixeira

v
Contents

1 Water Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Water Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Solid Boundary Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 Adsorption Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Porous Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.5 Sorption Isotherms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.6 Moisture Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.7 Hygroscopicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.8 BET Equation for Foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.9 GAB Equation for Food . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.10 Sorption Enthalpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.11 Other Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.12 Shelf Life of Food Related to Water Activity . . . . . . . . . . . . . . 41
1.13 Laboratory Determination of Sorption Isotherms . . . . . . . . . . . . 48
1.14 Standard for Sorption Isotherms . . . . . . . . . . . . . . . . . . . . . . . . 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2 Mass Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.1 Weighing and Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.2 Laboratory Methods for Determining Density . . . . . . . . . . . . . . 77
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3 Disperse Systems: Particle Characterization . . . . . . . . . . . . . . . . . . 101
3.1 Particle Size Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3.2 Measurement of Particle Size Distributions . . . . . . . . . . . . . . . . 131
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
4 Rheological Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.1 Elastic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.1.1 Stress, Pressure, Uniaxial Tension, Young’s
Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.1.2 Three-Dimensional Stress, Bulk Compression . . . . . . . 151
4.1.3 Shear, Shear Modulus . . . . . . . . . . . . . . . . . . . . . . . . . 152
4.1.4 Transverse Strain, Poisson’s Ratio . . . . . . . . . . . . . . . . 154

vii
viii Contents

4.2 Rheological Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

4.3 Viscous Behavior, Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
4.3.1 Shear Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
4.3.2 Newtonian Flow Behavior . . . . . . . . . . . . . . . . . . . . . 164
4.3.3 Non-Newtonian Flow Behavior . . . . . . . . . . . . . . . . . . 166
4.3.4 Comparison: Newtonian and Non-Newtonian Fluids . . . 167
4.3.5 Pseudoplastic Flow Behavior . . . . . . . . . . . . . . . . . . . 168
4.3.6 Thixotropic Flow Behavior . . . . . . . . . . . . . . . . . . . . . 169
4.3.7 Dilatant Flow Behavior . . . . . . . . . . . . . . . . . . . . . . . . 169
4.3.8 Rheopectic Flow Behavior . . . . . . . . . . . . . . . . . . . . . 170
4.3.9 Plastic Flow Behavior . . . . . . . . . . . . . . . . . . . . . . . . . 170
4.3.10 Overview: Non-Newtonian Flow Behavior . . . . . . . . . . 173
4.3.11 Model Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
4.3.12 Ostwald–de Waele Law . . . . . . . . . . . . . . . . . . . . . . . 177
4.3.13 Model Functions for Plastic Fluids . . . . . . . . . . . . . . . 180
4.4 Temperature Dependency of Viscosity . . . . . . . . . . . . . . . . . . . 183
4.5 Viscosity of Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
4.6 Viscoelasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
4.7 Rheological Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
4.7.1 Rotational Rheometers . . . . . . . . . . . . . . . . . . . . . . . . 193
4.7.2 Oscillation Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
4.7.3 Further Measurement Systems . . . . . . . . . . . . . . . . . . . 213
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
5 Texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
5.1 Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
5.2 Test Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
5.3 Stress Relaxation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
5.4 Creep Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
5.5 Deborah’s Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
5.6 Oscillating Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
5.7 Fracture Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
5.8 Texture Profile Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
6 Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
6.1 Interfacial Tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
6.1.1 Curved (Convex/Concave) Interfaces . . . . . . . . . . . . . . 250
6.1.2 Temperature Dependence of Interfacial Tension . . . . . . 256
6.1.3 Concentration Dependence of Interfacial Tension . . . . . 260
6.1.4 Emulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
6.1.5 Liquid–Liquid–Gas Interface . . . . . . . . . . . . . . . . . . . . 265
6.1.6 Solid–Liquid–Gas Interface . . . . . . . . . . . . . . . . . . . . . 269
6.2 Kinetic Phenomena at Interfaces . . . . . . . . . . . . . . . . . . . . . . . 271
6.3 Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Contents ix

7 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
7.1 Steady State Diffusion in Solids . . . . . . . . . . . . . . . . . . . . . . . . 285
7.2 Conductivity, Conductance, Resistance . . . . . . . . . . . . . . . . . . . 289
7.3 Steady State Transport Through Solid Multi-layers . . . . . . . . . . 290
7.4 Food Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
7.4.1 Composite Films . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
7.5 Permeation as a Molecular Process . . . . . . . . . . . . . . . . . . . . . . 298
7.6 Temperature Dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
7.7 Measurement of Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . 300
7.8 Analogous Transport Phenomena: Heat and Electricity . . . . . . . 304
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
8 Thermal Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
8.1 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
8.2 Heat and Enthalpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
8.3 Thermodynamics: Basis Principles . . . . . . . . . . . . . . . . . . . . . . 318
8.3.1 Laws of Thermodynamics . . . . . . . . . . . . . . . . . . . . . . 318
8.4 Heat capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
8.4.1 Ideal Gases and Ideal Solids . . . . . . . . . . . . . . . . . . . . 325
8.4.2 Heat Capacity of Real Solids . . . . . . . . . . . . . . . . . . . . 327
8.5 Classification of Phase Transitions . . . . . . . . . . . . . . . . . . . . . . 328
8.6 Heat Transfer in Food . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
8.6.1 Heat Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
8.6.2 Heat Conduction Transfer . . . . . . . . . . . . . . . . . . . . . . 335
8.6.3 Convection Heat Transfer . . . . . . . . . . . . . . . . . . . . . . 344
8.6.4 Heat Transfer by Phase Transition . . . . . . . . . . . . . . . . 349
8.6.5 Thermal Conductivity . . . . . . . . . . . . . . . . . . . . . . . . . 351
8.6.6 Thermal Diffusivity . . . . . . . . . . . . . . . . . . . . . . . . . . 360
8.6.7 Measurement of Thermal Conductivity and Thermal
Diffusivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
8.7 Caloric Value of Foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
8.7.1 Caloric (Energy) Requirement of the Human Body . . . . 364
8.7.2 Caloric Value of Food . . . . . . . . . . . . . . . . . . . . . . . . 369
8.7.3 Measurement of Caloric (Combustion) Values . . . . . . . 373
8.8 Thermal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
8.8.1 Thermogravimetry (TG) . . . . . . . . . . . . . . . . . . . . . . . 374
8.8.2 Heat Flow Calorimetry . . . . . . . . . . . . . . . . . . . . . . . . 379
8.8.3 Combustion Calorimetry . . . . . . . . . . . . . . . . . . . . . . . 392
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
9 Electrical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
9.1 Electrical Conductivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
9.1.1 Temperature Dependence of Electrical Conductivity . . . 407
9.1.2 Electrolyte Solutions . . . . . . . . . . . . . . . . . . . . . . . . . 408
9.1.3 Conductivity of Solid Foods . . . . . . . . . . . . . . . . . . . . 414
9.2 Capacitance and Inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
x Contents

9.3 Impedance and Admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . 418

9.3.1 Frequency Dependance of Impedance . . . . . . . . . . . . . 420
9.4 Measurement of Electrical Conductivity and Impedance . . . . . . 420
9.5 Zeta Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
9.6 Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
9.7 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
10 Magnetic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
10.1 Magnetic Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 432
10.1.1 Paramagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
10.1.2 Ferromagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
10.1.3 Diamagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
10.2 Magnetization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
10.2.1 Hysteresis in Magnetization . . . . . . . . . . . . . . . . . . . . 437
10.2.2 Metal Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
10.2.3 Lorentz Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
10.2.4 Induction Cooker . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
10.3 Magnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
10.3.1 Electron Spin Resonance . . . . . . . . . . . . . . . . . . . . . . 450
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
11 Electromagnetic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
11.1 Electromagnetic Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
11.2 Electric Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
11.2.1 Temperature Dependency . . . . . . . . . . . . . . . . . . . . . . 461
11.2.2 Frequency Dependency . . . . . . . . . . . . . . . . . . . . . . . . 463
11.2.3 Complex Permittivity . . . . . . . . . . . . . . . . . . . . . . . . . 465
11.3 Microwaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
11.3.1 Conversion of Microwaves into Heat . . . . . . . . . . . . . . 468
11.3.2 Penetration Depth of Microwaves . . . . . . . . . . . . . . . . 470
11.3.3 Microwave Heating . . . . . . . . . . . . . . . . . . . . . . . . . . 473
11.4 Terahertz Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474
11.5 NIR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
12 Optical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
12.1 Refraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
12.1.1 Measurement of Refractive Index . . . . . . . . . . . . . . . . 488
12.2 Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
12.3 Reflection, Absorption, and Transmission . . . . . . . . . . . . . . . . . 494
12.4 Scattering and Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
12.5 Colorimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
12.5.1 Color as a Result of Selective Absorption . . . . . . . . . . 497
12.5.2 Physiology of Color Vision . . . . . . . . . . . . . . . . . . . . . 499
12.5.3 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
Contents xi

13 UV and X-rays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

13.1 UV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
13.2 X-ray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
13.2.1 X-ray Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
13.2.2 Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . 517
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
14 Radioactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523
14.1 Types of Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
14.1.1 Activity and Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
14.1.2 Measurement of Ionizing Radiation (α-, β-, γ-) . . . . . . . 528
14.1.3 Natural Radioactivity . . . . . . . . . . . . . . . . . . . . . . . . . 530
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
15 Acoustic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
15.1 Sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
15.1.1 Speed of Sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
15.1.2 Loudness and Volume . . . . . . . . . . . . . . . . . . . . . . . . 543
15.1.3 Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544
15.2 Acoustic Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . 545
15.3 Ultrasonic Sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
16 On-line Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
16.1 Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
16.2 Control–Directing–Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . 555
16.3 Working Principles of On-line Sensors . . . . . . . . . . . . . . . . . . . 559
16.3.1 Piezoresistive Sensors . . . . . . . . . . . . . . . . . . . . . . . . . 559
16.3.2 Mechanical Oscillation . . . . . . . . . . . . . . . . . . . . . . . . 560
16.3.3 Induction (Flow Measurement, Metal Detection) . . . . . 564
16.3.4 Ultrasound Transit Time . . . . . . . . . . . . . . . . . . . . . . . 566
16.3.5 Capacity and Permittivity . . . . . . . . . . . . . . . . . . . . . . 567
16.3.6 Refraction and Absorption . . . . . . . . . . . . . . . . . . . . . 571
16.4 Chemo- and Bio-sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
16.5 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578

Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
Water Activity
1

Water is an important component of nearly all food materials and plays a decisive
role in dictating the physical properties, as well as quality, microbial, chemical, and
biochemical degradation of the food material [1–4]. For most food materials, unless
the moisture content is reduced below 50% (wet basis), much of the water content is
freely available to behave physically as pure water with properties such as vapor
pressure equal to pure water. As moisture content is lowered further, a point will be
reached at which the water becomes less active in that it cannot act physically or
chemically as pure water. For example, it cannot freeze or act as a solvent or reactant.
In this state, it is considered to be bound water.
The way in which water is bound to the internal structure of the food, the degree
to which it is freely available to act as a solvent, to vaporize or freeze, or the degree to
which it is chemically bound and unavailable can all be reflected by an ability to
specify the water activity of a food material. The matrix presented in Table 1.5
attempts to illustrate the range of conditions under which water may be bound and
the role it can play under each condition. The water activity of a food can be thought
of as the equilibrium relative humidity of the food material. When a food sample
comes into equilibrium with the atmosphere surrounding it, the water activity in the
food sample becomes equal to the relative humidity of the atmosphere surrounding
it. Once this equilibrium is reached, the food sample neither gains nor loses moisture
over time. A more comprehensive definition of water activity will be given in this
chapter.

1.1 Water Activity

To understand the concept of water activity, let’s first consider an aqueous solution
at a given temperature. The vapor pressure p above the solution is in the range
between zero and the saturation vapor pressure ps of water at this temperature. If we
have the solution in a closed vessel and track the vapor pressure over time in the gas
space of the vessel, we observe an increase in vapor pressure up to an equilibrium

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 1

L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_1
2 1 Water Activity

Fig. 1.1 Vessel with aqueous

solution of concentration c.
The water vapor pressure p in
the gas chamber above the
liquid is detected with a sensor

value p. How high this value p is depends on how “free” the water molecules of the
solution are. If the water molecules are restricted in their “freedom” by interaction
with dissolved substances in the solution, the equilibrium vapor pressure is lower.
This equilibrium vapor pressure can thus be used as an indicator of the freedom
(also: reactivity, activity) of the water molecules of a solution. One refers to the
relative vapor pressure, i.e. the quotient of vapor pressure p and saturation vapor
pressure ps above the solution as the water activity aW of the solution (Fig. 1.1)
p
aW = : ð1:1Þ
ps

In an aqueous solution in which the freedom of the water molecules is not

restricted, the water activity has its maximum value aW = 1. This is the case when
there are no interactions or when the concentration of the solution is zero, i.e. the
solution consists of 100% pure water. In thermodynamics, the reactivity of a
substance is expressed by means of the chemical potential of the substance i,
generally

∂G
μi = = μ0i þ RT  ln ai : ð1:2Þ
∂ni p,T,nj

In the case of water, the chemical potential is therefore a function of water activity
aW [5]:

∂G
μW = = μ0W þ RT  ln aW : ð1:3Þ
∂nW p,T,nS

μ chemical potential in Jmol-1

μ0 chemical potential under standard conditions in Jmol-1
G Gibbs energy in J
n amount of substance in mol
R general gas constant (8.314 JK-1mol-1)
T temperature in K
(continued)
1.1 Water Activity 3

p pressure in Pa
aW water activity

In our thought experiment with the aqueous solution in a closed vessel, we had
tacitly assumed that no other gases were present. Let us now mentally allow the
vessel to contain air in addition to water vapor: Now there may be interactions
between the different molecules of the gas phase, i.e. between water molecules and
the molecules of air such as N2, O2, CO2, and other air components. Through such
interactions, the above-called water vapor partial pressure can be changed,
e.B. lowered. For this reason a distinction is made between the fugacity f and the
partial pressure p of a gas.

f = γ  p: ð1:4Þ

The fugacity is the actual measured pressure of a gas, which can be lower than the
partial pressure of the gas as a result of interactions between the gas molecules. The
quotient of fugacity and partial pressure is called the fugacity coefficient of the gas
concerned [6].

f
γ= : ð1:5Þ
p
With the assumption that air behaves like an ideal gas at normal pressure and the
assumption that the water vapor molecules in atmospheric air also behave like
molecules of an ideal gas—i.e., there are no interactions between the molecules of
the gas phase—i.e., because of γ = 1 and therefore f = p we can dispend the
distinction between fugacity and partial pressure. In the following, we always
speak of partial pressure of the water molecules in the gas phase, i.e. of the water
vapor partial pressure.
Let’s look again at the closed vessel in which our aqueous solution and the gas
space with this water vapor partial pressure are located: Let us now replace in our
thoughts a part of the liquid water with a few moles of a solid, soluble substance. The
ratio of the molar concentrations of the solution is now
nW
nW
V
= : ð1:6Þ
ðnW þnS Þ nW þ nS
V

Assuming that the solution behaves like an ideal solution, i.e. there are no
interactions between the molecules, then the relative water vapor pressure of the
solution is

p nW
= : ð1:7Þ
ps n W þ n S

i.e., the water activity of an ideal solution is calculated

4 1 Water Activity

nW
aW = : ð1:8Þ
nW þ nS
This relation is also known as the law of Raoult. It states that the water activity
(or also: the relative water vapor pressure) depends on the ratio of the amount of
water to the total amount of material of the ideal solution. This ratio is also called
mole fraction or relative mass fraction.
Because of

N
n= : ð1:9Þ
NA
the water activity of the ideal solution depends on the number of molecules N

nW NW
NW
aW = V
= V
= ð1:10Þ
ðnW þnS Þ ðN W þN S Þ NW þ NS
V V

with

m
n= : ð1:11Þ
M
Raoult’s law can be written

nW
aW = ð1:12Þ
nW þ M
mS
S

aW water activity
p vapor pressure in Pa
p saturation vapor pressure in Pa
n amount of substance in mol
N number of particles
NA Avogadro constant
M molar mass in kgmol-1
m mass in kg
V volume in m3
W subscript for water
S subscript for solute

i.e., for the water activity it plays a role which molecular weight M the solute has.
Substances with a small molecular weight, e.g. simple sugars, lower the water
activity more than the same mass of a higher molecular weight substance,
e.g. soluble starch. Because of the influence of the number of molecules, salts
such as NaCl, which dissociate in two ions in aqueous solutions, have twice the
effect as the same amount of substance of a substance that goes undissociated into
solution. The vapor pressure reduction by dissolved materials is also referred to as a
colligative property because of the number of molecules dependence. We will recall
1.1 Water Activity 5

this when we talk about boiling point increase and freezing point reduction of
solutions.
These simple considerations made here help us to understand the concept of water
activity and its causes. Raoult Law is based on ideal solutions. However, the
experimental measurement of the water activity of aqueous solutions shows that
only very diluted solutions of, e.g., sucrose or NaCl behave like ideal solutions. With
increasing concentration, there are numerous interaction effects between dissolved
molecules and water molecules, which lead to non-ideal behavior and make it very
difficult to calculate the water activity of foods from recipe information [7].

Bottom Line
The relative water vapor pressure of an aqueous solution is a measure of the
thermodynamic activity of the water phase of the solution. This property of
liquid and solid food is referred to as water activity. Their value is between
0 and 1.
In the case of non-ideal solutions, Raoult’s law has the form

nW
aW = γ W  ð1:13Þ
nW þ nS
with

γW activity coefficient of water

aW water activity
n amount of substance in mol

In the following example, the water activity of a sugar solution according to

Raoult is calculated and compared with the experimentally determined value.

Example
Calculation of the water activity of a sugar solution of 9 g of sucrose and 100 g
of water according to Raoult:

mS 9g
nS = = = 0:0263 mol
M S 342:3 g  mol - 1
m 100 g
nW = W = = 5:55 mol
M w 18:016 g  mol - 1
nW
atheor =
W nW þ nS
5:55
atheor = = 0:995
W 5:55 þ 0:0263
By experiment we measure aexp
W = 0:995, i.e. the activity coefficient is

(continued)
6 1 Water Activity

aexp 0:995
γW = W
= = 1:
atheor
W
0:995

Example
Calculation of the water activity of a sugar solution of 200 g of sucrose and
100 g of water according to Raoult:

mS 200 g
nS = = = 0:584 mol
M S 342:3 g  mol - 1
m 100 g
nW = W = = 5:55 mol
M w 18:016 g  mol - 1
nW
atheor =
W nW þ nS
5:55
atheor = = 0:904
W 5:55 þ 0:584
By experiment we measure aexp
W = 0:860, i.e. the activity coefficient is

aexp 0:86
γW = W
theor
= = 0:955:
aW 0:90

These examples of aqueous sucrose solutions show the range of water activity in
which Raoult’s law is valid. Prediction of the water activity of an 8.3% (m/m)
sucrose solution by means of Raoult’s law succeeds well. Apparently, the solution
behaves similarly to an ideal solution (γ W = 1). In the case of the 66.7% (m/m)
sucrose solution, the deviations from the behavior of an ideal solution are clearly
visible, the activity coefficient is γ W = 0.995. Raoult’s law was derived for ideal
solutions. A diluted sugar solution can apparently be regarded as an ideal solution up
to a certain limit. For higher accuracies or for higher concentrations, there are other
models for calculating the water activity of solutions [2, 5, 8–11]. A selection of
models is listed here (Tables 1.1, 1.2, and 1.3).
There are numerous other approaches to mathematically model the water activity
of special systems, e.g. salt mixtures or aqueous carbohydrate mixtures, refer [2, 8–
12].
1.2 Solid Boundary Surfaces 7

Table 1.1 Water activity calculation formula for non-electrolyte solutions

Raoult aW = γ W  nWnþn
W
S
n—amount of substance
(1882) Suitable for low concentrations in mol
Norrish log axWw = - k i ð1 - zW Þ2 zw—mole fraction water
(1966) Suitable for high concentrations Ki—Norrish constant
Grover aW = 1:04 - 0:1  ðsi  ci Þ þ 0:0045  ðsi  ci Þ2 c—concentration
(1947) i i s—sucrose equivalent
Empirical model for sweets, confectionery s (lactose) = 1.0
s (invert sugar) = 1.3
Money and aw = ð1:0þ0:27n
1:0
Þ
n—amount of substance
Born (1951) Empirical model sugar per 100 g water

Table 1.2 Water activity calculation formula for electrolyte solutions

Pitzer ϕ—osmotic coefficient
aW = exp - 0:01802  ϕ  Mi M—molality of solution
i

Bromley aW = exp (-0.01802  ϕ  z  M ) ϕ—osmotic coefficient

z—ion number
M—molality of solution

Table 1.3 Water activity calculation formula for multicomponent solutions

Ross (1975) aW = ðaW,i Þ aW, i—Water activity of component
i i
Ferro-Fontan-Benmergui-Chirife aW = ðaW,i Þ I
Ii
aW, i—Water activity of component
(1980) i i
I—Ionic strength of component i

Attention
In the case of real gases, instead of pressure, we speak of fugacity. The
quotient of fugacity and partial pressure is called the fugacity coefficient.
With real solutions we use activity of the solute instead of concentration.
The quotient of activity and concentration is called activity coefficient of the
solute. Water activity, on the other hand, does not describe the solute, but the
solvent of the aqueous solution.

1.2 Solid Boundary Surfaces

Let us now consider a solid surface which is contact to gaseous phase containing
water molecules. When gaseous water molecules hit the solid surface and begin to
adhere to it, the process is called adsorption with respect to the solid. When the same
molecules are already present on the solid surface and are escaping from the surface
8 1 Water Activity

Fig. 1.2 Solid surface in

adsorption equilibrium with
the surrounding atmosphere,
schematic. 1: gaseous phase,
2: sorbate, 3: sorbent
1

2
3

by being attracted to the fluid phase, the process is called desorption with respect to
the solid (Fig. 1.2). For example, when a fresh moist food sample is being dried to
produce a dehydrated product, it is placed into an environment of very low relative
humidity. Under that condition, the free water molecules on the surface begin to
escape into the relatively dry fluid atmosphere surrounding it in attempt to reach
equilibrium relative humidity, and the food sample is undergoing desorption. Alter-
natively, when a previously dehydrated food sample is being rehydrated by being
placed into a relatively humid environment or surrounded by water, the water
molecules from the surrounding fluid will be attracted to the dry surface and begin
to adhere to it in attempt to reach equilibrium relative humidity, and the food sample
is undergoing adsorption.
The rate at which these adsorption and desorption processes take place is
governed largely by the physical and chemical characteristics of the surface bound-
ary. For example, how readily the fluid molecules can either adhere or escape from
the surface can depend upon special features of the surface physical structure, such
as roughness, smoothness, and porosity (having a porous structure affecting absor-
bance, etc.). Besides physical adsorption, there can also be chemical binding that can
enhance or interfere with the surface adsorption/desorption process (chemisorption),
as shown in Fig. 1.3. These phenomena of adsorption and absorption will be
explored more fully in the following section.
If there is a chemical bond between the sorptive—here our water molecule—and
the adsorbate—here our solid food—one speaks chemisorption, while the formation
of physical and secondary bonds (e.g., capillary forces, electrostatic forces, dipole–
dipole interactions, hydrogen bonds) is summarized under the term physisorption
(Fig. 1.3). Table 1.4 is listing terms of sorption [13].
Water that is bound in solid foods can be divided into different groups based on
the type of binding. This can be done by the binding heat, which is the energy
necessary for the desorption of the adsorbed water. A high heat of binding is a sign of
a strong bond. If you want to remove such bound water from a body by evaporation,
you have to apply the bond heat in addition to the heat of evaporation. Table 1.5
shows some examples of water bound to different degrees to the solid surface. With
increasing binding strength, the binding heat increases and the mobility of the water
molecules decreases. Water molecules with a very small bond behave similarly to
“free” water molecules. In order to evaporate free water molecules from solids, only
the heat of evaporation need to be used, since the binding heat is zero.
1.2 Solid Boundary Surfaces 9

H
physisorption O chemisorption
H H +HOH
O O O O HO OH
H H H H
- - R O Si O Si
Cl Cl O O
Cl- Cl-

capillary dipole hydrogen chemical bonding

forces bonding bonding

gaseous water molecules

adsorption absorption adsorption desorption

adsorbate

Fig. 1.3 Terms used in sorption, by binding (top) by place of residence (down left), by direction of
process (down right)

Table 1.4 Terms of rf. Fig. 1.2 Term

sorption
Gaseous phase Sorptive
Solid material Sorbents
Bound molecules Sorbate

These considerations apply not only to flat interfaces where water molecules
adsorb, but also to modified or structured surfaces with cracks or pores. In particular,
pores with a small diameter, which are open to the gas phase, can change the
adsorption behavior of solids. More about this in Sect. 1.4.

Definition
Interfaces exist between non-miscible phases, which can be solid, liquid, or
gaseous. Interfaces bordering a gas phase are called surfaces.

The theory of adsorption of fluid phases at solid-state interfaces includes adsorp-

tion from the gas phase and also from the liquid phase. Separation processes based
on the adsorption of substances from liquids (example: drinking water treatment) can
therefore be treated with the same laws. In the following, however, we want to focus
on the balance between solid food and the gas phase.
 10

Table 1.5 Types of water binding and related heats of binding [14]
Adhering water Water as a Water of
droplets Capillary water solute Adsorbed water Crystal water constitution
Description Water dropping Water in coarse Water in fine Water between Water sitting on Water belonging to a Water in a
from bodies capillaries capillaries molecules surfaces crystal lattice compound
Binding None Mechanical Physical Physicochemical Chemical
Binding type Nonstoichiometric Stoichiometric
Water mobility Free Nearly free Decreased Nonmobile
Heat of binding 0 0 0–300 0–1000 100–3300 300–2200 1000–6000
in J/g H2O
Examples Wet surfaces of Wet sand Wet filter Starch gel, Solid surfaces in Crystalline glucose- Ca(OH)2
solids tissue Gelatin gel atmospheric air mono- hydrate
1
Water Activity
1.3 Adsorption Equilibrium 11

Definition
Fluids are: liquids and gases.

1.3 Adsorption Equilibrium

The term adsorption equilibrium refers to the steady state condition that is ultimately
reached when adsorption and desorption are going on at the same time. This is a
dynamic state, in which molecules are leaving the surface in a desorption process,
while other molecules are attaching themselves to the surface in an adsorption
process. Eventually, a dynamic equilibrium state is reached when the number of
molecules leaving the surface and those attaching themselves to the surface is the
same, and the number of molecules resting on the surface remains constant on
average. This is adsorption equilibrium. The various factors affecting the rate at
which this equilibrium can come about are discussed in the following subsections.
Another way to describe the behavior of water in foods is by the mechanisms of
molecular adsorption, corresponding to “bound water” and capillary adsorption
corresponding to “free water” described earlier. Molecular adsorption occurs under
very low water activity when water molecules adhere to specific points in
the molecular structure of the cell walls within the solid food material. When the
distance between the water molecule and the cell wall becomes small enough, the
force of attraction is large enough to draw the water molecule into the structure of the
cell wall. The force of attraction at such low moisture content is so high that an
“adsorption compression” results in a net decrease in volume of the solid-water
aggregate. As the moisture content increases, the molecular attraction lessens and
there is a volume increase, which is roughly equal to the volume of water added.
Because of the initial adsorption compression, however, the total volume of the
aggregate remains smaller than that of the sum of the constituents. Normally, it is
undesirable to bring water activity of food materials to such low levels that irrevers-
ible damage from adsorption compression will occur.
The extent and nature of the surface on which adsorption compression can take
place are likely the primary factors governing molecular adsorption. Molecular
attraction can be due to electronic and van der Waals attraction, but it is mostly
due to hydrogen bonding in the case of water in foods. Thus, the greater the number
of ionic or polar type molecules, the more water is held in the food material in this
form. Molecular adsorption is the primary cause of swelling in hygroscopic food
materials, such as starches.
At still higher moisture contents, where the vapor pressure has not yet reached the
saturation point, most of the available attraction sites have been filled with water
molecules, and further holding of water molecules is possible only through the
formation of “water bridges,” chains of water molecules extending between those
molecules which have been directly adsorbed.
12 1 Water Activity

In the case of solid food materials with non-porous surfaces when placed in
contact with a gaseous phase at a different relative humidity, adsorption or desorp-
tion will take place freely at the solid–fluid boundary surface, as described earlier.
During adsorption, gaseous molecules will be attracted to the solid surface and begin
diffusion to the interior once the surface becomes saturated with gas-phase
molecules. In the case of desorption, water molecules from within the solid phase
will be attracted to the surface, and freely escape into the gaseous phase once they
reach the surface. The degree to which the molecules disperse themselves about the
surface before escaping to the gaseous phase will depend upon the strength of
bonding at the surface. In the case of a non-porous surface with no interference
from porosity, this bonding strength is a function of the energy or enthalpy of
adsorption or heat of vaporization.
Both adsorption and desorption processes take place at the same time, but very
different rates, depending on the initial difference in relative humidity of the two
phases. At a constant temperature and partial pressure of the gaseous phase, equilib-
rium will be reached when the results from both processes compensate for them-
selves, and conditions at the boundary surface remain constant. If partial pressure of
the gaseous phase is increased, the equilibrium is disrupted, and adsorption will
begin once again until the surface becomes saturated with gaseous molecules
forming a complete monolayer of molecules. This is known as monomolecular
adhesion of a complete monolayer. If partial pressure is increased even further,
then further adsorption from the gaseous phase causes formation of multiple molec-
ular layers. These multiple layers are held by much weaker bonds and begin to give
rise to term “free water” referred to drying technology that can be most easily
evaporated. When only lower layers are present approaching only the monolayer,
then the binding forces are very strong giving rise to “bound water” that is difficult to
remove by evaporation. Table 1.5 shows how these levels of binding depend upon
the way the water is held within the material.
For each partial pressure p of a component, the corresponding interface assign-
ment m is set. The adsorption equilibrium is characterized by a fixed relationship
between interface assignment and partial pressure.
m = fkt ðpÞ

m mass of sorbate in kg, mg, μg

p partial pressure of sorptive in Pa

Definition
A substance coming from the gas phase that sorbs and condenses at an
interface is called an sorptive.
1.3 Adsorption Equilibrium 13

If the partial pressure is increased, the occupancy increases. Finally, the entire
solid surface is covered with a complete layer of the adsorpt. In this state, one speaks
of a monomolecular occupation of the interface. The solid interface is covered with a
monolayer of the gas-phase molecules.

Definition
A continuous layer of adsorpt molecules on an interface is called a monolayer.
This chapter discusses monolayer formed by water molecules.

If the partial pressure is increased further, further adsorption of molecules from

the gas phase takes place, whereby further layers of molecules settle on the mono-
layer, the so-called multiple layers. The adsorptive binding of molecules in the
multiple layers is significantly reduced compared to the monolayer. Molecules at
higher altitudes with a bonding heat close to zero behave similarly to free molecules
in the gas phase. If the adsorpt is water, this water is also referred to as “free
water.“This is to express that, in contrast to bound water (lower layers), these layers
have practically no binding to the solid. For drying technology and for the shelf life
of food, it is essential whether there is free or bound water and how strong this water
binding is. We come back to this in Sect. 1.6.
If the sorptive load at the interface is drawn graphically above the partial pressure
for a given temperature, a so-called sorption isotherm is obtained. In the sorption
isotherm, all equilibrium states are summarized at different partial pressures or
interface assignments. Sorption isotherms provide important information in the
field of drying and preservation of food. Some examples of simple sorption
isotherms are available in [15].
The influence of temperature on the sorption equilibrium is understandable from
kinetic gas theory: A higher temperature is synonymous with a higher speed or
average kinetic energy of the gas molecules. The relationship between temperature
and the velocity distribution of the molecules of an ideal gas is shown in the
appendix using Maxwell-Boltzmann’s equation. A higher energy leads to the fact
that a larger number of molecules can leave the solid surface contrary to the sorptive
bond and a new equilibrium state with a lower interface load is achieved. The
appearance of the sorption isotherm of a food thus changes when the temperature
is increased or decreased.
14 1 Water Activity

Definition
Sorption isotherm: Graphical representation of equilibrium sorbate load over
the partial pressure of the sorptive.

Following are examples of adsorption equilibria in processing technology:

• Activated carbon filter for the removal of pollutants from air

• Adsorption of oxygen to zeolite for air separation
• Drying of air by water vapor adsorption to silica gel
• Unpackaged icing sugar adsorbs water vapor from the ambient air.
• Nitrogen oxides are adsorbed on the catalyst surface of the combustion engine in
order to be degraded there.

Definition
A steady state process is a process in which the state variables are constant
over time.

1.4 Porous Surfaces

Solids with pores are called porous bodies. The porosity of a body is the ratio of the
pore volume to the total volume of the solid body.
VP
ε= ð1:14Þ
VK
ε porosity
VP volume of pores in m3
VK total volume of solid body in m3

Attention
Information such as porosity, water content, alcohol content, relative humidity
are often in %. Avoid mistakes by providing complete information such as %
(V/V ), Vol.%, wt.%, % RH, % (m/m).

Surfaces with porous structure contain voids that promote transport of water by
capillary absorption. Depending on the size and number of pores, the specific area of
a porous surface is significantly larger than that of the same mass with a smooth
surface. The quotient of area and mass of the sample is called specific surface area
Am.
1.4 Porous Surfaces 15

Table 1.6 Pore size clas- Term Pore radius in nm

ses according to IUPAC
Micropores <1
Mesopores 1–25
Macropores >25

A distinction is made between closed and open pores. While closed pores lie
inside the solid and have no opening to the gas phase, open pores have an opening to
the gas phase and thus provide additional space for adsorption. In contrast to the
outer surface of a solid, this additional interface created by open pores is called the
inner surface. For highly porous substances such as activated carbon, the outer
specific surface area Am = 3–5 m2∙g-1, but the inner specific surface area may be
over 1000 m2∙g-1 [13].
Capillary absorption occurs when voids in the cellular structure are of the size to
hold water in liquid form by forces of surface tension. Pore sizes can be classified
like in would Table 1.6 [16]. There are proposals for other classifications, e.g. to
distinguish between nanopores and micropores [17].
In the case of natural products and also in food, the non-porous case is rare. The
far more common case is the adsorption of molecules from the gas phase on porous
solid surfaces [18, 19]. Examples are the adsorption of water molecules from the
ambient air on the surfaces of baked goods or at the interfaces of powder particles.
The pore radii of a material are not uniform but are subject (as well as, e.g.,
particle sizes) to a size distribution [20]. Likewise, the pore shape is subject to
distribution. In micropores, there is a very high adsorption potential due to the small
distance between opposite pore walls. Micropores are almost always covered with
traces of adsorpt, only in high vacuum can adsorbed molecules also be removed from
micropores. In mesopores, capillary condensation occurs. The adsorpt (e.g., water)
deposited in these pores has a lower vapor pressure than the free adsorptive (e.g., free
water). Therefore, a gaseous component condenses in these pores already at a partial
pressure that is well below the saturation vapor pressure. This pressure, the vapor
pressure of the adsorpt in a capillary, can be calculated using Kelvin’s equation. For
a cylindrical pore with the pore radius rP is:
p 2  σ  Vm
ln = ð1:15Þ
p0 r P  RT

with

Vm V V 1
= = = ð1:16Þ
R n  Rs M m  Rs ρ  Rs
16 1 Water Activity

it is

p 2σ
ln = ð1:17Þ
p0 r P  ρ  R s  T

p vapor pressure above curved interface

p0 saturation vapor pressure
R universal gas constant in JK-1mol-1
Rs specific gas constant in JK-1kg-1
M molar mass of liquid in kg × mol-1
r density of the liquid in kg × m-3
T temperature in K
rP pore radius in m
n amount of liquid substance in mol
V volume in m3
Vm molar volume in m3∙mol-1

This is the Kelvin’s equation for calculating the relative vapor pressure of
droplets. Droplets have a convex-shaped surface whereas liquid surfaces in
capillaries have a concave-shaped surface. In mathematics the difference is the
sign of the radius only. That is why we have to put into Kelvin’s equation here
negative values of the cylindrical pores, e.g.
p 1:0682 nm p
ln = = - 0:10682 → = 0:899:
p0 - 10 nm p0

For water vapor as an adsorptive at 20 °C, it is

p 1 2σ 1 2  72:25  10 - 3 N  m - 1
ln =  = 
p0 r P ρ  Rs  T rP 999 kg  m 3  461:9 J  K - 1  kg - 1  293:15 K
-

so
p 1
ln =  1:0682 nm
p0 r P

Example
Relative water vapor pressure in cylindrical pores: For a pore with a diameter
of 20 nm, a radius of

r P = - 10 nm

(negative sign due to the concave curvature of the interface) is put into
Kelvin’s equation

(continued)
1.4 Porous Surfaces 17

p 1
ln =  1:0682 nm = - 0:10682
p0 - 10 nm
p
= e - 0:10682 = 0:89868
p0

The vapor pressure of water in a pore of 20 nm diameter is lowered to

89.9% of the vapor pressure of free water. For other pore radii at 20 °C it is:

rP in nm p/p0
1 0.344
5 0.808
10 0.899
50 0.979
100 0.989
1000 0.999

It becomes clear that vapor pressure reduction, and thus capillary condensation
plays a role in mesopores, i.e. in nanostructured surfaces. For this reason, a
mesoporous substance has a higher hygroscopicity than the corresponding
non-porous solid. Nanostructured surfaces can also be used to increase food safety
[21]. In the case of macropores, the effect of reducing vapor pressure is negligible.
Pore size and pore size distribution are determined using the so-called porosimetry
[20, 22].

Bottle or Flask-Shaped Pores

Reference to cylindrical-shaped pores in the previous section is only an ideal case to
help explain the mechanism of capillary adsorption. A way to explain hysteresis of
the sorption isotherm is to assume surfaces having pores with the shape of a bottle or
flask, whose opening has a considerably smaller radius than the bottom of the pore
(see Fig. 1.4) [23]. Recall that the relative vapour pressure required for capillary
adsorption will depend upon the pore radius. In the case of such bottle-shaped pores,
the smaller radius at the top of the pore will govern the vapour pressure needed for
the process of desorption (drying), in which water molecules must be drawn out from

Fig. 1.4 The capillary radius

on adsorption (I) is different
from that on desorption
(II) when a pore has the shape
of a bottle or flask, schematic

I II
18 1 Water Activity

Table 1.7 Four main types of sorption isotherms

I II III IV

m m m m

p p p p
Freundlich Langmuir BET/GAB BET/GAB with pores

the bottom of the pore. However, in the case of the adsorption process when the
material starts out in the dry form and the pores are initially empty of free water, the
water molecules adsorb at the larger radius at the bottom of the pore. This will
require a greater vapour pressure to reach the same level of moisture content during
desorption than was required during adsorption. This difference in vapour pressure
that is needed to reach the same level of moisture content depending on the direction
of the process (adsorption or desorption) is often given as a possible explanation for
the hysteresis observed in most sorption isotherms of food materials (graph IV in
Table 1.7). Sorption isotherms will be discussed at some length in the following
sections.

1.5 Sorption Isotherms

Sorption isotherms are graphical plots of the equilibrium between surface adhesion
forces and the partial pressure of the gaseous adsorbent at the boundary surface over
a range of partial pressures at a constant temperature. Four classic types of sorption
isotherms encountered in scientific studies are shown in Table 1.7 along with names
of the mathematical models that are used to characterize each type [15, 24].

Freundlich Model
The Freundlich model is given by Eq. (1.4) and is intended to characterize the
isotherm when it shows nearly no saturation behavior (Type I) when adhesion at the
boundary surface takes place. Since the Freundlich model is a simple power law
equation, taking logarithms of both sides will produce a linear Eq. (1.4) from which
the Freundlich constants (aF and bF) in the model can be obtained by linear
regression of a log-log plot of Eq. (1.5). The constant bF is taken from the slope of
the straight-line log-log plot, and the constant aF can then be found by substitution.
1.5 Sorption Isotherms 19

The Freundlich model is the model of choice when sorption isotherms are to be
analyzed in regions of very low partial pressure. However, when regions of higher
partial pressure are important, other models like the Langmuir and BET/GAB
models described in the following discussion are better suited.

m = aF  p b F ð1:18Þ

or

lgm = lgaF þ bF  lgp ð1:19Þ

m mass of adsorbent in kg
aF, bF Freundlich constants (0 < bF < 1)
p partial pressure in Pa

Langmuir Model
The Langmuir model focuses on characterizing the saturation behavior of the
sorption process. This is the region in the sorption isotherm where the curve tends
to flatten out (Type II). This region of the isotherm is often explained by realizing
that it normally covers the range of partial pressures over which molecular adhesion
at the boundary surface is a saturated monolayer. Under this condition a relatively
wide shift in partial pressure may produce relatively little change in molecular
adhesion at the surface. The model is based upon the assumption that in adsorption
equilibrium the rates of adsorption k and desorption k´ must produce an equal end
result, as reflected in Eqs. (1.6) and (1.7):

k  p  ð1- mÞ = k 0  m: ð1:20Þ
kp
m= : ð1:21Þ
k  p þ k0
Over the range of the curve it is

p
m = mmax , ð1:22Þ
pþb
where

M mass of adsorbent in kg
mmax maximum mass of adsorbent in kg
k, k′ rate constants
b Langmuir constant in Pa
p partial pressure in Pa

The Langmuir model starts out with the assumption of a homogeneous monomo-
lecular layer in which the adsorbent is held with maximum adhesion at the surface
20 1 Water Activity

Table 1.8 Adsorption General: mmmax = p

pþb
models and its range of
application For small p For middle p For higher p
mmax = pþb
p
m
mmax = 1
1þbp
m = mmax 1þ1 b m
p

≈ m = mmax  k 0  pbF ≈ = const:

m p m p
mmax b mmax p
m=kp m = aF  p bF m = mmax
Henry’s Law Freundlich model Langmuir model

mmax. The model parameters (b, mmax) can be determined by writing the equation in
the form of
1 1 b 1
= þ  : ð1:23Þ
m mmax mmax p
This is the equation of a straight line in a graph of m1 versus 1p. The slope mmax
b
and
1
intercept mmax will give the constants.
The saturation behavior described by the Langmuir model applies to gas and
liquid phase adsorption, particularly in the case of chemisorption (refer Fig. 1.3)
when the monomolecular layer covering the boundary surface cannot be exceeded. If
the monolayer were to become covered with additional molecular layers, this would
give rise to multilayer adsorption, as in the case of physisorption (refer Fig. 1.3). In
this case the Langmuir model would fail, and the Brunauer, Emmett, and Teller
(BET) model [25], described next, should be used instead. Table 1.8 gives an
overview of the applicability of these various models at increasing partial pressure.
Note that the Langmuir model changes into the Freundlich model as partial
pressures decrease from the intermediate range. At very low partial pressures the
Freundlich model becomes practically identical with Henry’s Law.

BET Model
Once multilayer molecular adsorption is reached, further increase in partial pressures
will cause the isotherm to depart from the relatively flat region characterized earlier
by the Langmuir model, and it will begin to increase dramatically reflecting the
weakening bonds of multilayer adhesion. When the isotherm is examined over the
full range of partial pressure, it will take on a sigmoid shape (Types III and IV in
Table 1.7). The best known and most widely used mathematical representation of the
complete adsorption phenomenon in biological materials is given by the BET
Eq. (1.10), after Brunauer, Emmett, and Teller (5), because it mathematically
characterizes the entire isotherm over all three regions:

p 1 C-1 p
= þ  ð1:24Þ
V  ð p S - p Þ V a  C V a  C pS
1.5 Sorption Isotherms 21

with

1
V =m ð1:25Þ
ρ
and the abbreviation

p
=φ ð1:26Þ
pS

it is
p
pS 1 C-1 p
= þ  ð1:27Þ
m  1- p ma  C ma  C pS
pS

respective

φ 1 C-1
= þ φ ð1:28Þ
m  ð1 - φÞ ma  C ma  C

where

p partial pressure in Pa
r density of the adsorbent in kg m-3
j relative partial pressure
Va volume of the monolayer in m3
ma mass of the monolayer in kg
C BET constant
m mass of adsorbent in kg
V volume of adsorbent in m3
ps saturation vapor pressure in Pa

The BET equation can undergo a linear transformation by creating a graphical

plot of m1  1 -φ φ against φ, as shown in Fig. 1.5. This produces a straight line having an
equation in the form y = a + b  φ.

Fig. 1.5 BET plot of

adsorbed mass versus relative
vapor pressure
22 1 Water Activity

The BET constant C can be obtained from the slope and intercept of this straight
line using the expression, C = ba þ 1, the monolayer mass ma from

1 1
ma = = : ð1:29Þ
a b
a þ1 aþb

The BET model [25] has a strong advantage over the Freundlich or Langmuir
models because of its validity over the full range of partial pressures. It has become a
very useful tool for estimating the volume or mass of the monolayer on the surface
boundary of powders or materials with porous surfaces with mesopore size. For
example, if we adsorb nitrogen onto a powder surface, the monolayer area can be
calculated from the known area of a nitrogen molecule AN 2 = 1.62 × 10-19 m2. If
water vapor is used as the adsorbent, then the relative partial pressure pp = φ takes on
S
the meaning of equilibrium relative humidity (or water activity, which will be
explained further.
When isotherm data points taken from an adsorption process are plotted on
the same graph as those taken from a desorption process on the same material at
the same temperature, the two isotherms will follow different pathways revealing the
hysteresis behavior shown in Type IV isotherms (Table 1.7). Sorption isotherm
hysteresis is far from being understood [5]. As explained earlier, the presence of
bottle-shaped pores with openings that are much narrower than their interior is a
possible cause of this type of hysteresis. Another possible cause is that irreversible
changes may occur in the food material during the adsorption/desorption
experiment.
Both desorption and adsorption isotherms can be characterized well by the BET
equation. The desorption isotherm relates to a drying process, in which a fresh food
sample with initial high moisture content is exposed to a gaseous surrounding of
lower and lower partial pressures. The desorption isotherm relates to a rehydration
process, in which an initially dry food sample is exposed to a surrounding fluid of
higher and higher partial pressure. Understanding the distinction between both cases
is a necessary prerequisite for undertaking the experimental procedures used to
obtain the sorption isotherm for a given material.
With the help of the BET theory, the surface area of solids can be determined. For
this purpose, nitrogen molecules are adsorbed on the surface, e.g., of a powder
sample, and the mass (of the adsorbed molecules) in the state of monolayer covering
is determined. From the determined mass, the number of adsorbed nitrogen
molecules can be calculated and from the known area requirement of a nitrogen
molecule (AN 2 = 1.6210-19 m2) the entire monolayer area. This gas adsorption
method for determining the specific surface area of porous or powdery solids is also
known as the BET method [26].
1.6 Moisture Content 23

GAB Model
The GAB model (after Guggenheim, Anderson, de Boer) is a semi-theoretical multi-
molecular adsorption model intended for use over a wide range of water activity and
can be written as a 3-parameter model:

aW 1 C-1
= þ  aW ð1:30Þ
m  ð1 - k  a W Þ m a  C  k maC

aW water activity
m mass of adsorbate in kg
k GAB correction factor (0.7–1.0)
C Guggenheim constant
ma mass of monolayer in kg

The GAB isotherm equation is an extension of the 2-parameter BET model which
takes into account the modified properties of the adsorbate in the multilayer region
and bulk liquid (“free” water) properties through the introduction of a third parame-
ter k. If k is less than 1, a lower sorption is predicted than that by the BET model and
allows the GAB isotherm to be successful up to water activity of 0.9. The GAB
equation reduces to the BET equation when k = 1. The constants in the GAB
equation (k and C) are temperature dependent and are the means by which we can
extract information when we construct and refer to sorption isotherms at different
temperatures.

1.6 Moisture Content

For general purposes, the moisture content of a food is normally expressed simply as
the percent moisture in the food substance. Mathematically, this is the ratio of the
mass of water contained in the food sample (adsorbent) over the total mass of food
sample containing the moisture (adsorbate), expressed as a percent. However,
moisture content used as the variable plotted on the vertical axis of vapor sorption
isotherms often is expressed as the ratio of mass of water mW (adsorbate) divided by
mass of dry matter mdm, only, (absorbent).
24 1 Water Activity

Table 1.9 Conversion of Dry basis (db) Wet basis (wb)

moisture content on dry
xW:db = mmdm
W
xW,wb = mmtotal
W
basis and wet basis
xW,db = mW
mtotal - mW xW:wb = mW
mdm þmW
Division by mW yields
xW,db = 1 1 - 1 xW,wb = 1
xW,db þ1
1
xW,wb

Possibilities to present the result

xW (db) xW (wb)
xW on dry basis xW on wet basis
xW in g H2O/g dry matter xW in g H2O/g total

Attention
Water content of a foodstuff can be expressed based on dry matter or based on
wet matter.

Based on dry matter Based on wet matter

xW = mass
mass water
of dry matter xW = massmass water
of total food
xW = mW
mdm xW = mW
mtotal

These two different methods for expressing moisture content in a food

sample are known as “wet basis” (wb) and “dry basis” (db), respectively. The
distinction between both methods of expression, as well as the ability to
quickly calculate one from the other must be well understood. The example
below shows a vertical bar representing a 100 g food sample composed of 20 g
water and 80 g dry matter and shows how these quantities are used to correctly
express moisture content on either basis. For example, in this case the moisture
content on a wet basis is 20/100 = 20%. But, on a dry basis it is 20/80 = 25%.
Refer also to Table 1.9 for further elaboration on the difference between the
two expressions and how to convert the quantities from one to another.
1.6 Moisture Content 25

Example
A food consists of 20 g of water and 80 g of dry matter.

water 20g

dry matter 80g

The water content of the food based on dry matter (db) is

20 g
= 0:25 = 25%
80 g

The water content of the food based on total weight (wb) is

20 g
= 0:20 = 20%
100 g

The water content of food is one of the most important quality parameters and is
of crucial importance for the production processes and the shelf life or safety of food.
In order to avoid misunderstandings when specifying the water content, fixed
designations have proven their worth. Table 1.9 shows some names and the conver-
sion of the information.

Attention
Water content and water activity are not the same thing. The relationship
between the two is given by the sorption isotherm.

The content of ingredients is often given in %. Examples are the water content or
the alcohol concentration of a food or beverage. Expressing these simply as a
percentage without specifying the basis upon which the percentage is calculated
will lead to confusion. To say that alcohol concentration of beer is 5.5%, or water
content of crispbread is 15%, it is unclear whether these are about volume or weight
percentages. In order to avoid misunderstandings, the water content of a food is
26 1 Water Activity

therefore often expressed with the addition of wt.%, mass percentage or (m/m). This
makes it clear that the reference value is a mass. In order to specify which mass is the
reference value (dry matter, db, or total weight, which includes the water content,
wb), the abbreviations, db and wb are added to the percentage number (see above)
(Table 1.10).
The expression for the water content of our crispbread is therefore:

xW = 15% (m/m) db or xW = 15 wt.% based on the dry matter.

When determining the water content or the dry matter content by drying methods
in the laboratory [27] the results are often reported on a wet basis (wb). The
conversion to dry basis (db) or vice versa is shown in Table 1.9 and is shown here
using two examples:

Example
Moisture content of corn flakes (dry basis).
A sample of corn flakes has a moisture content of 7.5% (wet basis). Express
this moisture content on a dry basis.

mW 7:5 kg 7:5 kg
xW,db = = = = 0:081
mtotal - mW 100 kg - 7:5 kg 92:5 kg

so

xW,db = 8:1 g H2 O=100 g dm = 8:1%ðdbÞ

or more simply

1
xW,db = = 0:081
1
0:075-1

Table 1.10 Percentages with different bases and examples of their use
Relative mass Relative volume Mole Relative Relative water vapor
fraction fraction fraction number pressure
xi = mmi xi = Vi
V
xi = nni xi = Ni
N
φ = pp
S

15% (m/m) 5% (V/V) 5% (n/n) 50% r.h.

15% by weight 5% by volume 40% by 45% by 50% relative humidity
mole number
1.7 Hygroscopicity 27

Example
Moisture of corn flakes (wet basis).
A sample of corn flakes has a moisture content of 21.2% (dry basis).
Express the moisture content on a wet basis.

mW 21:2 kg
xW,wb = = = 0:175
mdm þ mW 100 kg þ 21:2 kg

xW,wb = 17:5 g H2 O=100 g = 17:5%ðwbÞ

When we refer to moisture content in food sorption phenomena in this book, we

always mean moisture content on a dry basis.

1.7 Hygroscopicity

Hygroscopicity is a term used to describe how readily a material will take up

moisture when subjected to a given shift (change) in relative humidity. The term is
often used in industrial practice to indicate materials that quickly become problem-
atic in the presence of the least increase in surrounding relative humidity. For
example, certain powdered ingredients are said to very hygroscopic if they become
sticky and cause caking problems that prevent free flowing from the storage hopper
in a food process.
In this regard the term is used in a relative qualitative sense, and no quantitative
scale has ever been assigned to hygroscopicity. We have no means to quantify what
is the difference between strong and weak hygroscopic behavior. It may be possible
to consider hygroscopicity as a material property that could be quantified if we
understand that this type of moisture uptake behavior in response to change in
relative humidity can be observed from the shape of the sorption isotherm for this
material. For example, a highly hygroscopic powder will show a much higher uptake
of water when the relative humidity surrounding the powder is shifted from 45% to
75% than would a powder with lower hygroscopicity.
To help better understand this, consider the four different sorption isotherms
shown in Fig. 1.6. Each isotherm belongs to a different material, and all four

Fig. 1.6 Exposing four

materials exhibiting different
degrees of hygroscopicity to a
shift in water activity from
aW1 to aW2 . The same shift
causes different amounts of
water uptake Δx
28 1 Water Activity

materials exhibit different degrees of hygroscopicity. Let us assume that we shift the
equilibrium relative humidity (water activity) from water activity aW1 to water
activity aW2 on all four materials. Then, we can observe the different quantities of
water uptake experienced by each material as an indicator of different degrees of
ΔxW
hygroscopicity. The mean slope of the sorption isotherm in this range is Δa W
.
It clearly can be seen that material 2 will have greater water uptake than material
1. The slope of the secant in the segment of the curve between the starting point and
final point of the shift in water activity can be taken as a water uptake potential (wP-
value) of the given material. The wP-value serves only as an indicator of the moisture
difference Δx, which occurs on a given change ΔaW. It does not take into account the
level of moisture already absorbed into the material earlier. So, when we examine
Fig. 1.6 once again, we would have to note that although material 1 is a very good
sorbent, when we make our shift in water activity aW1 → aW2 experiment, material
2 will show the higher water uptake. Therefore, we can say that material 2 has a
greater water wP than material 1 in the range of water activities of our experiment.
When we use wP as an indicator of hygroscopicity, the points aW1 and aW2 have to be
clearly specified. Also, a constant temperature was assumed throughout this
example.
For infinitesimally small changes in water content or water activity, the secant
slope (mean slope) of the sorption isotherm turns into the tangent slope:

ΔxW dxW
W P = lim = : ð1:31Þ
aw → 0 ΔaW daW

The hygroscopicity wP of a material can thus be specified by the tangent slope of

the sorption isotherm in a specified point (aW/xw).

Bottom Line
Hygroscopicity of a material is the slope of the water vapor adsorption
isotherm for that material. For a comparison of materials, it is necessary to
specify in which state (aW, xw, T, p) the comparison should be made.

Example
When water activity is increased from 0.45 to 0.75, a substance shows an
increase in water content from 20% (m/m) db to 40% (m/m) db. The hygro-
scopicity is

ΔxW 0:4 - 0:2

WP = = = 0:66 ðdbÞ
ΔaW 0:75 - 0:45
1.8 BET Equation for Foods 29

This example provides the hygroscopicity in a range aW = 0.45 to 0.75 or water

content between 20 and 40% (m/m) db, i.e. an average value for this part of the
sorption isotherm. It also shows the need for conventions and indications such as
“water content related to dry matter” or “water content related to weighting.”

1.8 BET Equation for Foods

In the case of water vapor sorption in foods, when the isotherms are obtained as plots
of moisture content (db) versus water activity, the BET equation gets the form
1 aW 1 C-1
 = þ a ð1:32Þ
x W 1 - aW xW,a  C xW,a  C W

aW water activity
xW moisture content (db) in kg/kg dry matter
C BET constant
xW, a monolayer moisture content (db) in kg/kg dry matter

If the quantity x1W  1 -aWaW were plotted against water activity aW, a straight line is
obtained (see Fig. 1.7). The monolayer moisture content xW,a as well as the BET
constant can then be obtained from the slope and intercept of the line. The value of
xW,a is usually the moisture content at which the water is tightly bound with water
molecules in a single monolayer, and it cannot participate as a solvent. Thus, it is the
moisture content that should be reached for maximum stability of dehydrated foods
(see also Sect. 1.12).
The evaluation of the plot is performed again by obtaining the intercept and slope
of the straight-line curve in the diagram.
C-1
intercept : a = 1
xW,a C slope : b = xW,a C

Because of ba = C - 1 or C = ba þ 1 the moisture content of the food at the point

where a complete monolayer exists is

Fig. 1.7 BET plot of

adsorbed moisture content
versus water activity

1 aw
⋅
x 1 − aw

aw
30 1 Water Activity

1 1
xW,a = = : ð1:33Þ
a b
aþ1 aþb

For the water activity at reaching the monolayer is [28] because of

xW ðaW,a Þ = xW,a :
1 aW,a 1 C-1
 = þ a : ð1:34Þ
xW,a 1 - aW,a xW,a  C xW,a  C W,a
p
C-1
aW,a = : ð1:35Þ
C-1
The monolayer moisture content xW,a and BET constant C are the only two
parameters needed to describe the sorption isotherm for a given food product.
Published tables or on-line databases listing values of these parameters for various
food stuffs can be found in references [29–33]. Also, sorption enthalpies needed for
estimating energy requirements in the engineering design of drying or dehydration
processes can be obtained from the Arrhenius temperature dependency of these
parameters.

Example
BET—isotherm for a food product:
The moisture content (dry basis) of the product in equilibrium with differ-
ent relative humidity is listed below: Construct the BET plot from these data,
and calculate the values for the monolayer moisture content xW,a and BET
constant C.

φ/% r.h. xW/% (db)

0 0.0
10 2.5
20 4.3
30 5.2
50 8.3
75 18.8

First we prepare a table of values; then we draw the BET diagram (Fig. 1.8):

φ/% r.h. xW/% (db) xW (db) aW

aW
1 - aW
1
xW  1 -aWaW
0 0.0 0 0 0 –
10 2.5 0.025 0.1 0.11 4.4
20 4.3 0.043 0.2 0.25 5.8
30 5.2 0.052 0.3 0.43 8.3
50 8.3 0.083 0.5 1.0 12.0
75 18.8 0.188 0.75 3.0 16.0
1.8 BET Equation for Foods 31

For the slope, we get b = 19.0 and for the intercept a = 2.4. So the
monolayer moisture content is

1 1
xW,a = = = 0:0467
a þ b 2:4 þ 19:0
And the BET constant C is
b 19:0
C= þ 1= þ 1 = 8:92:
a 2:4
p
C-1
The water activity with monolayer reached is aW,a = C-1

p
8:9 - 1
aW,a = = 0:25:
8:9 - 1

BET: One-Point Method

For the purpose of rough estimation, the straight-line BET plot can be approximated
by constructing a straight line from the origin of the coordinate axes through a single
data point. The single data point should be taken at value of water activity at which
the monolayer is fully developed (saturated) but no multiple layer formation exists.
As a rule of thumb for most foods, this value is normally chosen in the range of water
activity between 0.3–0.4. The BET parameters for the monolayer moisture content
xW,a and BET constant C can be estimated from the slope of this straight line and will
normally serve sufficiently well for most purposes of approximating the profile of
the sorption isotherm for the given food substance, as shown by Eq. (1.32).

Fig. 1.8 Evaluation of the 20

BET diagram: The slope of
the straight line provides the
value of the constant b, and
the intercept provides the 10
value for the constant a 1 aw
⋅
x 1 −aw

0.1 0.5 1
aw
32 1 Water Activity

1
xW,a = : ð1:36Þ
b
Because of a = 0 a value for C is not needed in the BET—one-point method

Example
Estimation of the monolayer water content of a food using the BET—one-
point method:
A food product has a moisture content of 5.2% (db) in equilibrium with air
at a relative humidity of 30% r.h. We construct the one-point BET straight line,
and estimate the monolayer moisture content xw,a for this food.
With that single point we get the following:
1 aW 1 0:3 0:3
 =  = = 0:3  27:5 = 8:25
x w 1 - aW 0:052 1 - 0:3 0:036

φ/% r.h. xW/% (db) xW (db) aW

aW
1 - aW
1
xW  1 -aWaW
30 5.2 0.052 0.3 0.43 8.2

Therefore, without diagram we see the slope will be b = 8:25

0:3 = 27:5.
The monolayer water content is xW,a = b = 27:5 = 0:036 = 3:6%ðm=mÞ db.
1 1

The same result we get by xW  (1 - aW). It is 0.052  0.7 = 0.036.

The BET model has its strengths in the range of monomolecular adsorption
(around aW = 0⋯0.45). For higher water activities we have adsorption of multiple
layers and here the GAB model is more accurate.

1.9 GAB Equation for Food

The GAB model is most appropriate when we deal with water activities above
0.40.5. A small disadvantage of this model is the increased mathematical com-
plexity in dealing with a 3-parameter equation compared to the BET model. It is
1 k  aW 1 C-1
 = þ  k  aW ð1:37Þ
x W 1 - k  aW xW,a  C xW,a  C

or

1 aW 1 C-1
 = þ a : ð1:38Þ
xW 1 - k  a W xW,a  C  k xW,a  C W

aW water activity
xW moisture content (db) in kg∙kg-1 dry matter
(continued)
1.9 GAB Equation for Food 33

C Guggenheim constant
k GAB correction factor (0.7–1.0)
xW, a monolayer moisture content (db) in kg∙kg-1 dry matter

By knowing the GAB parameters k,C and xW,a, complete sorption isotherms of a
material can be drawn or aW – xW – values can be calculated. To calculate the water
content of a food at a given water activity, we use the GAB equation in the form
xW,a  C  k  aW
xW = : ð1:39Þ
ð1 - k  aW Þð1 þ ðC - 1Þ  k  aW Þ

or

xW,a  C  k  aW
xW = : ð1:40Þ
ð1 - k  aW Þð1 - k  aW þ C  k  aW Þ

By taking experimental data of xW versus aW and bringing them into a function

like Eq. (1.37) we can determine the coefficients a, b, g of the polynomial, and from
it we can calculate the GAB parameters [2].

aW
= α  a2W þ β  aW þ γ ð1:41Þ
xw
with

k 1
α= -1 : ð1:42Þ
xW,a C
1 2
β= 1- : ð1:43Þ
xW,a C
1
γ= : ð1:44Þ
xW,a  C  k

So we get the GAB parameters

1
1 2
xW,a = : ð1:45Þ
β2 - 4α  γ
2  α  xW,a
k= - : ð1:46Þ
β  xW,a þ 1
1
C= : ð1:47Þ
xW,a  γ  k

With the aid of appropriate computer software, these parameters can be calculated
within a few seconds for daily routine use of the BET or GAB models. With the use
34 1 Water Activity

of such software and appropriate data input, food engineers and scientists can
quickly determine the information needed to design and specify optimum
processing, packaging, storage, and handling conditions with respect to water
activity requirements for shelf-life stability of foods.

1.10 Sorption Enthalpy

Up to this point, we had only looked at a single temperature or kept the temperature
of our samples constant. Let’s now go one step further and take sorption isotherms of
a food at different temperatures. The evaluation of the BET parameters shows that
they depend on the temperature. If the BET constant C is drawn logarithmically
above the reciprocal value of the absolute temperature (the so-called Arrhenius
diagram), a straight line is obtained, from whose slope the excess enthalpy of
adsorption, or monolayer-bonding enthalpy ΔhC, can be determined. The slope
(m) of the straight line on the graph of log C over 1/T is:
ΔhC
m= - ð1:48Þ
2:3  Rs
with

ΔhC = Δhs - Δhvap ð1:49Þ

respective

Δhs = ΔhC þ Δhvap ð1:50Þ

ΔhC specific bonding enthalpy of H2O monolayer in J kg-1 (desorption)

Δhvap specific vaporization enthalpy H2O in J kg-1 (desorption)
Δhs specific sorption enthalpy H2O monolayer in J kg-1 (desorption)
T absolute temperature in K
Rs specific gas constant of H2O
kJK - 1 mol - 1
Rs = MR = 8:314
18:02 gmol - 1
= 461:4 J  g - 1

The specific sorption enthalpy Δhs, mono will be found at the drying condition
which leaves only the monolayer, but fully intact. At this point the monolayer is a
complete layer of water molecules tightly bound to the boundary surface of the food
material, and the enthalpy at this point marks the distinction between “free” and
“bound” water in the food. As explained earlier, the BET theory covers a monolayer
only and there is no regard for multilayers (like in the GAB model, see below). That
means water in the second layer is treated like “free” water. So in the BET theory the
“excess” monolayer-bonding enthalpy ΔhC disappears and plays no further role. In
this case, water is freely available to behave as normal water (vaporize, condense,
freeze, sublime and thaw), and only the enthalpy of vaporization (in the case of
desorption) or condensation (in the case of adsorption) applies. In the opposite
1.10 Sorption Enthalpy 35

direction when the “excess” monolayer-bonding enthalpy ΔhC takes on high values,
the bonding strength at the monolayer surface becomes very strong. Further water
removal beyond this point through normal drying processes becomes very costly and
time consuming. Therefore, it is of critical importance to identify this point of
distinction with respect to the “excess” monolayer-bonding enthalpy ΔhC in order
to design and specify optimum drying, storage, and packaging conditions to assure
long-term stability of dehydrated foods.
A fast method for quickly estimating this “excess” monolayer-bonding enthalpy
ΔhC is possible by comparing water activity of a food at different temperatures and
determining the temperature dependency of water activity for the given food. This
can be done with the Clausius-Clapeyron equation:
∂ ln pS ΔH Δh
= = : ð1:51Þ
∂T RT 2 RS T 2

∂ ln pS Δhvap
= : ð1:52Þ
∂T Rs T 2
here is

∂ ln pS  aW ΔhS
= : ð1:53Þ
∂T Rs T 2

∂ ln pS ∂ ln aW Δhvap ΔhC
þ = þ : ð1:54Þ
∂T ∂T Rs T 2 Rs T 2

∂ ln aW ΔhC
= : ð1:55Þ
∂T Rs T 2

ΔhC
d ln aW =  dT: ð1:56Þ
Rs T 2

aW1 T1
ΔhC
∂ ln aW =  ∂T: ð1:57Þ
Rs T 2
aW2 T2

ΔhC 1 1
ln aW2 - ln aW1 = -  - : ð1:58Þ
Rs T2 T1

lgaW2 - lgaW1 ΔhC

=- : ð1:59Þ
1
T2 - 1
T1
2:3  Rs

or
36 1 Water Activity

ΔðlgaW Þ ΔhC
=- : ð1:60Þ
ΔðT1 Þ 2:3  Rs

i.e., the slope of the lg aW over 1/T diagram is

ΔhC
m= - : ð1:61Þ
2, 3  Rs
The net isosteric sorption enthalpy is

ΔhC = - 2, 3  m  Rs ð1:62Þ

because of

ΔhS = ΔhC þ Δhvap : ð1:63Þ

we distinguish between net isosteric sorption enthalpy ΔhC and total sorption
enthalpy ΔhS.
The isosteric enthalpy of sorption ΔhC determined in this way (net isosteric heat
of sorption) applies to a defined water content xW of the food. If one determines for
different water contents xW, it shows that ΔhC with decreasing water content
increases strongly. A closer look at the isosteres ln aW above 1/T shows that it is
only approximately a linear function. In fact, however, the slope is slightly depen-
dent on the temperature.

Example
Starch powder monolayer sorption enthalpy.
From two sorption isotherms taken at different temperatures we read the
water activities at moisture content xW = 10% (db):

ϑ/°C aW
30 0.25
80 0.45

We transform these data to.

-1
i ϑ/°C T/K T =K
1
aW log aW
1 30 303.15 3.29910-3 0.25 -0.602
2 80 353.15 2.83210-3 0.45 -0.347

From equation

(continued)
1.10 Sorption Enthalpy 37

ΔðlgaW Þ ΔhC
=-
Δ T1 2, 3  Rs

we get

Δ log aW
ΔhC = - 2:3  Rs 
Δ T1

or

lgaW2 - lgaW1
ΔhC = 2:3  Rs 
T2 - T1
1 1

so

- 0:34678 - ð - 0:6020Þ
ΔhC = - 2:3  0:461 kJ  K - 1  kg - 1 
2:832  10 - 3 K - 1 - 3:299  10 - 3 K - 1
that is

ΔhC = 579 kJ  kg - 1 :

The result is a positive enthalpy. It means that ΔhC = 579 kJ  kg-1 has to
be introduced into the material to the sorption binding strength of the mono-
layer molecules. To bring the water molecules form the monolayer phase into
the gaseous phase according to Eq. (1.13) we need the total enthalpy of

ΔhS = ΔhC þ Δhvap = 579 kJ  kg - 1 þ 2200 kJ  kg - 1 = 2779 kJ  kg - 1 :

When we calculate a monolayer sorption enthalpy based on two points only (here
30° and 80 °C) we should not overestimate the precision of the result. Although in
this example we got the information that the heat needed for desorption of the
monolayer molecules is about 20 to 30% higher than for free water:
ΔhC 579
= = 0:26:
Δhvap 2200

If the enthalpy is replaced by the specific free enthalpy Δg = Δh - T  Δs

(specific Gibbs enthalpy) in the Arrhenius plot, the temperature dependence can be
recognized.
ΔðlgaW Þ Δh - T  Δs
=- :
Δ T1 2:3  Rs

Then from the slope of the Arrhenius plot (lg aW over 1/T ), then we get:
38 1 Water Activity

ΔgC
m= -
2:3  Rs
and from that specific Gibbs enthalpy

ΔgC = ΔhC - T  ΔsC

This makes it possible to determine the change in the specific entropy of sorption
ΔsC in addition to the specific enthalpy of sorption ΔhC. Negative values for ΔgC
indicate that there is an exergetic process, i.e. that the sorption takes place
spontaneously.
Similarly, the temperature dependence of the BET constant C and the
Guggenheim correction factor k can be determined to obtain information about the
sorption process [34]. Unlike the BET model, the GAB model takes into account the
fact that when multi-molecular layers begin to develop, adsorbed water molecules do
not instantly behave as free water. But instead, bonding forces gradually weaken
with increasing multilayers, such that vapor pressure is slightly reduced from that of
pure water. This is taken into account by the parameter k, which is used as a
multiplier coefficient that serves as a correction factor (with values between 0.7
and 1.0) in order to take this behavior into account. The BET model neglects this
effect. From the temperature dependency of the GAB parameter k, we can determine
the excess enthalpy from bonding of the multilayers:
ΔhC
C = C0  eRs T ð1:64Þ

respectively
Δh
k = k 0  eRs T ð1:65Þ

Bottom Line
The drying of a food is the reduction of the water content and thus also the
water activity. The sorption isotherm provides information on what the water
activity must be at a specified water content. The shelf life is usually higher the
lower the water activity of the dried food. For deviations from this rule,
see text.

Therefore, the specific sorption enthalpy of the monolayer can be determined

from the temperature dependency of the constants (C, C′) in the BET equation. The
“excess” monolayer-bonding enthalpy ΔhC disappears and plays no further role in
the case of multilayer adsorption. In this case, other values of ΔhC apply depending
on the adhesive bonding of the second, third, and additional multiple layers. With
each additional layer, less and less bonding energy (enthalpy) is needed until the
adsorbed water becomes freely available to behave as normal water (vaporize,
1.11 Other Models 39

condense, freeze, sublime, and thaw), and only the enthalpy of vaporization
(or condensation) applies.
Knowledge of sorption enthalpy gives indications of the binding strength, as well
as the type of bond between water and the solid food. It is therefore helpful in
estimating the shelf life of food or the design of necessary packaging. In addition, the
sorption enthalpy provides information for specifying or designing the most appro-
priate drying processes. Since the focus when drying solid foods is often on the
removal of multilayers, the GAB model is preferable here.

Bottom Line
During desorption, i.e. drying, the isosteric enthalpy of water increases: at the
beginning of drying, we have to apply the isosteric enthalpy of water in
different multilayers, then of water in the monolayer.

1.11 Other Models

Several other models to mathematically describe the profile of sorption isotherms in

foods have been proposed and reported in the literature [15, 17]. They differ in
regard to their physical approach, method for mathematical derivation, and number
of parameters. Tables 1.11, 1.12, and 1.13 contain a listing of these other models
organized by the number of parameters required of each model. Models with only
two parameters are listed in Table 1.11, 3-parameter models are listed in Table 1.12,
and 4-parameter models in Table 1.13. If water-rich foods are treated approximately

Table 1.11 Two-parameter models for sorption isotherms

Freundlich (1906) xW = A  aBW
Langmuir (1916) xW = xmax aW
aW þB

Smith (1947) xW = A + B  (ln(1 - aW))

Oswin (1946) B
xW = A aW
1 - aW
1
Henderson (1952) lnð1 - aW Þ B
xW = -A
xW,a CaW
Brunauer, Emmett, Teller (1938) (BET) xW = ð1 - aW Þð1þðC - 1ÞaW Þ
C-1
1
xW  1 - aW = xW,a C þ xW,a C
aW 1
 aW
1
Halsey (1948) B
xW = - A
ln aW

Chung, Pfost (1967) xW = - 1

B ln ln aW
-A
1
Iglesias, Chirife (1978) ln xW þ ðxW 2 þ x0:5W Þ2 = A  aW þ B
Lewicki (2000) B-1
xW = A  1
aW -1 -1
40 1 Water Activity

Table 1.12 Three-parameter models for sorption isotherms

Cubic model xW = p1 þ p2  aW þ p3  a2W þ p4  a3W
xW,a CkaW
Guggenheim (1966), Andersen (1946), De Boer (1953) xW = ð1 - kaW Þð1 - kaW þCkaW Þ
(GAB) aW
xW = α  aW þ β  aW þ γ  C
2
GAB formulated by Bizot:
Chung, Pfost (1967)—modified [36] ln aW ðTþC Þ
xW = - 1
B ln -A
1
Henderson (1952)—modified [36] lnð1 - aW Þ B
xW = - AðTþCÞ
1
Halsey (1948)—modified [36] B
xW = - A
ln aW

Oswin (1946)—modified [36] C

x W = ðA þ B  T Þ aW
1 - aW

Table 1.13 Four-parameter models for sorption isotherms

Peleg (1993) xW = A  aCW þ B  aD
W
DLP (double log xW = A  ln - ln a3W þ B  ln - ln a2W þ C  lnð- lnðaW ÞÞ þ D
polynomial)
Isse (1993) ln xW = A  ln aW
1 - aW þ B, q > 0, q < 0

like aqueous solutions, the sorption isotherm can also be calculated on the basis of
Raoult‘s law [35]:

nW
aW = : ð1:66Þ
nW þ nS
1
aW = : ð1:67Þ
1 þ nnWS

xW moisture content (db) in kg∙kg-1 dry matter

aW water activity
xm monolayer moisture content (db) in kg∙kg-1 dry matter
x0.5, m moisture content in kg∙kg-1 (db) at aW = 0.5
A, B, C, D constants

Definition
A model is a representation of reality. It is a simplified reproduction of an
effect, behavior, or object in which the properties considered essential are
emphasized and the aspects considered incidental are disregarded in order to
simplify understanding. A mathematical model is a mathematical expression

(continued)
1.12 Shelf Life of Food Related to Water Activity 41

of the relationship between variables that is capable of predicting the outcome

for given values of the variables. Because of the simplifications, there are
always limitations to the applicability of a model.

1.12 Shelf Life of Food Related to Water Activity

Water plays an important role in the shelf life of food. For spoilage caused by
microbial activity and biochemical reactions, water activity—not the water content
of the food—plays the decisive role. The reason for this is that water activity
characterizes the availability of water molecules, while water content is a sum
parameter for the water it contains, both bound and free water. For this reason,
knowledge of the water activity of a food is important and for this reason sorption
isotherms are needed, from which the relationship between water activity and water
content of a food can be seen.
In addition to the kinetics of microbial inactivation, kinetics of quality-degrada-
tion reactions such as enzymatic tanning or non-enzymatic browning are also
dependent on water activity. In addition to numerous other factors, water activity
also influences the kinetics of oxidation processes. Figure 1.9 schematically shows
the dependence of the speed of some spoilage reactions on water activity. If you
move on the abscissa of Fig. 1.9 from right to left, it can be seen that with decreasing
water activity, the speed of the spoilage reactions decreases and these can be
gradually brought to a standstill. An exception is oxidation, where the speed of the
reaction increases again as soon as the monolayer cover is undercut (Fig. 1.9). This
phenomenon is explained by the fact that the monolayer performs a protective

Fig. 1.9 Relative rate (Vrel) of different spoilage reactions as a function of water activity aW in
food. 1 lipid oxidation, 2 non-enzymatic browning reactions, 3 hydrolytic reactions, 4 enzymatic
reactions, 5 molds, 6 yeasts, 7 bacteria. The dashed line indicates the sorption isotherm of the
sample material. From [43]
42 1 Water Activity

function against atmospheric oxygen. As soon as the monolayer is “damaged,” this

protective function is no longer given. For this reason, the knowledge of the
monolayer water content, as a lower limit during drying, is also important for fatty
foods.
In addition to water activity, the glass transition temperature of a food is a critical
parameter for the kinetics of conversion processes in food. The glass transition is a
so-called phase transition of the second order, it is characterized in that the substance
shows a solid–liquid transition without absorbing melting enthalpy. Such a phase
transition occurs in substances that are non-crystalline solidified. This solid,
non-crystalline phase is also called amorphous phase or glass. A glass can also be
understood as a solidified liquid, i.e. as a substance with an extremely high viscosity
that drops extremely when passing through the glass transition temperature.
Numerous solid foods show glass transitions. Typical examples are confectionery
such as sweets, which are created by solidifying highly concentrated solutions.
Many powders created by rapid drying processes (spray drying, roller drying, freeze
drying) and frozen foods created by rapid cooling processes show glass transitions.
But pasta and baked goods can also contain non-crystalline-solid areas and are then
prone to physicochemical changes when the glass transition temperature is
exceeded. Undesirable changes of this kind are the loss of flow properties in
powders, clumping in bulk solids, the stickiness of carbohydrate-rich solids, or the
delayed crystallization of ingredients.
When passing through the glass transition temperature, the fluidity of the material
and the mobility of the molecules in the substance increase sharply, and the water
activity also increases by leaps and bounds. For this reason, the speed of quality-
reducing reactions can increase by leaps and bounds at this temperature.
In order to exclude such transformations in affected foods, the temperature of the
food should remain well below its glass transition temperature. For this reason,
knowledge of the glass transition temperature is important for solid foods.
The glass transition temperature depends on the composition of the solid, i.e. on
the recipe. Here, the water content of the food again plays a major role. An initial
estimation of the glass transition temperature of a simple system can be made after
with the help of the composition and the glass transition temperatures of the pure
components [5]:
1 xi
= ð1:68Þ
Tg i
T g,i

x mass fraction
Tg glass transition temperature in K
k constant of Gordon Taylor equation
1.12 Shelf Life of Food Related to Water Activity 43

Example
Estimation of the glass transition temperature of a sugar product from 90% (m/
m) sucrose and 10% (m/m) water:

1 xi x x
= = W þ S
Tg i
T g,i T g,W T g,S

1 0:1 0:9 1
= þ = 0:0033697 =
T g 138:15 K 340:15 K 296:8 K

ϑg = 23:6 ° C

A higher water content thus leads to a reduction in the glass transition temperature
and can thus limit the stability of a food or its shelf life. On the other hand, the
transfer of a system to a “safe” glass condition can be used to maintain quality.
Details on the glass condition, on phase diagrams of food [37–40] and their experi-
mental measurement are discussed in Chap. 8.

Bottom Line
In the case of solid substances, a distinction is made between the crystalline
state and the glassy state. Crystalline substances pass into the liquid phase at
their melting point, glass-like substances soften as soon as the glass transition
temperature is reached. The stability and shelf life of food thus depend on the
glass transition temperature, among other factors. The glass transition temper-
ature, in turn, depends on the water content of the food, i.e. the adsorption of
water influences the glass transition temperature.

In practice, the shelf life of food is the time between harvest or production and
intended consumption. In European law, the minimum shelf life is understood to
mean the period of time within which “the food” retains its specific properties if
stored correctly [41]. According to the European Food Information Regulation, the
minimum shelf life is shown in the form of a date that is affixed to the packaging.
This best-before date thus characterizes the period of time guaranteed by the
manufacturer after packaging, within which the food retains its specific properties,
i.e. does not show excessively high quality losses.
44 1 Water Activity

Bottom Line
The minimum shelf life of a food is the length of time it can remain “on the
shelf” before any quality degradation can occur that would compromise
consumer acceptance. Manufacturers of a food product will stamp a “best
by” date on the package serving as a guarantee, that with proper storage, the
desirable properties of the food remain preserved until that date. With perish-
able food products in which microbial spoilage is of greatest concern, shelf life
is the length of time within which such a food can be safely consumed when
held under proper storage conditions. Dried or intermediate moisture foods
have sufficiently low water activity that microbial growth is not possible
allowing them to have much longer shelf life.

The length of a specified shelf life depends on the type and speed of the microbial
and quality-degradation reactions taking place in the food. A distinction is made
between microbiological, chemical, biochemical and physical processes that lead to
a loss of quality. Often these reactions are also summarized as spoilage-causing. If
you want to roughly divide food into shelf life classes, you can differentiate between
shelf life in days, weeks, and years. While, e.g., untreated beverages such as milk or
fruit juices (aW = 0.99) have a shelf life of only a few days, jams (aW = 0.85) can be
stored for weeks or months and low-fat biscuits or rusks (aW = 0.2) can be stored for
years.
However, shelf life depends not only on water activity, but also on the type and
number of microorganisms present and the storage temperature. As we know, the
shelf life of milk or fruit juices can be increased to several weeks by thermal
inactivation of the microorganisms (pasteurization, sterilization), while the water
activity remains at its value of 0.99. On the other hand, the shelf life of food can be
reduced by storing it under improper conditions, such as being left unpackaged or
held at too high a temperature. If biscuits or rusks are stored unpacked in ambient air,
they absorb water from the atmosphere and their water activity increases making
them stale. Depending on the water activity that occurs, spoilage reactions can occur,
which reduce the shelf life of the baked goods from a few months or years to only a
few weeks or days. From these examples, we can see that shelf life is not determined
solely by water activity. However, the water activity of a food is a prerequisite for a
number of spoilage reactions. The value of the water activity of a non-sterilized food
is thus an indication of the shelf life, i.e. the time in which the food loses its specific
properties due to spoilage, unless technical measures are taken against it, such as
cooling or pasteurization. Table 1.14 shows such indicative values for the shelf life
of foods of different water activity.
Microbial spoilage reactions are the multiplication of microorganisms (e.g.,
bacteria, fungi) and the associated formation of microbial metabolic products or
toxins. Chemical and biochemical spoilage reactions include fat oxidation,
1.12 Shelf Life of Food Related to Water Activity 45

Table 1.14 Rule of thumb Water activity Shelf life of food amounts to few...
values for the shelf life of
aW > 0.95 Days
foods of different water
activity, explanations in aW ≈ 0.85 Weeks
the text aW < 0.75 Months
aW < 0.65 Years
aW < 0.6 Decades

Table 1.15 Rule of thumb values of water activity for propagation of microorganisms [4]
aW minimum water activity for propagation of...
0.97 Clostridium botulinum E, Pseudomonas fluorescens
0.95 Escherichia coli, Clostridium perfringens, Salmonella spp., Vibrio cholerae
0.94 Clostridium botulinum A, B
0.93 Bacillus cereus
0.92 Listeria monocytogenes
0.91 Bacillus subtilis
0.90 Staphylococcus aureus (anaerobe), Saccharomyces cerevisiae
0.86 Staphylococcus aureus (aerobe)
0.80 Penicillium citrinum, Saccharomyces bailii
0.77 Aspergillus niger
0.61 Saccharomyces rouxii
<0.60 Nor propagation

hydrolysis, enzymatic and non-enzymatic reactions such as tanning reactions. Phys-

ically measurable quality reductions are, for example, the deterioration of a desired
texture, viscosity, elasticity, powder flowability, color or acoustic properties. An
essential principle of food preservation is the slowing down of such spoilage
processes. The objectives here are the safety of the consumer and the preservation
of the highest possible quality.
With decreasing water activity, the optimal growth conditions for specific
microorganisms are gradually undercut, so that the toxin formation of
microorganisms and finally their multiplication can be reduced or inhibited. This is
an essential principle of preservation by drying food. Table 1.15 lists reference
values for the minimum water activity for the growth of specific microorganisms.
As a rough classification, it can be said that below aW = 0.90 most bacteria cease to
multiply, below 0.88 most yeasts and below 0.80 most molds. However, specialized
microorganisms (e.g., halophilic bacteria, xerophilic mold, osmophilic yeasts) can
be active at aW = 0.6 - 0.75.
Labuza [42] published a widely-referenced hypothetical plot illustrating relative
spoilage reaction rates as a function of water activity, see Fig. 1.9.
Table 1.16 shows typical values of water activity for various food products. Since
the water activity depends on the raw material composition and the recipe, as well as
on the temperature and the storage conditions, the stated values may only be
46 1 Water Activity

Table 1.16 Typical water aW Food

activities of food
0.98–0.99 Lemonade
(examples, average data),
from [4] 0.97–0.99 Juices
0.99 Fish, fresh
0.97 Egg
0.96–0.97 Liver sausage
0.98–0.99 Beef
0.94–0.95 Margarine, unsalted
0.93–0.96 Mayonnaise
0.94–0.96 Bread
0.93–0.94 Ketchup, mustard
0.82–0.95 Butter, salted
0.95–0.98 Cheese, cheddar
0.82–0.94 Marmalade
0.84 Blueberry muffins
0.72 Pepper, powder
0.72-0.80 Dried fruits
0.63–0.64 Marshmallows
0.6 Milk chocolate
0.68–0.70 Wheat, cornels
0.48–0.6 Honey
0.34–0.36 Oatmeal
0.35–0.38 Corn flakes
0.37–0.39 Pasta, spaghetti, dry
0.51–0.54 Resins
0.5 Rice, dry
0.5–0.6 Coffee beans, green
0.1–0.3 Coffee beans, roasted
0.20–0.27 Skimmed milk powder
0.227 Sucrose, crystal
0.11–0.12 Ritz cracker
0.07 Popcorn

understood as indicative values. More detailed information can be found, e.g., in

[4, 44, 45].
It can now be apparent how all the different relationships discussed in this chapter
begin to fit together. Water activity dictates whether or not a reaction will take place
in a food, and if so, at what rate. Therefore, the objective of most food dehydration
processes is to bring the food product to a specified water activity. However, the
objective of the engineer responsible for the design and operation of the food
dehydration process is to achieve a target level of moisture content. Food drying
equipment is designed only to remove moisture, and only moisture content can be
measured and controlled on the factory floor (not water activity). The sorption
1.12 Shelf Life of Food Related to Water Activity 47

Fig. 1.10 Generic sorption

isotherm showing main
regions of water bonding

xw

I II III
0.2 0.4 0.6 0.8 1
aw

isotherm becomes the fundamental tool by which it is possible to specify the

moisture content needed that will assure the required water activity.
The shape of the sorption isotherm also helps us to know what to expect about the
quality of physical characteristics of a food product as a function of water activity.
For example, we often divide a sorption isotherm into the three regions of water
bonding states, as shown in Fig. 1.10.
Region I is the region of very lowest water activity and moisture content. In this
region, the very small amount of water in the material is completely bound within the
monolayer, and unavailable to contribute any kind of moisture behavior to the
material. Most food materials would be completely shrunken, hard, and brittle in
this state. It is also very difficult and costly to attempt to remove sufficient water by
drying processes to reach such low levels of water activity. For these reasons, we
often avoid bringing foods into region I of their sorption isotherms.
Region II is the region in which the water activity is sufficiently low that most
food spoilage reactions cannot proceed. This is the region in which the monolayer is
completely saturated, but no free moisture is available for supporting any chemical,
biochemical, or biological activity. Therefore, this is the region into which we bring
dehydrated foods for long-term storage stability, such as food powders and dry
snacks. These dry foods normally have a crispy or crunchy texture.
Region III is the region in which moisture uptake occurs most rapidly as
multilayers are formed and the moisture content reaches the levels in which the
water can act freely once again to support chemical and biological reactions. This is
also the region in which the food material will exhibit the textural properties of being
“fresh” with soft to firm elastic properties while being moist and swollen (fresh fruit
and vegetables).
Some foods called “intermediate moisture” foods are dehydrated to a carefully
controlled intermediate level of water activity in the region of transition between
regions II and III. The precise level of water activity is chosen to be sufficiently low
so that no free moisture is available for supporting any spoilage reactions, but
sufficiently high to still give a soft flexible texture. These foods are ready to eat
and require no rehydration (beef jerky, salami, etc.). At the same time they are shelf
stable for long periods of storage just as any dehydrated foods.
48 1 Water Activity

1.13 Laboratory Determination of Sorption Isotherms

In order to establish sorption isotherms, it is necessary to measure the water activities

of a sample at different moisture contents (or vice versa). Numerous analytical
methods exist for the determination of moisture content [46]. The standard method
is the gravimetric determination of the moisture loss during drying. This is the
so-called oven drying method [47]. In this method, a preweighed sample of fresh,
moist food is placed in a drying oven and withdrawn after specified time intervals. At
each time interval, the sample is weighed, and the difference in weight from the
initial fresh sample is the amount of moisture that was lost up to that time interval.
Once no further weight change is observed, the sample has reached equilibrium and
drying has ended. This final weight at equilibrium is the weight of bone-dry solid
matter that was contained in the initial fresh sample. By knowing the weight of solid
matter in the sample, as well as the total sample weight, the moisture content can be
calculated. For the purpose of plotting moisture content against water activity in
establishing a sorption isotherm, moisture content must be calculated on the dry
basis and reported as % moisture dry basis (db), (see Table 1.9). The easiest way to
determine water activity, aW, is to measure the relative humidity of the atmosphere
that is in equilibrium with the food. This is done by placing the sample in a closed
system in which the relative humidity of the trapped air surrounding the sample is
measured or known (Fig. 1.1).

Bottom Line
Procedure for determining a point of the sorption isotherm:

– Place the sample in a sealable vessel with humidity sensor

– Keep temperature constant, wait for equilibrium adjustment
– Reading equilibrium humidity = water activity of the sample
– Take a sample
– Determine the water content of the sample (e.g., by means of a drying oven)

A classic method for determining the sorption isotherm is the desiccator method.
Here, the sample is placed in a desiccator jar with defined humidity and temperature
(Fig. 1.11). The sample is weighed at intermittent time intervals until no further
weight change occurs. At that point, the sample has come into equilibrium and has a
water activity equal to the relative humidity of the atmosphere in the chamber. The
moisture content can be determined from the final weight of the sample that was
reached at the point of equilibrium.
The defined desiccator atmosphere is created by placing a selected saline solution
in the desiccator whose water activity is known (Raoult’s Law). By using a saturated
saline solution, the concentration of the liquid phase is known at a given tempera-
ture (Table 1.17). This eliminates the need for an exact weighing of a salt, it only has
to be ensured that there is always a visible sediment of undissolved salt.
1.13 Laboratory Determination of Sorption Isotherms 49

Fig. 1.11 Taking a point on

the sorption isotherm by the
desiccator jar method. The
sample is placed in a closed
chamber maintained at a
constant known relative
humidity by a saturated salt
solution at the bottom of the
chamber

The advantage of this classic method is that we only have to carry out water
content determinations and do not need a humidity sensor. However, the disadvan-
tage without a humidity sensor is that you have no information about the equilibrium
situation. Reaching the equilibrium can take a few h to a few days, depending on the
weight and sample conditions. The desiccator method is therefore limited to foods
with a higher shelf life, i.e. lower water activity.
In order to arrive at meaningful measured values for the sorption isotherm, it is
important to keep the temperature constant during the equilibrium adjustment or the
water activity measurement and to ensure that the samples do not change during the
measurement. To measure the desorption isotherm, samples are used whose equilib-
rium adjustment has been achieved by water release. Adsorption isotherms are
obtained by using samples that have reached their equilibrium value by water
absorption.
For more convenient recording of sorption isotherms, climate chambers are used
instead of the desiccator jars, in which the samples are located at a defined tempera-
ture and humidity. In addition, there are climate chambers with built-in analytical
balances to determine the water content of the samples simultaneously. With the help
of suitable measuring programs, complete sorption isotherms and sorption isosteres
can be automatically recorded with these devices. Such measuring methods, in
which the water vapor pressure over the sample is changed over time, are used as
dynamic vapor sorption (DVS) methods [48], also dynamic isopiestic method (DIM)
[49] or dynamic dewpoint isotherm (DDI). Isopiestic means that there is an equilib-
rium between the water vapor pressure of the sample and the atmosphere (ίσoς,
Greek: equal, πιστóς, Greek: pressure).
There are numerous metrological approaches to record water contents and water
activities analytically in a time-saving manner using various methods to carry out
water vapor sorption measurements, as follows:

• Cryoscopy: One tries to induce the water activity of the solution from the
solidification temperature of an aqueous solution [52].
• Dew point measurements: The relative humidity of the atmosphere under consid-
eration can be calculated from the temperature at which water vapor from the air
condenses into liquid water (dew point).
 50

Table 1.17 Water activity standards

Saturated aqueous Water activity aW at

solution of: From 10 °C 15 °C 20 °C 25 °C 30 °C 35 °C
NaOH [50] – 0.0957 ± 0.028 0.0891 ± 0.024 0.0824 ± 0.021 0.0758 ± 0.017 0.0692 ± 0.015
Sodium hydroxide
LiCl [50] 0.1129 0.1130 0.1131 0.1130 0.1128 0.1125
Lithium chloride ± 0.0041 ± 0.0035 ± 0.0031 ± 0.0027 ± 0.0024 ± 0.0022
ZnCl2 [51] – – 0.119 – – –
Zinc chloride
K-CH3CO2 [50] 0.2338 0.2340 0.2311 0.2251 0.2161 –
Potassium acetate ± 0.0053 ± 0.0032 ± 0.0025 ± 0.0032 ± 0.0053
MgCl2 [50] 0.3347 0.3330 0.3307 0.3278 0.3244 0.3205
Magnesium chloride ± 0.0024 ± 0.0021 ± 0.0018 ± 0.0016 ± 0.0014 ± 0.0013
K2CO3 [50] 0.4314 0.4315 0.4316 0.4316 0.4317 –
Potassium carbonate ± 0.0039 ± 0.0033 ± 0.0033 ± 0.0039 ± 0.0050
MgNO3 [50] 0.5736 0.5587 0.5438 0.5289 0.5140 0.4991
Magnesium nitrate ± 0.0033 ± 0.0027 ± 0.0023 ± 0.0022 ± 0.0024 ± 0.0029
NH4NO3 – – 0.66 0.66 – –
Ammonium nitrate
NaCl [51] 0.7567 0.7561 0.7547 0.7529 0.7509 0.7487
Sodium chloride ± 0.0022 ± 0.0018 ± 0.0014 ± 0.0012 ± 0.0011 ± 0.0012
KCl [50] 0.8677 0.8592 0.8511 0.8434 0.8362 0.8295
1

Potassium chloride ± 0.0039 ± 0.0033 ± 0.0029 ± 0.0026 ± 0.0025 ± 0.0025

KNO3 [50] 0.9596 ± 0.014 0.9541 0.9462 0.9358 0.9231 0.9079
Potassium nitrate ± 0.0096 ± 0.0066 ± 0.0055 ± 0.0060 ± 0.0083
Water Activity
1.14 Standard for Sorption Isotherms 51

• Dielectric spectroscopy: By measuring the microwave permittivity the water

content of the sample is measured [53] and by this water activity is estimated [54].
• Optical methods: The determination of water activity by laser absorption is
described [55].
• Isopiestic method: This is based on the refractometry of polyol droplets that are in
equilibrium with the atmosphere above the food [56, 57].

In order to be able to compare different methods for the measurement of sorption

isotherms or isosteres, it is necessary to precisely define the boundary conditions, the
models used, the selected statistical parameters, and the comparison criteria [58].

1.14 Standard for Sorption Isotherms

In order to validate the accuracy of any given laboratory procedure or analytical

method for measuring a physical or chemical property, most laboratories rely upon
testing their procedures with a “standard” sample to see if the result is in accordance
with the expected known outcome for the standard. In the case of water vapor
sorption isotherms in foods, the recommended standard is microcrystalline cellulose
(MCC). This material is widely available to the scientific community under the trade
name “Avicel” and possesses a signature isotherm, that is published in books
physical properties of foods [59]. Figure 1.12 shows the sorption isotherm for this
standard material. Equation (1.69) gives the mathematical model for this isotherm
when the numerical value of the constants A and b are taken as 0.0540 and -0.4343,
Respectively (Avicel PH1 at 25 °C).

Definition
A standard is a universally accepted way of doing something. In analytics, the
term is used not only for standard procedures but also for reference materials.

Fig. 1.12 Sorption isotherm 15

of microcrystalline cellulose
(25 °C, adsorption) (from
[59]) 10

xw in % (db)
5

0
0 0.2 0.4 0.6 0.8 1.0
aW
52 1 Water Activity

Table 1.18 Predicted and measured data of Avicel

aW xW in % (db) predicted xw in % (db) measured (from [59])
0.1 2.80 2.0 ± 0.5
0.2 2.96 3.0 ± 0.5
0.4 4.53 5.0 ± 0.5
0.6 6.44 7.0 ± 0.5
0.8 9.86 10.0 ± 0.5
0.9 14.02 –

b
1
xW = A -1 : ð1:69Þ
aW

The following table compares model predicted with experimentally measured

moisture content of a sample of Avicel (Table 1.18).

Further Reading

Deep-fried food: effect of water sorption on glass transition and texture [60]
Freeze-dried food: effects of water on glass transition and texture [61]
Powdered milk: sorption isotherms and thermodynamic properties [62]
Lactose crystallization by water sorption [37]
Effect of water on the flowability of hygroscopic powders [63]
Orange juice powder—sorption isotherm and net isosteric heat of sorption [64]
Blueberries: state diagram, sorption isotherm [65]
Moisture sorption characteristics and modeling of energy sorghum (Sorghum [36]
bicolor)
Apricots: influence of sugar composition on water sorption isotherms and glass [66]
transition
Orange (Citrus sinensis) peel and leaves: experimental and mathematical [67]
investigation of moisture sorption isotherm
Potato: determination of sorption isotherm and isosteric heats of sorption [68]
Apple: GAB model for water activity and moisture content during osmotic [69]
dehydration
Fruits and vegetable: simplified calculation of sorption isotherms [70]
Pepper: sorption isotherm [71]
Starch: water sorption isotherm [72]
Leguminose: simplified algorithm for the prediction of water sorption isotherms [73]
Starch: adsorption and desorption isotherms [74]
Beef and pork: desorption isotherms [75]
Cacao: Sorption isotherm and isosteric enthalpy [76]
Strawberry powder, freeze-dried: critical water activity [77]
Meat: dielectric spectroscopy for prediction of water activity [54]
Use of polyols to measure equilibrium relative humidity [56]

(continued)
References 53

Cellulose: dynamic vapor sorption [7]

Critical water activity as indicator for powder stability [78]
Amorphous lactose/whey protein systems: water sorption-induced crystallization [37]
Banana, dried: sorption isotherm [79]
High pressure inactivation of Bacillus cereus—effect of water activity [80]

Summary
The shelf life of food is strongly dependent on water activity. Water activity is
linked to moisture content by use of a sorption isotherm, which graphically
describes the relationship between moisture content of a food and its water
activity. Isotherms are unique to any one food material and are therefore a
fundamental physical property of food materials. In this chapter, the creation
of such isotherms and the mathematical modeling of them is demonstrated
using examples. With the help of the sorption isotherm, the hygroscopicity of
materials can be quantitatively specified. Finally, application examples are
listed, which can be used for further studies and as suggestions for your own
investigations

References
1. Isengard HD (2017) Water - only a simple molecule? Food Chem 236:1. https://doi.org/10.
1016/j.foodchem.2017.05.061
2. Lewicki PP (2004) Water as the determinant of food engineering properties. A review. J Food
Eng 61(4):483. https://doi.org/10.1016/S0260-8774(03)00219-X
3. Al-Muhtaseb AH, McMinn WAM, Magee TRA (2002) Moisture sorption isotherm
characteristics of food products: a review. Food Bioprod Process 80(2):118. https://doi.org/
10.1205/09603080252938753
4. Barbosa-Cánovas GV, Fontana AJ, Schmidt SJ, Labuza TP (2008) Water activity in foods:
fundamentals and applications. Blackwell Publishing, Hoboken, NJ. https://doi.org/10.1002/
9780470376454
5. Walstra P (2003) Physical chemistry of foods. Marcel Dekker, New York. https://doi.org/10.
1201/9780203910436
6. Atkins PW, De Paula J (2011) Physical chemistry for the life sciences, 2nd edn. W. H. Freeman,
Oxford
7. Driemeier C, Mendes FM, Oliveira MM (2012) Dynamic vapor sorption and thermoporometry
to probe water in celluloses. Cellulose. https://doi.org/10.1007/s10570-012-9727-z
8. Cazier J-BB, Gekas V (2007) Water activity and its prediction: a review. Int J Food Prop 4(1):
35–43. https://doi.org/10.1081/JFP-100002187
9. van der Sman RGM (2017) Predicting the solubility of mixtures of sugars and their replacers
using the Flory-Huggins theory. Food Funct 8(1):360
54 1 Water Activity

10. Maneffa AJ, Stenner R, Matharu AS, Clark JH, Matubayasi N, Shimizu S (2017) Water activity
in liquid food systems: a molecular scale interpretation. Food Chem 237:1133. https://doi.org/
10.1016/j.foodchem.2017.06.046
11. Multon LJ, Bizot H (1978) Intermediate moisture foods and water activity determination. Ann
Nutr Aliment 32:631
12. Walstra P (2002) Physical chemistry of foods. Marcel Dekker, New York
13. Römpp H, Falbe J (2014) Römpp-Lexikon Chemie. Thieme, Stuttgart
14. Bauer KH, Frömming KH, Führer C (1999) Pharmazeutische Technologie. Deutscher
Apotheker Verlag, Stuttgart
15. Limousin G, Gaudet JP, Charlet L, Szenknect S, Barthes V, Krimissa M (2007) Sorption
isotherms: a review on physical bases, modeling and measurement. Appl Geochem 22(2):
249. https://doi.org/10.1016/j.apgeochem.2006.09.010
16. IUPAC (1972) Manual of symbols and terminology for physicochemical quantities and units.
Butterworths, London
17. Mays TJ (2007) A new classification of pore sizes. Stud Surf Sci Catal 160:57
18. Gueven A, Hicsasmaz Z (2013) Pore structure in food: simulation, measurement and
applications. Springer, New York
19. Aouaini F, Knani S, Ben Yahia M, Ben Lamine A (2015) Statistical physics studies of
multilayer adsorption isotherm in food materials and pore size distribution. Phys A Stat Mech
Appl 432:373. https://doi.org/10.1016/j.physa.2015.03.052
20. Lawrence M, Jiang Y (2017) Porosity, pore size distribution, micro-structure. Springer,
New York, pp 39–71. https://doi.org/10.1007/978-94-024-1031-0_2
21. Srividya N, Ghoora MD, Padmanabh PR (2017) Antimicrobial nanotechnology: research
implications and prospects in food safety. In: Grumezescu AM (ed) Nanotechnology in the
agri-food industry. Academic Press, New York, pp 125–165. https://doi.org/10.1016/b978-0-
12-804303-5.00004-3
22. Krantz WB, Greenberg AR, Kujundzic E, Yeo A, Hosseini SS (2013) Evapoporometry: a novel
technique for determining the pore-size distribution of membranes. J Membr Sci 438(0):153.
https://doi.org/10.1016/j.memsci.2013.03.045
23. Zdravkov B, Čermák J, Šefara M, Janků J (2007) Pore classification in the characterization of
porous materials: a perspective. Open Chem 5(2):385. https://doi.org/10.2478/s11532-007-
0017-9
24. Van den Berg C, Bruin S (1981) Water activity and its estimation in food systems: theoretical
aspects. In: Rockland Stewart GFLB (ed) Water activity: influences on food quality. Academic
Press, New York, pp 1–61
25. Brunauer S, Emmett PH, Teller E (1938) Adsorption of gases in multimolecular layers. J Am
Chem Soc 60:309
26. Safefood 360° (2014) Water activity (aw) in foods - White Paper. https://safefood360.com/free-
resources/whitepapers/
27. Matissek R, Fischer M, Steiner G (2018) Lebensmittelanalyik. Springer, Berlin. https://doi.org/
10.1007/978-3-662-55722-8
28. Igathinathane C, Womac AR, Sokhansanj S, Pordesimo LO, Womac AR (2007) Moisture
sorption thermodynamic properties of corn Stover fraction. Trans ASABE 50(6):2151
29. Timmermann EO, Chirife J, Iglesias HA (2001) Water sorption isotherms of foods and
foodstuffs: BET or GAB parameters? J Food Eng 48(1):19. https://doi.org/10.1016/S0260-
8774(00)00139-4
30. Iglesias HA, Chirife J (1982) Handbook of food isotherms: water sorption parameters for food
and food components. Academic Press, San Diego. https://doi.org/10.1016/B978-0-12-
370380-4.50005-0
31. Rao MA, Rizvi SSH, Datta AK (2005) Engineering properties of food. Food Science and
Technology, vol 144, 3rd edn. CRC Press, Boca Raton, FL
References 55

32. Nesvadba P, Houška M, Wolf W, Gekas V, Jarvis D, Sadd PA, Johns AI (2004) Database of
physical properties of agro-food materials. J Food Eng 61(4):497. https://doi.org/10.1016/
S0260-8774(03)00213-9
33. Rao MA, Rizvi SSH, Datta AK (2005) Engineering properties of foods. CRC Press, Boca
Raton, FL
34. Chirife J, Timmermann EO, Iglesias HA, Boquet R (1992) Some features of the parameter k of
the GAB equation as applied to sorption isotherms of selected food materials. J Food Eng 15(1):
75. https://doi.org/10.1016/0260-8774(92)90041-4
35. Lewicki PP (2000) Raoult’s law based food water sorption isotherm. J Food Eng 43(1):31.
https://doi.org/10.1016/S0260-8774(99)00130-2
36. Bonner IJ, Kenney KL (2013) Moisture sorption characteristics and modeling of energy
sorghum (Sorghum bicolor (L.) Moench). J Stored Prod Res 52:128. https://doi.org/10.1016/
j.jspr.2012.11.002
37. Fan FH, Mou T, Nurhadi B, Roos YH (2017) Water sorption-induced crystallization, structural
relaxations and strength analysis of relaxation times in amorphous lactose/whey protein
systems. J Food Eng 196:150. https://doi.org/10.1016/j.jfoodeng.2016.10.022
38. Telis VRN, Tonon RV (2017) The effects of non-equilibrium states and storage conditions on
glass transitions in food. Woodhead Publishing, Sawston, UK, pp 379–403. https://doi.org/10.
1016/b978-0-08-100309-1.00020-1
39. Fan F, Roos YH (2017) Structural strength and crystallization of amorphous lactose in food
model solids at various water activities. Innov Food Sci Emerg Technol 40. https://doi.org/10.
1016/j.ifset.2016.06.011
40. Roos YH, Drusch S (2015) Phase transitions in foods, 2nd edn. Woodhead Publishing,
Sawston, UK. https://doi.org/10.1016/C2012-0-06577-5
41. LMIV - Lebensmittelinformationsverordnung (2011)
42. Labuza TP, Fu B (1993) Growth kinetics for shelf-life prediction: theory and practice. J Ind
Microbiol Biotechnol 12(3):309
43. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, Munich
44. Bhandari B, Bansal N, Zhang M, Schuck P (2013) Handbook of food powders: processes and
properties. Woodhead Publishing, Cambridge UK. https://doi.org/10.1533/9780857098672
45. Belitz H-D (1999) Food chemistry: with 531 tables, 2nd edn. Springer, Berlin
46. Isengard HD (2010) Water in food - the story continues. Food Chem 122(2):387. https://doi.org/
10.1016/J.Foodchem.2010.02.056
47. Matissek R, Steiner G (2014) Lebensmittelanalytik. Springer, Berlin
48. Domian E, Brynda-Kopytowska A, Ciesla J, Gorska A (2018) Effect of carbohydrate type on
the DVS isotherm-induced phase transitions in spray-dried fat-filled pea protein-based powders.
J Food Eng 222:115
49. Rahman MS, Al-Belushi RH (2006) Dynamic isopiestic method (DIM): measuring moisture
sorption isotherm of freeze-dried garlic powder and other potential uses of DIM. Int J Food Prop
9(3):421
50. Greenspan L (1977) Humidity fixed points of binary saturated aqueous solutions. J Res Natl Bur
Stand 81(1):89
51. Rahman S (2009) Food properties handbook. CRC Press, Boca Raton, FL. https://doi.org/10.
1201/9781420003093
52. Esteban MA, Marcos A, Fernández-Salguero J (1987) Cryoscopic approach to water activity
measurement of non-liquid foods: application to cheese. Food Chem 25(1):31. https://doi.org/
10.1016/0308-8146(87)90051-3
53. Blakey RT, Morales-Partera AM (2016) Microwave dielectric spectroscopy – a versatile
methodology for online, non-destructive food analysis, monitoring and process control. Eng
Agric Environ Food 9(3):264. https://doi.org/10.1016/j.eaef.2016.02.001
54. Traffano-Schiffo MV, Castro-Giraldez M, Colom RJ, Fito PJ (2015) Study of the application of
dielectric spectroscopy to predict the water activity of meat during drying process. J Food Eng
166:285. https://doi.org/10.1016/j.jfoodeng.2015.06.030
56 1 Water Activity

55. Allan M, Mauer LJ (2017) Dataset of water activity measurements of alcohol:water solutions
using a Tunable diode laser. Data Brief 12:364. https://doi.org/10.1016/j.dib.2017.04.017
56. Steele RJ (1987) Use of polyols to measure equilibrium relative humidity. Int J Food Sci
Technol 22:377
57. Al-Khalili M, Al-Habsi N, Rahman MS (2021) Application of dynamic temperature-humidity
chamber for measuring moisture sorption isotherms of biomaterials as compared to the conven-
tional isopiestic method. Adsorp Sci Technol 2021:1236427. https://doi.org/10.1155/2021/
1236427
58. Labuza TP, Acott K, TatiNl SR, Lee RY, Flink J, McCall W (1976) Water activity determina-
tion: a collaborative study of different methods. J Food Sci 41(4):910. https://doi.org/10.1111/j.
1365-2621.1976.tb00751_41_4.x
59. Spiess WEL (1984) The water-vapour sorption isotherms of microcrystalline cellulose (MCC)
and of purified potato starch. Results of a collaborative study. J Food Eng 3:51
60. Jothi JS, Ebara T, Hagura Y, Kawai K (2018) Effect of water sorption on the glass transition
temperature and texture of deep-fried models. J Food Eng 237:1. https://doi.org/10.1016/j.
jfoodeng.2018.05.014
61. Sogabe T, Kawai K, Kobayashi R, Jothi JS, Hagura Y (2018) Effects of porous structure and
water plasticization on the mechanical glass transition temperature and textural properties of
freeze-dried trehalose solid and cookie. J Food Eng 217:101. https://doi.org/10.1016/j.jfoodeng.
2017.08.027
62. Abdenouri N, Idlimam A, Kouhila M (2010) Sorption isotherms and thermodynamic properties
of powdered milk. Chem Eng Commun 197(8):1109. https://doi.org/10.1080/
00986440903412936
63. Juarez-Enriquez E, Olivas GI, Zamudio-Flores PB, Ortega-Rivas E, Perez-Vega S, Sepulveda
DR (2017) Effect of water content on the flowability of hygroscopic powders. J Food Eng 205:
12. https://doi.org/10.1016/j.jfoodeng.2017.02.024
64. Edrisi Sormoli M, Langrish TAG (2015) Moisture sorption isotherms and net isosteric heat of
sorption for spray-dried pure orange juice powder. LWT Food Sci Technol 62(1 Part 2):875.
https://doi.org/10.1016/j.lwt.2014.09.064
65. Vásquez C, Díaz-Calderón P, Enrione J, Matiacevich S (2013) State diagram, sorption isotherm
and color of blueberries as a function of water content. Thermochim Acta 570(0):8. https://doi.
org/10.1016/j.tca.2013.07.029
66. Djendoubi Mrad N, Bonazzi C, Boudhrioua N, Kechaou N, Courtois F (2012) Influence of
sugar composition on water sorption isotherms and on glass transition in apricots. J Food Eng
111(2):403. https://doi.org/10.1016/j.jfoodeng.2012.02.001
67. Kammoun Bejar A, Boudhrioua Mihoubi N, Kechaou N (2012) Moisture sorption isotherms –
experimental and mathematical investigations of orange (Citrus sinensis) peel and leaves. Food
Chem 132(4):1728. https://doi.org/10.1016/j.foodchem.2011.06.059
68. McLaughlin CP, Magee TRA (1998) The determination of sorption isotherm and the isosteric
heats of sorption for potatoes. J Food Eng 35(3):267. https://doi.org/10.1016/S0260-8774(98)
00025-9
69. Prothon F, Essen T, Ahrné LM (2004) Application of the Guggenheim, Anderson and De Boer
model to correlate water activity and moisture content during osmotic dehydration of apples. J
Food Eng 61(3):467. https://doi.org/10.1016/S0260-8774(03)00119-5
70. Murgante B (2012) Computational science and its applications. In: ICCSA 2012: 12th interna-
tional conference, Salvador de Bahia, Brazil, June 18–21, 2012, Proceedings. Part II. Springer,
Berlin
71. Kaymak-Ertekin F, Sultanoglu M (2001) Moisture sorption isotherm characteristics of peppers.
J Food Eng 47(3):225
72. Al-Muhtaseb AH, McMinn WAM, Magee TRA (2004) Water sorption isotherms of starch
powders: part 1: mathematical description of experimental data. J Food Eng. https://doi.org/10.
1016/S0260-8774(03)00133-X
References 57

73. Moreira R, Chenlo F, Torres MD (2009) Simplified algorithm for the prediction of water
sorption isotherms of fruits, vegetables and legumes based upon chemical composition. J
Food Eng 94(3):334. https://doi.org/10.1016/j.jfoodeng.2009.03.026
74. Al-Muhtaseb AH, McMinn WAM, Magee TRA (2004) Water sorption isotherms of starch
powders. J Food Eng. https://doi.org/10.1016/S0260-8774(03)00133-X
75. Delgado AE, Sun D-WW (2002) Desorption isotherms for cooked and cured beef and pork. J
Food Eng 51. https://doi.org/10.1016/S0260-8774(01)00053-X
76. Akmel DC, Kakou KE, Kone KY, Assidjo NE, Kouame P (2015) Desorption isotherms and
isosteric heats of fermented cocoa beans (Theobroma cocoa). J Food Res 4(3):138. https://doi.
org/10.5539/jfr.v4n3p138
77. Mosquera LH, Moraga G, Martinez-Navarrete N (2012) Critical water activity and critical water
content of freeze-dried strawberry powder as affected by maltodextrin and arabic gum. Food
Res Int 47(2):201. https://doi.org/10.1016/j.foodres.2011.05.019
78. Carter BP, Galloway MT, Campbell GS, Carter AH (2015) The critical water activity from
dynamic dewpoint isotherms as an indicator of pre-mix powder stability. J Food Meas Charac
9(4):479. https://doi.org/10.1007/s11694-015-9256-1
79. Yan Z, Sousa-Gallagher MJ, Oliveira FAR (2008) Sorption isotherms and moisture sorption
hysteresis of intermediate moisture content banana. J Food Eng. https://doi.org/10.1016/j.
jfoodeng.2007.10.009
80. Xu J, Janahar JJ, Park HW, Balasubramaniam VM, Yousef AE (2021) Influence of water
activity and acidity on Bacillus cereus spore inactivation during combined high pressure-
thermal treatment. LWT 146:111465. https://doi.org/10.1016/j.lwt.2021.111465
Mass Density
2

The mass density of a material is the ratio of mass to volume of the material. It can be
thought of as the relative “heaviness” of a material. The exact term “mass density” is
used to recognize the difference in the term “density” as a material property having
mass, versus “density” used with other phenomena, such as energy density or
population density. Mass density of a material is a fundamental physical property
that must be known and quantified in order to design, specify, or size materials
handling and processing equipment, as well as for process control and quality
characterization. Because of the wide spread need for accurate measurement of
density in the food industry, many different methods have been developed for
different types of food and beverages. Some of these are unique to specific
industries.
In order to calculate the density of a material, we must be able to measure the
mass and volume of a material sample or “body” as accurately as possible. There-
fore, we will first cover the basics of measuring weight or mass.

Definition
Density is the mass density of a material, i.e. the ratio of mass and volume of a
material.

2.1 Weighing and Mass

In physics the mass measures for the inertia and gravity of a body. The SI unit of the
mass is the kilogram (kg). The definition of the kilogram based on fundamental
constants is:

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 59

L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_2
60 2 Mass Density

h
Δν
1 kg = 1:475521
1040
:
c2
The unit kg is one of the seven basic units of the International System de Unite
(SI). Seven basic units are defined with the help of seven fundamental constants. The
use of fundamental constants for the definition of the kilogram allows to have a
worldwide kg standard which is not based on a haptic kilogram prototype any longer
[1]. In Chap. 17 (appendix), the SI units are explained in more detail.
Gravity of a body is caused by the gravitational force between that body and
planet earth. Gravitational force is proportional to its gravitational constant, mass of
the planet Earth mass and mass of the body, and reciprocal to the square of the
distance between them [2].

m
mE
FG = G
: ð2:1Þ
r 2E

Taking together gravitational constant G, planet Earth mass mE, and radius rE2,
we get another constant g which is called gravitational acceleration.

mE
g=G
: ð2:2Þ
r 2E

Gravitational force (weight) is the product of gravitational acceleration and the

mass of the body we are weighing.
FG = m
g ð2:3Þ

The gravitational force FG is a vector, which means in physics that it has

direction. The direction of the gravitational force is identical to the direction of
gravitational acceleration, g.
→ →
FG =m
g ð2:4Þ

FG gravitational force in N
m mass in kg
g gravitational acceleration in m
s-2
rE radius of Earth in m
G gravitational constant: 6.67259
10-11 N
m2
kg-2
mE mass of Earth in kg

One instrument for measuring the gravitational force FG of a sample is an

electronic balance. Such a balance does not truly “weigh” the mass of a body.
Instead, it measures the attraction force between the body and planet Earth. To get
the mass value on the display of the balance, the measured gravitational force is
divided by the value of g:
2.1 Weighing and Mass 61

Table 2.1 Weighing results of a 1 kg-reference weight if the balance was calibrated for the
gravitational acceleration in Zurich only
Location g (local) in m
s-2 Balance display shows
Zurich 9.80665 1000.0 g
Bogota 9.77390 996.7 g
Reykjavik 9.82265 1001.6 g

FG
mdisplayed = : ð2:5Þ
g
As such, it is necessary that the balance inherently knows the correct value of g.
This is accomplished by adjustment of the balance during calibration by placing a
certified weight on the balance, such as a one-kilogram weight. Then, the balance is
adjusted until the display shows 1 kg. This calibration procedure can be performed
by the operator of the balance or by an official institution. Modern electronic
balances have built-in calibration weights that can be remotely calibrated.
Gravitational acceleration (acceleration of free fall) is slightly different at differ-
ent places on Earth. However, a single international standard is needed for world-
wide use. For this purpose, the value for the gravitational acceleration in Zurich,
Switzerland is used; g = 9.80665 m
s-2.
Reasons for geographical differences in gravitational acceleration stem from lack
of homogeneity in the shape and surface composition of our planet. These include
different mineral composition of the soil in different regions, variation in distance
between the center of the planet and the surface (mountains), and because the planet
is not a perfect sphere (actually, somewhat pear-shaped). Furthermore, because of
the Earth’s rotation, bodies at the equator experience higher centrifugal acceleration
than bodies located far from the equator which will experience a slightly smaller
value of g. Because of this effect, it is necessary to calibrate a balance at the place
where it is to be used. Table 2.1 shows examples of weighing results if that rule is not
followed.

Attention
A balance must be calibrated at the place where it is to be used because
gravitational acceleration is not the same at every place on Earth.

Reference weights are available with different nominal values and in categories
with different levels of accuracy. Reference weights must be made from a material
with density of 8000 kg
m-3 [3]. Companies or commercial laboratories that offer
calibration services must, in turn, have their reference weights tested and certified by
various national metrology institutes that maintain standards for industrial customers
and authorities. In Germany, this is the Physikalisch-Technische-Bundesanstalt
(PTB). In the USA, it is the National Bureau of Standards. In order to continuously
62 2 Mass Density

improve the accuracy in determining the natural constants required for the kg, PTB
works e.g. with monocrystalline spheres made of pure silicon. The spheres are
geometrically so precise that the number of atoms contained can be determined
(Avogradro project [1]).

Bottom Line
The principle of adjustment—calibration—measurement presented here using
the example of weighing also applies to other measuring instruments and
measured variables. In order for sensors and measuring devices (the
so-called measuring equipment) to be able to display correctly, they must be
correctly adjusted or calibrated. This requires reference methods and reference
materials, the so-called standards. The regular inspection of the measuring
equipment available in operation with such standards is a part of routine
quality assurance in production operations and laboratory services.

Air Buoyancy During Weighing

Atmospheric air is a gaseous fluid, and as such, possesses buoyancy. A body placed
on a balance will show a slightly lower weight in normal atmosphere than in a
vacuum. We are familiar with this effect when we sense an apparent weight
reduction from diving into water. The higher the displaced volume and the higher
the density of the displaced fluid, the greater will be the buoyancy force. With air, the
effect is much smaller than with water. Figure 2.1 schematically illustrates how the
indicated weight of a body increases when the air is removed from the weighing
chamber.
So there is a difference between the mass displayed in the atmosphere and the
mass displayed in a vacuum. Since there is no air buoyancy in a vacuum, vacuum
weighing provides the true mass mK of a body. The value displayed under the
influence of air buoyancy is slightly lower, we call this value the apparent mass
mK . Both differ in the factor K the air buoyancy correction factor. With the help of
this factor, we can calculate the true mass of a body weighed in air.

Fig. 2.1 Effect of air

buoyancy: I zero position of
the balance, II balance with
weight, III balance with
weight in vacuum

0.0g

1000.0g
1001.05g

I II III
2.1 Weighing and Mass 63

Table 2.2 Composition of air

Gas Molar mass M in g
mol-1 Concentration in % (V/V)
Nitrogen N2 28.0 78.1
Oxygen O2 32.0 20.9
Argon Ar 40.0 0.9
Carbon dioxide CO2 44.0 0.04
Water vapor H20 18.0 0.01

mK = mK
K: ð2:6Þ

The correction factor K takes into account that the balance has been adjusted with
a reference weight that is equally affected, i.e. that also has a different true mass than
the displayed mass. K is calculated from the density of the reference weight, the
density of the sample, and the air density.
1
ρK - 1
ρR
K = 1 þ ρL
ρL : ð2:7Þ
1- ρK

mK displayed weighing value of the body in kg

mK true mass of the body in kg
ρL density of air in kg
m-3
rK density of the body in kg
m-3
ρR density of the reference weight in kg
m-3
K air buoyancy correction factor

Bottom Line
Air consists largely of nitrogen and oxygen. The exact composition, especially
water content, depends on the location, weather, and environmental activities.
For exact values, see [4]. For estimates, the following values for the composi-
tion of air can be taken (Table 2.2):

The density of the air depends on its composition, as well as on air pressure, air
temperature, humidity, CO2 content, etc. The values can be extracted from tables [5]
or with empirical eqs. A simple calculation formula for air density is [6]:

p=Pa p
ρL =kg
m - 3 = 3:4849
10 - 3
1- 0:3780
W : ð2:8Þ
T=K p

p atmospheric pressure in Pa
pW water vapor partial pressure in Pa
T temperature in K
64 2 Mass Density

Example
Calculation of air density

Atmospheric pressure p = 1000 hPa

Air temperature ϑ = 20 ° C, i.e. T = 293.15 K
Air humidity φ = 50 % r. h.
Water vapor partial pressure pW = φ
ps = 0.5
23.27 hPa = 11.7 hPa

105 11:7
ρL =kg
m - 3 = 3:4849
10 - 3
1- 0:3780
3 = 118
293:15 10
so

ρL = 1:18 kg
m - 3 :

Air buoyancy during weighing depends not only on atmospheric conditions, but
also on the density of the body to be weighed. The density of most foods is in the
range 800–1200 kg
m-3 .Foods with the highest values are crystalline substances
(e.g., NaCl, ρ = 2165 kg
m-3) and those with the lowest values are foams (e.g.,
marshmallows, ρ = 200 kg
m-3). Polymer materials are in the order of magnitude of
1000 kg
m-3, steel and stainless steels between 7500 and 8000 kg
m-3. Table 2.3
shows that the air buoyancy correction has a very small effect on the weighing result.
Deviations are in the order of 0.1% and become smaller as the density of the sample
approaches that of the reference weight. Let us imagine a sale of precious metals
requiring the use of a calibrated scale: In the case of platinum, the scale will read
slightly less than the customer receives. In the case of food foams like ice cream and
marshmallows, the scale will show a little more than the customer receives
(Table 2.3). It does not matter whether or how exactly the scale was calibrated.
The deviation is due to air buoyancy and the fact that a reference weight was used in
the calibration that has a different density than our sample. A deviation of this kind is
called a systematic deviation. For each type of measurement, systematic deviations
and random deviations must be determined and observed when thinking about the

Table 2.3 Examples for weighing without correction for air buoyancy: mK is the mass value
shown on the display of the balance by placing a material sample which has a true mass of
mK = 1 kg
Material ρ/kg
m-3 K mK/g Difference Difference
Marshmallows 350 1.00322 1003.2 +3.2 g +0.32%
Ice cream 500 1.00221 1002.2 +2.2 g +0.22%
Water 1000 1.00103 1001.0 +1.0 g +0.10%
Sucrose 1590 1.00059 1000.6 +0.6 g +0.06%
Stainless steel 7800 1.00000 1000.0 +0.0 g +0.00%
Platinum 21,450 0.99991 999.9 - 0.0 g - 0.01%
2.1 Weighing and Mass 65

reliability of measurements. Table 2.3 shows some examples for the magnitude of
the air buoyancy correction with food materials (calculated with ρL = 1.18 kg
m-3
and ρR = 8000 kg
m-3).

Remark
In some countries, ice cream is sold not by mass units but by volume.

Attention
Air density influences the weighing result because of the air buoyancy. Air
density depends on the pressure, temperature, and humidity of the air, i.e. on
the environmental conditions of the scale and the weather. If air buoyancy is
neglected, these influences are set to zero.

Nevertheless, a fluctuating temperature should be avoided in the laboratory, as

this increases the measurement uncertainty due to random deviations. Fluctuating
humidity in the air of the laboratory has similar effects and can shorten the service
life of the laboratory instruments. Normally, air buoyancy is not taken into account
in everyday production in the food industry because the added precision is unneces-
sary. However, air buoyancy is routinely taken into account when precision
measurements are required, e.g. in metrology institutes or national standards
laboratories.

Bottom Line
Measurements generally have uncertainties. A distinction is made between
random measurement deviations and systematic measurement deviations, both
of which should be kept as small as possible. Deviations due to the incorrect
calibration of a balance or due to the lack of air buoyancy correction are
examples of systematic measurement deviations. For mathematical handling
of random measurement, refer to example in Appendix.

Weighing
The measuring principle of electronic balances is often capacitive or resistive.
Capacitive means that a capacitor in the balance causes electrical capacitance to
change with the position of the weighing plate (ref. Chap. 9). The load of the balance
can thus be determined by measuring the capacity. The electric resistive measuring
principle is based on strain gauges which use the effect that their resistance changes
when an electrical conductor is strained. The load on a strain gauge changes its
electrical resistance. Such strain gauges can be integrated into machines and then
work as online load cells (see also Chap. 16). Mechanical spring scales use the
66 2 Mass Density

increase in length of an elastic steel spring to indicate the load. Here, the linearity
between force and change in length is used (see Chap. 4). In mechanical beam
balances, counterweights are used to compensate for the weight force of the sample.
For the use of scales as online sensors, see also Chap. 16.

1. Relative Density and Absolute Density

The quotient of mass and volume is called mass density or density for short and
has the SI unit kg
m-3.

m
ρ= : ð2:9Þ
V
This definition applies regardless of the physical state of a material, i.e. gaseous,
liquid, and solid bodies. In this way, densities can be specified for materials that are
not homogeneous and, e.g., consist of several phases. Examples of this are
emulsions, suspensions, foams, or powders (see also Chap. 3).
The ratio of the density of a material ρ to the density of a reference material ρR,
e.g. water is called relative density d of a material. In the USA, when the reference
material is water, it is normally called specific gravity. Relative density is a number
without physical units The relative density is a number without physical unit.

ρ
d= : ð2:10Þ
ρR

m mass in kg
V volume in m3
ρ density in kg
m-3
v specific volume in m3
kg-1
d relative density
ρR density of reference material in kg
m-3

Since the density is temperature-dependent, the temperature at which it is read

must be indicated. This can be done by including the temperature as a suffix to the
symbol for density. Then ρ20 then means “density at 20°C.”

Example
Relative density of a chocolate sample
A chocolate sample has a mass density at 20 °C of ρ20= 1.35 g
cm-
3
= 1350 kg
m-3.
Water (reference material) has ρW, 20 = 998.20 kg
m-3,
while at 4 °C, it is ρW, 4 = 999.97 kg
m-3.
The relative density of the chocolate, or its specific gravity, is

(continued)
2.1 Weighing and Mass 67

ρ20 1350 kg
m - 3
d20=20 = = = 1:352
ρW,20 998:20 kg
m - 3

or

ρ20 1350 kg
m - 3
d20=4 = = = 1:350:
ρW,4 999:97 kg
m - 3

For many liquids, the relative density or specific gravity is used as an easy way to
determine quality parameters. The relative density of many aqueous solutions (saline
solutions, sugar solutions, alcoholic solutions) is tabulated for this purpose in many
engineering and scientific handbooks.
When searching values for physical quantities of food, one finds that a clear
tabulation is not so easy due to numerous influencing variables. The density of a food
product or a powder depends not only on the raw material (e.g., plant genus, growth
and harvest conditions), but also on process-related product properties like particle
size, porosity, water content, and storage conditions, such as temperature, water
vapor, and oxygen content of the atmosphere. In addition, it may happen that
determined quantities have to be sorted according to the measurement methods
used in order to establish comparability. For these reasons, a literature search often
does not provide the value sought, but clues for the range of values in which the
product sought should lie. One then calculates further with this assumption of
uncertainty or decides to measure the desired quantity experimentally. Table 2.4
lists some examples of density values along with sources where material data can be
found.
The reciprocal quantity of the (absolute) density r is a quantity with the SI unit
m3
kg-1, i.e. the volume of a certain mass of the material. This quantity is called the
specific volume v.

V 1
v = = : ð2:11Þ
m ρ

Definition
Specific volume: The attribute “specific” usually refers to physical quantities
that are related to the amount of substance, i.e. either to the mass (in kg) or to
the amount of substance (in moles).
volume V in m3

(continued)
68 2 Mass Density

Table 2.4 Examples of density values and reference works for substance data
Material Density in kg m-3 Reference
Whole milk powder 320 Kessler [7]
Whipped cream 300–400 Hui [8]
Coffee, ground 400 Singh [9]
Coffee, instant powder 470 Barbosa-Canovas [10]
Wood, oak 700 Hayes [11]
Pea, dried 800 Singh, Heldman [9]
PP, polypropylene 900 Wächter [12]
Sunflower oil 20 °C 916 Tschubik [13]
Fat 900–950 Lewis [14]
Peas, frozen 970 Gruda [15]
Frozen fish 967 Gruda [15]
Ice, 0 °C 916 Kohlrausch [4]
Water, 4 °C 1000 Grigull [16]
Milk 3.5%, 10 °C 1038 Kessler [7]
Curd cheese 1080 Tschubik [13]
Meat, lean 970-1000 Kessler [7]
Fruit and vegetables 800–1100 Moshenin [17]
PA, polyamide 1130 Wächter [12]
Protein Ca. 1400 Peleg [18]
Spray dried skimmed milk powder 1440–1460 Töpel [19]
Starch 1500 Peleg [18]
Sucrose 1586 Rosenplenter [20]
Fondant, 20 °C 1392 Tschubik [13]
Cellulose 1270–1610 Peleg [18]
NaCl 2163 Kohlrausch [4]

V
Specific volume v = in m3
kg - 1 : ð2:12Þ
m
V
Molar volume v = in m3
mol - 1 : ð2:13Þ
n
Molar-specific quantities are practical and most commonly used in chemis-
try and physical chemistry. However, in the engineering sciences, mass-
specific quantities are preferred, i.e. with “quantities related to the kilogram.”
It is customary to use the lowercase letter of the quantity for specific sizes. We
will see this convention again frequently in the coming chapters, examples are:
Heat capacity C in J
K-1 specific heat capacity c in J
K-1
kg-1
Enthalpy H in J specific enthalpy h in J
kg-1
Caloric value E in kcal caloric value e in kcal
kg-1
2.1 Weighing and Mass 69

Bottom Line
A quantity related to the kilogram is called a specific quantity, more precisely:
mass-specific quantity. Specific quantities are symbolized with lowercase
letters.

Temperature and Pressure Dependence

The primary reason why density depends on temperature is that most materials
experience thermal expansion as temperatures increase, and thermal contraction
when temperatures decrease. Specifically, it is the volume that expands with increas-
ing temperature while mass remains constant, i.e. density decreases with increasing
temperature. This effect is most evident in gases and is much smaller in liquids and
solids.
In order to calculate an approximate value for the density of a food material, the
specific volumes are added according to their mass fraction x:

1 1
= xi
vi = xi : ð2:14Þ
ρ i i
ρi

For a first estimate, it is sufficient to take into account only the major components
of a food, which are usually water, proteins, carbohydrates, and fats. These four
components make up what is normally referred to as the gross composition of a food.
Table 2.5 lists equations for the density of these substances at any specified
temperature.

Bottom Line
Main components of food are: water, proteins, carbohydrates, and fats.

Table 2.5 Temperature dependence of density according to [21, 22] in the range -40 °C to 150 °C
ρ in kg
m‐3, ϑ in ° C
Water ρ = 997.18 + 0.0031439
ϑ - 0.0037574
ϑ2
Fat ρ = 925.59 – 0.41757
ϑ
Carbohydrates ρ = 1599.1 – 0.31046
ϑ
Proteins ρ = 1329.9 – 0.5184
ϑ
Dietary fiber ρ = 1311.5 – 0.36589
ϑ
Ash ρ = 2423.8 – 0.36589
ϑ
70 2 Mass Density

Example
Density of chicken egg yolk at 0 °C, knowing the fraction x of each
component [23].

x= m
mtotal Density at 0 °C xi ρ1
i

Water 0.5 ρW = 997.18 kg
m-3 0:5

997:18 kg
m - 3
5:0
10 - 4
kg
m - 3
Proteins 0.16 ρP = 1329.9 kg
m-3 0:16
1329:kg
m - 3
1:2
10 - 4
kg
m - 3
Fat 0.32 ρFat = 925.59 kg
m-3 0:32
925:59 kg
m - 3
3:5
10 - 4
kg
m - 3
Carbohydrates 0.003 ρKH = 1599.1 kg
m-3 0:003
1599:1 kg
m - 3
1:9
10 - 6
kg
m - 3
Minerals 0.0017 ρAsche = 2423.8 kg
m-3 0:0017
2423:8 kg
m - 3
7:0
10 - 7
kg
m - 3
xi ρ1 = 1028 kg
1

m - 3
i
i

1 1 10 - 4 9:7
10 - 4
= xi = ð5:0 þ 1:2 þ 3:5 þ 0:019 þ 0:007Þ =
ρ i
ρi kg
m - 3 kg
m - 3
1
= :
1028 kg
m - 3

ρ = 1028 kg
m - 3 :

More accurate values of the density of airless water at a pressure of 1013.25 hPa
can be used with the help of the polynomial equations of Kell [24] as a function of
temperature (ϑ in °C):

5
ρW = an
ϑn =ð1 þ b
ϑÞ:
n=0

The coefficients updated in 1990 [25] for the Kell formula are (Table 2.6):

a0 = 9:9983952
102 kg
m - 3
a1 = 1:6952577
101 ° C - 1
kg
m - 3
a2 = - 7:9905127
10 - 3 ° C - 2
kg
m - 3
a3 = - 4:6241757
10 - 5 ° C - 3
kg
m - 3
a4 = 1:0584601
10 - 7 ° C - 4
kg
m - 3
a5 = - 2:8103006
10 - 10 ° C - 5
kg
m - 3
b = 1:6887236
10 - 2 ° C - 1
2.1 Weighing and Mass 71

Table 2.6 Examples of density calculation formulas of some foods. Further formulas are given in
the Annex
Product ρ in kg
m-3, ϑ in ° C, dry matter content: dm in %(m/m)
Whey ρ = ρwater + 4.039
dm + 1.273
10-2
dm2 + 9.62
10-5
dm3 [7]
UHT milk ρ = 1040.7 - 0.2665
ϑ - 2.3
10-3ϑ2 [7]

Fig. 2.2 Anomaly of water

(maximum density at 4 °C),
normal density curve (N),
schematic

Water shows abnormal behavior as it undergoes thermal expansion. Figure 2.2

shows this behavior schematically. Normally, a substance in the solid state has a
higher density than in the liquid state. In case of water the anomaly continues here:
the solid phase (ice) has a significantly lower density than the liquid phase.
If the density of a liquid or solid material is not known at a certain temperature, a
calculation can be made with the thermal volume expansion coefficient γ. The
volume of the sample is calculated at the desired temperature and from this the
desired density.
1 dV
γ=
ð2:15Þ
V dT
and
V = V 0 þ γ
dT ð2:16Þ

γ thermal volume expansion coefficient in K-1

ρ density in kg
m-3
T temperature in K
V volume in m3

Pressure Dependence of Density

If pressure on a body at a constant temperature increases, it will be counteracted by a
decrease in its volume. Ideal compressibility is characterized by a linear volume-
pressure relationship. When the external pressure is removed, the body regains its
72 2 Mass Density

original volume exhibiting elastic behavior. The extent of this reversible volume
change is determined by the compressibility coefficient κ of the material, often called
simply compressibility.
1 dV
κ= - ð2:17Þ
V dp
i.e.,

dV
= - κ
dp ð2:18Þ
V
respectively

ΔV
= - κ
Δp: ð2:19Þ
V
The relative change in density at variable pressure is thus linearly dependent on
the compressibility of a material.
Δρ ΔV
= = jκ
Δpj: ð2:20Þ
ρ V
The reciprocal quantity of compressibility is called compression modulus K and
has the SI unit Pa. The compression modulus characterizes the volume elasticity of a
material.

1
K= : ð2:21Þ
κ
p pressure in Pa
ρ density in kg
m-3
m mass in kg
V volume in m3
κ compressibility in Pa-1
K compression modulus in Pa

The compressibility of real liquids and solids depends not only on the temperature
but also slightly on the pressure. Water (20 °C) at atmospheric pressure has a
compressibility κ 20 = 4.59
10-10 Pa-1, at a pressure of 1000 bar it is
κ20 = 3.88
10-10 Pa-1. Tabulated substance data of water and ice at elevated
pressures can be found in [5, 4, 16, 26].
2.1 Weighing and Mass 73

Example
An aqueous food is subjected to a high-pressure treatment of 10 min at
5400 bar. The change in volume and density of the food should be estimated.
We use the substance data of water.

Δp ≈ 5399 bar = 539:9 MPa:

κ20 = 4:59
10 - 10 Pa - 1 :
Δρ ΔV
= = jκ
Δpj:
ρ V
Δρ ΔV
= = 4:59
10 - 10 Pa - 1
539:9
106 Pa = 0:25:
ρ V
Result: The density of the product increases by 25% during treatment. The
volume decreases by 25%.

Example
Estimate the specific energy input when water is compressed to 3000 bar.
Volume work is

dW = - p
dV
W= - p
dV
mit
dV = - κ
V 0
dp
W= - p
ð - κ
V 0 Þ
dp

W =κ p
V 0
dp

W = κ
V0 p
dp
1
W= κ
V 0
p2 :
2
To obtain the specific volume work, we divide by mass m:

(continued)
74 2 Mass Density

W 1 V κ
p2
= κ
0
p2 = :
m 2 m 2ρ0

For ρ0, the original density of the sample we take 1000 kg
m-3 (value of
water).
For p we approximate ≈3000 bar = 300 MPa
2
W 4:59
10 - 10 Pa - 1
300
106 Pa
=
m 2
1000 kg
m‐3
4:59
10 - 10
9
1016 N
m - 2
= ≈ 20
103 N
m
kg - 1 :
2
103 kg
m - 3
W
≈ 20 kJ
kg - 1 :
m

Example
What temperature increase can be expected by high-pressure treatment of an
aqueous solution if the specific energy input is about 20 kJ
kg-1?
solution:
The sensible heat (see Chap. 8) is Q = m
cp
ΔT.
With an assumed specific heat capacity of cp ≈ 4 kJ
kg-1
K-1 we get

Q = m
cp
ΔT
Q
= cp
ΔT
m
1Q
ΔT =
cp m
1 20 kJ
ΔT =

4 kJ
kg - 1
K - 1 kg
ΔT = 5 K

Gases
Some foods contain gases and therefore have a very large thermal expansion and a
strong temperature dependence on density (Table 2.7).

In order to calculate the density of a gas depending on temperature and pressure,

we first want to consider the gas as an ideal gas.

p
V = n
R
T: ð2:22Þ
2.1 Weighing and Mass 75

Table 2.7 Examples of gases in food production

Gas Molecule Molar mass M in g × mol-1 Application example in
Oxygen O2 32.0 Packaging
Nitrogen N2 28.0 Packaging
Carbon dioxide CO2 44.0 Packaging, beverages
Nitrous oxide N20 44.0 Whipped cream

This equation is known as the ideal gas law and applies to gases in which no
interaction forces occur between the molecular particles. At low particle
concentrations in a gas at moderately low temperatures, i.e. when we can neglect
the intermolecular interactions in a gas, we can apply the theory of the ideal gas law
to many gases. To calculate the density of an ideal gas, we first replace the universal
gas constant R with the specific gas constant Rs.

R
RS = ð2:23Þ
M
so

n
M R
p
V=
R
T = m

T = m
RS
T: ð2:24Þ
M M
By this, we get the ideal gas law in a form where the mass m is used instead of
moles n. And the specific gas constant RS is used instead of the universal gas constant
R:
p
V = m
RS
T: ð2:25Þ

this is

m p
= ð2:26Þ
V RS
T
or

p
ρ= : ð2:27Þ
RS
T
This means that the density of an ideal gas at a constant temperature is directly
proportional to the pressure of the gas: ρ~p.
76 2 Mass Density

p pressure in Pa
Rs specific gas constant in J
K-1
kg-1
ρ density in kg
m-3
T temperature in K
m mass in kg
V volume in m3
M molar mass in g
mol-1

Example
Can we handle dry air as an ideal gas?
With an average molecular weight of air (with rounded values from [27]) of
M air = 0:78
28:0 g
mol - 1 þ 0:21
32:0 g
mol - 1 þ 0:01
40:0 g
mol - 1
= 28:96 g
mol - 1

.
it is

R 8:314 J
K - 1
mol - 1
RS = = = 287:1 J
K - 1
kg - 1
M 28:96 g
mol - 1

at

p = 1 atm = 1:013 bar = 1013 mbar = 1013 hPa = 1:013
105 Pa

and a temperature of 0 °C = 273.15 K.

air density results to

p
ρ= :
RS
T

1:013
105 Pa
ρ= = 1:2 kg
m - 3 :
287:1 J
K - 1
kg - 1
293:15 K

This value matches the literature value of ρL = 1.20 kg
m-3 [4], which confirms
the assumption that atmospheric air can be treated here approximately like an ideal
gas. In this calculation example, dry air was assumed. At higher humidities we can
quickly reach the limits of this approximation.
2.2 Laboratory Methods for Determining Density 77

2.2 Laboratory Methods for Determining Density

There are many different methods for determining the density of beverages, viscous
fluids, solids, or powders. In addition, many industries have developed their own
methods and, in some cases, use traditional density scales. We start with an overview
and then discuss some methods in detail.

Pycnometric Determination
By weighing a known volume of a liquid, the density of that liquid can be measured
in a simple way. Glass bulbs with precisely known volume that are used for this
purpose are called pycnometers. A pycnometer can also be any other instruments
designed for the same purpose. It must have sample chambers of precisely known
volume, but made of other materials (not glass bulbs). The glass bulb or sample
chamber will have a marker to which the liquid sample must be carefully filled. Then
the density of the fluid can be calculated by:

mF - m0
ρF = : ð2:28Þ
V
Because of thermal expansion of the glass, the pycnometer volume is known for
the temperature at which it was calibrated, only. So, for measurement of the absolute
density, pycnometers should be used at the same temperature at which they were
calibrated. Another way is to measure the relative density (specific gravity) rather
than the absolute density. For this purpose, the pycnometer is weighed with the
sample liquid and again weighed with the reference liquid (often water). The ratio of
both weights gives the relative density d, or specific gravity of the sample.
mF m
V ρ
= F = F = d: ð2:29Þ
mW V
mW ρW

m0 mass of empty pycnometer in kg

mF mass of pycnometer filled with sample in kg
mW mass of pycnometer filled with water in kg
V volume in cm3
ρF density of sample in kg
m-3
ρW density of water in kg
m-3
d relative density of sample (specific gravity)

Once the relative density d, or specific gravity, of the sample is known, and the
density of the reference material is known from literature, the absolute density of the
sample can be calculated as:

ρF = d
ρW : ð2:30Þ

As mentioned before, a weighing result can be corrected for the effect of

atmospheric buoyancy. The corrected mass is the true mass of a body and is slightly
78 2 Mass Density

Fig. 2.3 Pycnometer, examples

Fig. 2.4 Pycnometer designs: (a) Reischauer, (b) Bingham, (c) Gay-Lussac, (d) Sprengel, (e)
Lipkin, and (f) Hubbard

higher than the mass which the balance displays. When a true mass is used to
calculate density this is called a true density. If the atmospheric buoyancy correction
is not applied, the result can be called an apparent mass, and respectively, an
apparent density. When relative density is measured by using the same pycnometer
for both sample and reference material, the buoyancy effect is eliminated by forming
the weight ratio. Figures 2.3 and 2.4 show different designs of glass pycnometers.
2.2 Laboratory Methods for Determining Density 79

Hydrostatic Balances (Buoyancy Weighing)

The principle of a hydrostatic balance is based on Archimedes’ law of buoyancy. If a
body is submersed in a fluid, its weight will be lowered because of the buoyancy
force. The buoyancy force is the weight force of the volume of water displaced by
the submersed body and is therefore directly proportional to the volume of the
submersed body and the density of the fluid. By measurement of the buoyancy
force with the balance, the volume of the body can be determined quite accurately,
and together with the measured mass of the body, the density is obtained. A simple
technique for making this type of measurement is to place a beaker partially filled
with water on top of a top-loading balance, with the weight of the beaker and water
tared-out to read zero on the display. Then, fully submerge the solid body beneath
the water surface, taking care that it neither touches the bottom nor the sides of the
beaker. The weight reading shown on the display of the balance will be the weight of
the volume of water displaced by the solid body. Since density of water is known, the
precise volume of the solid body is determined.
mL
ρK = : ð2:31Þ
VK
F A = ρF
V K
g: ð2:32Þ

with Δm as the difference in weight (in kg) of the body before and after
submersion:
Δm = mL - mF : ð2:33Þ

F A = Δm
g: ð2:34Þ

ðmL - mF Þg mL - mF
VK = = ð2:35Þ
ρF
g ρF

mL
ρK =
ρ : ð2:36Þ
mL - mF F
So the density of a body can be obtained by taking first the weight prior to
submersion (that means weighing in air) and its weight when submersed in a fluid
with a known density ρF using Eq. (2.34).

mL mass of the body in air in kg

mF mass of the body submersed in fluid in kg
FA buoyancy force in N
g gravitational acceleration in m∙s-2
VK volume of body in m3
ρF density of fluid in kg
m-3
ρK density of body in kg
m-3
d relative density
80 2 Mass Density

Fig. 2.5 Hydrostatic balance

design. 1 balance, 2 platform,
3 small beaker, 4 large beaker,
5 support bracket, 6 pan, and
7 thermometer

With the ratio

ρK mL
= ð2:37Þ
ρF mL - mF

the relative density d (specific gravity) of the body is

ρK
d= ð2:38Þ
ρF
mL
d= ð2:39Þ
mL - mF
or

mL - mF
ρF = ð2:40Þ
VK
That means d can be calculated very quickly after two readings from the balance
using an immersion body with a known volume VK [28].
Figure 2.5 shows a special design of hydrostatic balance that is suitable for
measuring the density of a solid or the density of the liquid in the reservoir when
used with a solid body of precisely known volume. To obtain the density of a solid,
the sample is weighed first in air, and then it is submersed and the weight is taken
again. As can be seen in Fig. 2.5, there is a small pan mounted on the weighing plate
of the balance. A small beaker on a cable is suspended within a larger beaker
containing the fluid of interest. The large beaker is resting on a raised platform so
its weight is not transmitted to the balance.
To obtain the fluid density in the large beaker, a test body with known volume is
first placed on the pan and its weight in air is measured. Then the test body is placed
into the small beaker, submersed in the fluid, and weighed once again. Then the
density of the liquid is calculated using Eq. (2.38). It should be remembered that the
2.2 Laboratory Methods for Determining Density 81

density of the fluid is dependent on temperature so the temperature must be con-

trolled and recorded.

Example
Starch in corn kernels.
Agronomists have determined that degree of ripeness of an ear of corn can
be closely correlated with relative density (specific gravity) of individual corn
kernels. A relative density between 1.080 and 1.118 is the range expected at
peak ripeness. Values below that range would indicate insufficient ripeness,
while values above would indicate the corn has gone beyond peak ripeness
[29]. With a hydrostatic balance a fast check of incoming raw material is
possible.

Example
Rapid determination of the starch content of potatoes.
The density of potatoes also is a function of the starch content. By weighing
a basket with potatoes in air, and then submersed in water, the starch content
can be estimated quickly.
The diagram in Fig. 2.6 shows the relationship between the relative density
(specific gravity) of potatoes and their starch content. For convenience, the
scale on the left side vertical axis of the diagram shows the underwater weight
mUW of a potato sample weighing 5050 g in air. A standard test with that
sample size is published by the EU [30]. The fact that the weight of the basket
in air and under water is not the same is neglected.
The underwater weight force of the sample is the weight force in air GL
lowered by the buoyancy force FA

GUW = GL - F A ð2:41Þ

mUW
g = mL
g - ρF
V K
g ð2:42Þ

so the volume of the sample is

mL - mUWG
VK = ð2:43Þ
ρF

and the density of the sample

mL ρF
m L
ρK = = ð2:44Þ
VK ðmL - mUW Þ

the relative density then is

(continued)
82 2 Mass Density

ρK mL
d= = ð2:45Þ
ρF mL - mUW

g gravitational acceleration in m2.s-1

GUW underwater weight force in N
GL weight force in air in N
FA buoyancy force in N
mUW apparent mass of submersed sample in kg
mL mass of sample in air in kg
ρF density of fluid in kg.m-3
ρK density of sample in kg.m-3
Vk volume of sample in m3
d relative density (specific gravity) of sample

Mohr–Westphal Balance
The Mohr–Westphal balance (Fig. 2.7) is another type of hydrostatic balance. It is
designed as a non-symmetric beam balance for measuring the density of liquids. At
the free end of the arm of the balance a “buoyancy body” is suspended in air. The
buoyancy body is normally made of glass and can have a built-in thermometer. Then
the buoyancy body is submersed into the liquid of interest. Because of the effect of
buoyancy, the weight of the submersed glass body will appear lower than it was in
air and will bring the balance out of zero. The buoyancy force can be measured by
successively adding small weights to the arm until the balance is restored to zero.
The measurement is then repeated with water as a reference liquid. The ratio of both
readings provides the relative density or specific gravity of the liquid as can be
shown below. With the buoyancy force

Fig. 2.6 Relative density d of

potatoes versus starch content.
The scale on the left shows the
underwater weight of a 5050 g
sample
2.2 Laboratory Methods for Determining Density 83

Fig. 2.7 Mohr–Westphal

balance. 1 beam, 2 weights,
3 buoyancy body, and 4 liquid
sample

FA = ρ
V
g ð2:46Þ

FA = m
g ð2:47Þ

with the fluid of interest

F A,F = ρF
V K
g = mF
g ð2:48Þ

with water as the reference material

F A,W = ρW
V K
g = mW
g ð2:49Þ

so the ratio is

mF
τ= ð2:50Þ
mW
Because the volume of the glass body is the same for both readings, it is

mF m
V ρ
τ= = F = F =d ð2:51Þ
mW V
mW ρW

so the specific gravity simply is

mF
d= ð2:52Þ
mW
m mass of weights in kg
V volume in m3
FA buoyancy force in N
ρ density in kg
m-3
g gravitational acceleration in m∙s-2
τ submersed weights ratio
d relative density
subscript F for fluid
subscript W for reference material (mostly water)
84 2 Mass Density

The result should be recorded along with the temperature of the measurement.
Often both readings are taken at 20 °C and the result is written as d20/20. The quantity
d20/4 would mean that the density of the liquid was compared to the density of water
at 4 °C:

ρF,20 ° C
d20=20 =
ρW,20 ° C
ρ
d20=4 = F,20 ° C :
ρW,4 ° C

Hydrometer
Hydrometers (sometimes called aerometers) are hollow glass bodies with the shape
of a buoy (Fig. 2.8). Hydrometers are designed with a volume to mass ratio in such a
way that the glass body will float at a certain depth in the liquid under investigation.
Depending upon the density of that liquid the hydrometer will float at a higher or
lower position. The upper part of the hydrometer has a scale for reading the
non-submersed part h of the floating glass body (Fig. 2.9).
The floating depth position of the hydrometer depends on weight force and
interfacial force (force due to surface tension, see Chap. 6). It is
FG þ Fσ = FA ð2:53Þ

that means:

Fig. 2.8 Hydrometer.

1 scale, 2 body (with and
without thermometer), and
3 keel [31]

°C
25

2 20

15

10

3
2.2 Laboratory Methods for Determining Density 85

Fig. 2.9 Reading of a

hydrometer scale at the liquid
surface (example)

m
g þ σ
π
d = VS
ρ
g ð2:54Þ

so

π
d2
m
gþσ
π
d= V þ
h
ρ
g ð2:55Þ
4

the non-submersed part of the hydrometer has the length

4 m
gþσ
π
d
h= V- ð2:56Þ
π
d 2 ρ
g

that means h = function(ρ).

Without consideration of interfacial tension effects, the hydrometer equation
simplifies to

4 m
h= V- : ð2:57Þ
π
d2 ρ

The non-submersed length of the hydrometer can be read with the aid of a scale
on the upper part of the hydrometer. A weight at the bottom of the hydrometer acts
like the keel of a sailboat to ensure that it will float in the liquid in a vertical
orientation.
The scale can be calibrated directly in units of density or, e.g., in concentration
units (see Fig. 2.9). Hydrometers with special scales are available for specific
applications, such as for sugar solutions (saccharimeter), alcohols (alcoholometer),
acids (acid hydrometer), Baumé hydrometer for salt solutions, milk (Quevenne
lactometer), beer wort hydrometer, urine checker, liquid gas aerometer, etc.
Sometimes the combination of two physical properties will give the information
needed about a process or a product. For example, by knowing both the density and
refractive index of beer wort, the alcohol content can be calculated, and by this, the
progress of fermentation [29].

m mass of hydrometer in kg
V total volume of hydrometer in m3
VS volume of hydrometer submersed in m3
(continued)
86 2 Mass Density

FG weight force of hydrometer in N

FA buoyancy force in N
Fσ interfacial tension force (force of surface tension) in N
σ interfacial tension (surface tension) in N∙m-1
d diameter of hydrometer neck in m
ρ density of liquid in kg
m-3
g gravitational acceleration in m∙s-2
h length of hydrometer neck not submersed in m

There are hydrometers available which are corrected for interfacial tension
offered in different range categories called L (low, 15–35 mN
m-1), M (medium,
35–65 mN
m-1), and H (high, for higher values). For high accuracy requirements, a
correction factor for the exact interfacial tension of the fluid can be used [32].
Winemakers and fruit growers determine the density of juice or must (“must
weight”) in order to estimate the sugar content of the fruit, and to estimate from it the
alcohol content achievable by fermentation. The conversion from relative density
into Oechsle units (°Oe) is: densitye in ° Oe = (d - 1)
1000.

Example
A 10% (m/m) sucrose solution (that is a sugar concentration of 10° Bx) has a
relative density of d = 1.04. Conversion to Oechsle: (1.04 -
1)
1000 = 40 ° Oe.

Submersion Technique
Pycnometers and hydrometers do not work very well with liquids of high viscosity.
For highly viscous liquids, measurement of density can be performed with the
submersion technique (refer Table 2.8). A beaker with the viscous liquid sample is
put on a balance. The display value is recorded, or the display may be set to zero
(tare). Then a test body with known volume is pressed into the sample (Fig. 2.10).
The buoyancy force caused by the submerged test body is transferred to the balance
and appears on the display as an apparent increased weight Δm. This increased
weight force is the buoyancy force, and is the weight of the displaced liquid, which is
equal in volume to the volume of the submersed solid body VK:

ΔG = F A ð2:58Þ

and

Δm
g = ρ
V K
g ð2:59Þ

so
2.2 Laboratory Methods for Determining Density 87

Table 2.8 Laboratory methods for measurement of density

Pycnometer Hydrostatic Immersion Aerometer Floating

Gases, liquids, Liquids, solids Liquids Liquids Solids

solids
Mohr–Westphal X-ray Bulk density Tapped Resonator
density

Liquids Solids Powders Powders Gases, liquids,

solids

Fig. 2.10 Submersion

technique for density
measurement. 1 depth mark,
2 liquid sample, and
3 buoyancy body of known
volume

Δm
ρF = ð2:60Þ
VK
ΔG apparent weight force increase in kg
Δm apparent mass increase in kg
VK volume of test body in m3
ρ density of liquid kg
m-3
88 2 Mass Density

Particle Flotation Technique

A solid particle will remain suspended within a liquid (meaning it will neither rise to
float at the surface nor sink to rest at the bottom) when both the liquid and solid
particle have the same density. Then the buoyancy force acting upward on the
particle will be equal to the gravitational force acting downward, and the particle
will remain suspended. To measure the density by the floating method, two different
liquids must be chosen so that one has a density lower than the particle, and the other
higher. Also, both liquids must be miscible with each other when mixed together and
remain optically transparent in order to observe the particle when placed into the
liquids. The liquids must also have no adverse effect on the particle, meaning the
particle will not dissolve or be altered in any way. Appropriate liquids for this
purpose are listed in Table 2.9.
The solid particle of unknown density is first placed into a beaker partially filled
with the low-density liquid. Since the liquid density is lower than that of the particle,
the particle should sink to rest at the bottom of the beaker. Then, small amounts of
the high-density liquid are lowly added and gently mixed until the particle just
begins to float upward. When this point is reached the density of the liquid mixture
and the density of the solid sample are the same (Fig. 2.11). By measuring the
density of the liquid mixture, e.g., by a pycnometric method, the density of the solid
particle sample is known. For measuring the density of agricultural materials or food
products, the floating method liquids will need to have densities ranging between
900 and 1500 kg
m-3. This range of densities can be obtained, e.g., with mixtures of
water and alcohols or with sugar solutions or salt solutions.
The temperature of the liquid mixture must be held constant for this procedure to
yield accurate results. If this is ensured, there is no need to perform the floating
technique at a fixed or controlled temperature so it can be carried out very easily. The
floating technique is recommended for solid samples which cannot be handled in a
hydrostatic balance. Problems may arise because the sample may dissolve or alter in

Table 2.9 Liquids for the floating technique (examples)

Liquid Density at 20 °C in kg
m-3 [33]
Water 998.2
Aqueous sucrose solution 50% (m/m) 1229.5
Aqueous NaCl solution 20% (m/m) 1147.8
Ethanol 90% (m/m) 818.0

Fig. 2.11 Floating

technique. The sample P is
suspended in the fluid F
2.2 Laboratory Methods for Determining Density 89

aqueous solutions. In that case, the samples should be packaged or coated or another
liquid should be chosen.

Density Gradient Column

An interesting variation of the floating technique is the use of a density gradient
column as a fast and accurate method for determining the density of small solid
particles like agricultural seeds and grains [34]. The method is based on observing
the depth to which a particle will remain suspended in a tall column of liquid that
exhibits a gradual change in density along the depth of the column, as shown in
Fig. 2.12. The column is first prepared by filling it carefully with layers of mixtures
in different proportions of two miscible transparent liquids of different densities
similar to those used in the floating methods. The first layer at the bottom would
consist only of the high-density liquid. The next layer above would be a mixture of
10% low-density liquid and 90% high-density liquid, followed by a layer of 20%
low-density liquid with 80% high-density liquid, and so on, until the layer at the top
would consist only of the low-density liquid. In this way, the column will exhibit a
gradual gradient in density ranging from the high density at the bottom up to the low
density at the top. The column is normally calibrated with color-coded glass beads of
known density commercially available for this purpose. The calibration beads are
dropped into the column, and each bead comes to rest suspended at the depth to
which the liquid density is the same as the bead. A graph of bead density versus
column depth serves to calibrate the column. Then, small particles of unknown
density are dropped into the column. They will come to rest suspended at the depth
to which the liquid density is the same as the particle. This depth is read from the

Fig. 2.12 Density gradient

column [35]: 1 clock motor,
2 water jacket, 3 elevating
basket, and 4 stand
90 2 Mass Density

Table 2.10 Liquids suit- Liquid Specific gravity

able for density gradient
Isopropanol—Water 0.79–1.00
column [17]
Isopropanol—Diethylene glycol 0.79–1.11
Water—Calcium nitrate 1.00–1.60

graduated scale on the column. The particle density is obtained from the calibration
graph, which will give the density at that depth. Table 2.10 gives a listing of various
liquid mixtures suitable for preparation of a density gradient column.
For the time-consuming preparation of density gradients, there are technical
solutions with automated dosing pumps. The density gradient method can also be
used as a centrifuge separation process. Then there is the density gradient column in
a centrifuge vessel. Under the influence of centrifugal force, the components of the
sample material take their respective floating position. Subsequently, the individual
density fractions can be removed separately from the centrifuge tube. With the help
of ultracentrifuges, proteins and microorganisms can be separated by their
density [36].

Resonator Frequency Techniques

When we have an oscillating tube filled with the fluid of interest, its resonant
frequency depends on the density of the liquid. So with electronic means and
appropriate calibration, it is possible to measure the density of a fluid by monitoring
the resonant frequency of an oscillating (vibrating) tube. Density meters based on
this effect are available as stationary laboratory devices, portable handheld sensors
and as in-line process sensors. The working principle is described in detail in
Chap. 16.

Overrun
Overrun is a term used in the manufacture of products involving whipping or
aeration, such as in ice cream or whipped dairy toppings. The whipping process
causes air to become incorporated into the liquid being mixed causing the physical
phenomenon of foaming. Here the density is an important physical property which is
directly related to the degree foaming. Foaming causes the volume of the foamed
liquid to increase far beyond the original volume of the liquid prior to whipping. This
volume increase, when divided by the original volume, is called overrun A. It is a
dimensionless ratio, and can be mathematically expressed as follows:

V foam - V liquid
A= ð2:61Þ
V liquid

since:

mfoam = mliquid þ mgas ≈ mliquid ð2:62Þ

and with
2.2 Laboratory Methods for Determining Density 91

m = mfoam ≈ mliquid ð2:63Þ

with

1 V
=
ρ m
overrun can be expressed:

V foam - V liquid
m m
A= m : ð2:64Þ
V liquid

i.e.,
1
ρfoam - 1
ρliquid
A= 1
: ð2:65Þ
ρliquid

that means

ρliquid ρliquid
A= - : ð2:66Þ
ρfoam ρliquid

so

ρliquid
A= - 1: ð2:67Þ
ρfoam

A overrun
m mass in kg
V volume in m3
ρ density in kg∙m-3

Example
Overrun of ice cream.
Ice cream is produced by whipping and freezing of a liquid mixture (see,
e.g., [37]). Typical values are given to ρice cream = 550 kg
m-3. and
ρice mix = 1150 kg
m-3. The overrun is calculated using eq. (2.67) to

(continued)
92 2 Mass Density

ρice mix
A= - 1
100%
ρice cream

1150 kg
m - 3
A= - 1
100% = 109%:
550 kg
m - 3

This value for the overrun means the volume of the liquid ice cream mix
was raised by 109%, or more than doubled.

Density of Solids
Many agricultural materials and food and feed ingredients are in the form of granular
materials (grains, meals, and powders), which are bulk solids made up of small
particles. The weight or size of the individual particles within any of these types of
materials may vary over a large range, e.g. from frozen diced vegetables to corn
kernels to fine powder particles. The term “solid density” means the density of the
solid material of which a particle is made, no matter what type of fluid or other
material may exist between the particles. On the other hand, if the solid particles
contain pores or hollow cavities filled with gases or liquids, this contributes to the
density of the solid. When pores or cavities occur, it is important to state whether
they are closed or open. If they are closed, meaning they are located completely
within the solid particle, they belong to the solid. If they are open to the surroundings
at the particle surface, e.g., the atmosphere, they do not belong to the solid body. To
avoid communication errors the density of solids should be given with a note like
“including pore volume” or “without pore volume.” This can be accomplished by
how the system boundaries are defined and can be calculated as follows:

mS
ρ= : ð2:68Þ
VS
mS mass of solid material in kg
VS volume of solid material in m3
ρ density of solid in kg
m-3

To measure the density of a solid particle often simply means to measure its
volume, because its mass is known upon weighing. To get the volume of the solid
sample without its open pores, a pycnometer technique with an appropriate liquid
can be used. The liquid must not alter or dissolve the sample. For this purpose, a
sequence of weighings is conducted as indicated in Fig. 2.13. After weighing the
empty pycnometer m0 the pycnometer is weighed with the sample mP. Then the
pycnometer is filled up to a designated mark with an appropriate reference liquid of
known density and weighed again mP,F. Finally, the pycnometer is weighed when
filled to the same mark with only the reference liquid mF [38]:
2.2 Laboratory Methods for Determining Density 93

Fig. 2.13 Pycnometer measurement of the density of a solid granular material, e.g. a powder

mS = mP - m0 : ð2:69Þ

VS = V0 - VF: ð2:70Þ

mP,F - mP
VF = : ð2:71Þ
ρF

mP - m0
ρS = : ð2:72Þ
V 0 - mP,Fρ- mP
F

mP - m0
ρS = : ð2:73Þ
mF
ρF - mP,Fρ- mP
F

mP - m0
ρS = ρF : ð2:74Þ
mF - mP,F þ mP
ρS
d= : ð2:75Þ
ρF
mP - m0
d= : ð2:76Þ
mF - mP,F þ mP

m0 mass of empty pycnometer in kg

mS mass of sample in kg
mP mass of pycnometer with sample in kg
mP,F mass of pycnometer with sample and filled with reference liquid in kg
mF mass of pycnometer with reference liquid only in kg
V0 volume of pycnometer in m3
VS volume of sample in m3
VF volume of reference liquid filled in pycnometer in m3
ρS density of solid in kg
m-3
ρS density of reference liquid in kg
m-3

For precision measurements the true mass instead of apparent mass (see Sect. 2.1)
should be used. In that case, the true density of the reference liquid should also
be used.
94 2 Mass Density

Bulk Density
Powders and bulk goods contain hollow spaces or voids filled with gas, normally air.
The density of that type of bulk material, including the void spaces, is called bulk
density. Using Eq. (2.7), the bulk density can be calculated by weighing a sample of
the bulk material and measurement of its volume. The volume of the whole bulk
material must be taken “as is.” To measure this volume, the sample material can be
poured into a beaker or cylinder up to a known volumetric mark. Different
techniques in filling the beaker or cylinder may lead to different distributions of
solid particles and hollow spaces. So, to get repeatable results the technique of filling
has to be standardized.
To overcome problems with repeatable filling technique, the bulk material can be
tapped before reading of the volume. By tapping the material, the solid particles will
“settle” into the most stable situation they can reach. The void spaces will get smaller
as the solid particles settle step by step into a spatial situation where the bulk density
will reach a maximum. The time needed to reach this maximum depends on tapping
speed and tapping amplitude. There are laboratory devices available which perform a
chosen number of tappings in a continuous motion. Depending on the sample
properties, sometimes 10, 100, 1000, or more tappings may be appropriate to
reach maximum bulk density.
Figure 2.14 shows an example of a device which can be used to measure bulk
density and tapped bulk density subsequently. First the bulk material is filled into a
1000 cm3 cylinder until it is overflowing under repeatable technique conditions.
Then with aid of a flat spatula, the excess overflow of sample material is scraped
away from the top of the cylinder to leave the sample perfectly level at the top, and
the 1000 cm3 sample is weighed to get the bulk density.

Fig. 2.14 Device for tapping

of bulk goods, 1 rotating cam,
2 housing, 3 powder sample,
4 cylinder, and 5 overring
2.2 Laboratory Methods for Determining Density 95

Table 2.11 Characterizati- Hausner ratio Powder flowability

on of powder flowability by
1.0–1.1 Free flowing
Hausner ratio [11]
1.1–1.25 Medium flowing
1.25–1.4 Difficult flowing
>1.4 Very difficult or non-flowing

Now a cylindrical extension overring is slipped onto the 1000 cm3 cylinder and
more sample material is filled in. The cylinder is mounted on the tapping device and
moved for a fixed number of tappings. In German testing standards [39] a number of
2500 with a frequency of 250 s-1 is specified.
After this the sample material is adjusted to 1000 cm3 again and weighed. The
tapped bulk density should be recorded with the parameters of its measurement.
The difference between bulk density and maximum tapped bulk density provides
information about the ability of the bulk material to be compressed by gravity or
pressure. Powders can be characterized for this property by the Hausner ratio, which
is the quotient of tapped bulk density over untapped bulk density (see Table 2.11).

Porosity
Also, the volume of the hollow void space (pores) can be calculated. The ratio of the
volume of the void space (pores) and the total volume of the bulk is called porosity ε.

VH
ρ= : ð2:77Þ
VB
so

VB - VS V
ε= =1- S : ð2:78Þ
VB VB
because mB ≈ mS = m

VS
m
ε=1- ð2:79Þ
m
VB
so

ρB
ε=1- : ð2:80Þ
ρS

With ε as a relative volume of the hollow pore space, and α as a relative volume of
the solid particle space it is evident that:
α þ ε = 1: ð2:81Þ

VB total volume bulk in m3

VH volume of hollow space in bulk in m3
VS volume of solid material in bulk in m3
(continued)
96 2 Mass Density

Fig. 2.15 Particle of a

powder, schematic, with open
(II) and closed (I) pores

ρB density of bulk in kg
m-3

ρS density of solid in kg
m-3
mB mass of bulk in kg
mS mass of solid in kg
ε porosity of bulk
α relative volume of solids in bulk

When describing porosity, we took into account the volume of pores in a powder.
On closer inspection, we have to distinguish between open pores and closed pores
(Fig. 2.15). Open pores are detected when measuring cavity volume, while closed
pores are not detected. If closed pores exist in the solid particles of a fill, they will
change the solids density but not the bulk density. Conversely, the open pores have
an influence on the bulk density but not on the solids density. A distinction is made
here between external porosity and internal porosity of particles. The porosity of an
unconsolidated mass of bulk agricultural materials such as silage, straw, and hay is a
very important physical property that is needed in air flow and heat flow processes,
as well as other applications.
Measuring the solid volume of the particles in these types of materials can be very
difficult. They are not suitable for being in contact with liquids used with pycnome-
ter methods. For these types of materials, their natural porosity can be determined
directly with the use of a simple apparatus called porosity tanks. In this apparatus,
two tanks of equal volume that can be closed air-tight are connected by a manifold of
tubing with shut-off valves and a manometer as shown in Fig. 2.16.
Tank 2 is filled with sample material and sealed. Valves 2 and 3 are closed, and
compressed air is brought into tank 1. When a suitable manometer displacement is
achieved, valve 1 is also closed, and the pressure in tank 1 p1 is read and recorded
from the manometer. Then, valve 2 is quickly opened, and the new lower equilib-
rium pressure in the system p3 is measured and recorded from the manometer. From
the perfect gas law at constant temperature, the mass of air added to the system when
tank 1 was initially pressurized m can be expressed as:

p1
V 1
m= : ð2:82Þ
RS
T 1
2.2 Laboratory Methods for Determining Density 97

Fig. 2.16 Differential

pressure gas pycnometer to manometer
(porosity tanks) [17]
valve 1 valve 2 valve 3
air in air out

tank 1
tank 2

After valve 2 is opened, this same mass becomes distributed between the empty
volume of tank 1 m1 and the available hollow pore spaces in tank 2 m2 containing the
sample. Then, by conservation of mass:

m = m1 þ m2 : ð2:83Þ
p1
V 1 p
V p
V
= 3 1þ 3 2: ð2:84Þ
RS
T 1 RS
T 1 RS
T 1
and

V 2 p1 - p 3
ε= = ð2:85Þ
V1 p1

ε porosity of bulk
p pressure in Pa
V volume in m3
m mass in kg
T temperature in K
RS specific gas constant in J
K-1
kg-1

This differential pressure gas pycnometer can be operated with dry compressed
air in the simplest case [17]. When measuring, make sure that the temperature of the
gas pycnometer is constant. For approximation of air as ideal gas, see example in
Sect. 2.1.
98 2 Mass Density

Definition
The term density is used in numerous contexts, e.g. energy density, population
density, particle density, electron density, defect density, star density, charge
density, momentum density, flux density, storage density, stock density, area
density, distribution density, and much more. This chapter was about mass
density.

Further Reading

Bakery products: measuring volume after Fornet [40]

Alcohol content of beverages by pycnometry [41] method 37.00-
1
Milk: density by means of a hydrometer [41] method
L01.00-28
Fruit and vegetable juices: relative density [41] method
L31.00-1
Beer and beer wort: relative density [41] method
L36.00-3
Density determination with X-ray scanner [42]
Powder: moisture and density determination by means of [43, 44]
microwaves
Density and viscosity determination by mechanical resonance [45]
Porosity: measurement by pycnometer [46]
Density measurement by means of ultrasound [47]
Potato starch content by underwater weighing [30]
Piezoelectric MEMS resonator for density and viscosity [48]
Lysimeter: online weighing of soil and crops [49]
Mango: ripeness detection by means of relative density [50]
Density determination of food by means of X-ray image [42]
processing methods

Summary
From the density of food, first conclusions can be drawn about the composi-
tion of a food, e.g. the concentration of an ingredient or the dry matter content.
The density is used for quality characterization and process control. Classical
density determinations require a volume determination and a mass determina-
tion, therefore the knowledge of basic rules of weighing is necessary. There

(continued)
References 99

are numerous food-specific and industry-specific measurement methods for

gaseous, liquid, semi-solid, and solid materials, some of which are explained
in this chapter using examples. In addition to the methods based on a mechan-
ical, static force equilibrium, there are resonance techniques that use the
influence of mass on the frequency of an oscillator. At the end of the chapter,
application examples are listed, which can be used for further studies and as
suggestions for your own scientific work.

References
1. Becker P, Bettin H (2016) Atome zählen für Masse und Stoffmenge. PTB-Mitteilungen 126
2. Meschede D (2006) Gerthsen Physik. Springer, Berlin
3. OIML International Recommendation R111 (2004) International Organization of Legal Metrol-
ogy, Paris
4. Kohlrausch F (1996) Praktische Physik, vol 3. Teubner, Stuttgart
5. VDI-Wärmeatlas (2013) Springer
6. Kohlrausch F (1996) Praktische Physik, vol 1. Teubner, Stuttgart
7. Kessler H-G (1996) Lebensmittel- und Bioverfahrenstechnik: Molkereitechnologie. Verlag
A. Kessler, Munich
8. Hui YH (1991) Data sourcebook for food scientists and technologists. VCH, New York
9. Singh RP, Heldman DR (2001) Introduction to food engineering. Academic Press, San Diego
10. Barbosa-Canovas GV (ed) (2005) Food powders - physical properties, processing, and func-
tionality. Kluwer, New York
11. Hayes (1987) Food engineering data handbook. Wiley, New York
12. Wächter (2012) Tabellenbuch zur Chemie. Wiley VCH, Weinheim
13. Tschubik IA, Maslow AM (1973) Wärmephysikalische Konstanten von Lebensmitteln und
Halbfabrikaten. VEB Fachbuchverlag, Leipzig
14. Lewis MJ (1996) Physical properties of foods and food processing systems. Woodhead
Publishing, Cambridge, UK
15. Gruda Z, Postolski J (1980) Gefrieren von Lebensmitteln. VEB Fachbuchverlag, Leipzig
16. Grigull U (1989) Zustandsgrößen von Wasser und Wasserdampf in SI-Einheiten. 0-800 Grad
Celsius+ 0-1000 bar, Berlin
17. Mohsenin NN (1986) Physical properties of plant and animal materials. Gordon and Breach
Science Publishers, New York. https://doi.org/10.1124/mol.105.016220.marker
18. Peleg M, Bagley EB (1983) Physical properties of foods. AVI Publishing, Westport, CT
19. Töpel A (2004) Chemie und Physik der Milch. Behr’s Verlag, Hamburg
20. von Rymon Lipinski GW, Schiweck H (2007) Handbuch Süßungsmittel. Behr’s Verlag,
Hamburg
21. Choi Y, Okos MR (1986) Effects of temperature and composition on the thermal properties of
foods. J Food Process Appl 1
22. Fricke BA, Becker BR (2001) Evaluation of thermophysical property models for foods.
HVAC&R Res 7(4):311. https://doi.org/10.1080/10789669.2001.10391278
23. Souci SW, Fachmann W, Kraut H (2016) Die Zusammensetzung der Lebensmittel, Nährwert-
Tabellen. Wissenschaftliche Verlagsgesellschaft (WVG), Stuttgart
24. Kell GS (1975) Density, thermal expansivity, and compressibility of liquid water from 0 °C to
150 °C: correlations and tables for atmospheric pressure and saturation reviewed and expressed
on 1968 temperature scale. J Chem Eng Data 20
100 2 Mass Density

25. Bettin HF, Spieweck F (1990) Die Dichte des Wassers als Funktion der Temperatur nach
Einführung der Internationalen Temperaturskala von 1990. PTB Mitteilungen 100
26. Li L, Bo M, Li J, Zhang X, Huang Y, Sun CQ (2020) Water ice compression: principles and
applications. J Mol Liq 315:113750. https://doi.org/10.1016/j.molliq.2020.113750
27. D'ans J, Lax E (1949) Taschenbuch für Chemiker und Physiker. Springer, Heidelberg
28. Standards E (2011) DIN 51757 Prüfung von Mineralölen und verwandten Stoffen -
Bestimmung der Dichte. Beuth, Berlin
29. Nielsen S (2014) Food analysis. Springer, New York
30. Unterwassergewicht (1995) EG-Verordnung Nr 97/95
31. DIN 12790 - Laborgeräte aus Glas - Aräometer (2018) Beuth, Berlin
32. Lauth JG, Kowalczyk J (2016) Einführung in die Physik und Chemie der Grenzflächen und
Kolloide. Springer Spektrum, Berlin
33. Lide DR (ed) (2008) Handbook of chemistry and physics, 99th edn. CRC Press, Boca Raton, FL
34. DIN EN ISO 1183 Kunststoffe - Verfahren zur Bestimmung der Dichte von nicht verschäumten
Kunststoffen - Teil 2: Verfahren mit Dichtegradientensäule (2004) Beuth, Berlin
35. Moshenin NN (1986) Physical properties of plant and animal material. Gordon and Breach
Science Publishers, New York
36. Fukushima H, Katsube K, Hata Y, Kishi R, Fujiwara S (2007) Rapid separation and concentra-
tion of food-borne pathogens in food samples prior to quantification by viable-cell counting and
real-time PCR. Appl Environ Microbiol 73(1):92
37. Kessler HG (2002) Food and bio process engineering: dairy technology, 5th edn. A. Kessler,
München
38. ISO 8130-3 Coating powders — Part 3: Determination of density by liquid displacement
pycnometer (2021) Beuth, Berlin
39. DIN EN 1237 fertilizers - determination of bulk density (tapped) (ISO 5311) (1992) Beuth,
Berlin
40. AACC-Method 10-05. Guidelines for measurement of volume by rapeseed displacement (2000)
St. Paul, Minnesota
41. Amtliche Sammlung von Untersuchungsmethoden nach § 64 LFGB Beuth, Berlin
42. Kelkar S, Boushey CJ, Okos M (2015) A method to determine the density of foods using X-ray
imaging. J Food Eng 159:36. https://doi.org/10.1016/j.jfoodeng.2015.03.012
43. Kent M, Kress-Rogers E (1986) Microwave moisture and density measurements in particulate
solids. Trans Inst Meas Control 8(3):161
44. Austin J, Rodriguez S, Sung PF, Harris M (2013) Utilizing microwaves for the determination of
moisture content independent of density. Powder Technol 236:17. https://doi.org/10.1016/j.
powtec.2012.06.039
45. Abdallah A, Reichel EK, Heinisch M, Clara S, Jakoby B (2014) Symmetric plate resonators for
viscosity and density measurement. Procedia Eng 87:36. https://doi.org/10.1016/j.proeng.2014.
11.260
46. Sereno AM (2007) Determination of particle density and porosity in foods and porous materials
with high moisture content. Int J Food Prop 10(3):455
47. Hoppe N, Schonfelder G, Hauptmann P (2002) Ultrasonic density sensor for liquids - its
potentials and limits. Tech Mess 69(3):131. https://doi.org/10.1524/teme.2002.69.3.131
48. Manzaneque T, Ruiz-Díez V, Hernando-García J, Wistrela E, Kucera M, Schmid U, Sánchez-
Rojas JL (2014) Piezoelectric MEMS resonator-based oscillator for density and viscosity
sensing. Sens Actuators A 220:305. https://doi.org/10.1016/j.sna.2014.10.002
49. Liu X, Xu C, Zhong X, Li Y, Yuan X, Cao J (2017) Comparison of 16 models for reference crop
evapotranspiration against weighing lysimeter measurement. Agric Water Manag 184:145.
https://doi.org/10.1016/j.agwat.2017.01.017
50. Wanitchang P, Terdwongworakul A, Wanitchang J, Nakawajana N (2011) Non-destructive
maturity classification of mango based on physical, mechanical and optical properties. J Food
Eng 105(3):477. https://doi.org/10.1016/j.jfoodeng.2011.03.006
Disperse Systems: Particle Characterization
3

Systems that are made up of particles in an ambient medium are called disperse
systems. These include, e.g. for powders with which we have already worked in the
first chapters (water activity of powders, porosity of powders, etc.). Powders consist
of small solid particles surrounded by a gaseous phase. In the disperse system
“powder,” the solid particles are the disperse phase while the gaseous phase
represents the continuous phase (Fig. 3.1).
There are numerous examples of disperse systems with which we come into
contact on a daily basis. Mists and sprays, e.g., consist of liquid droplets in a gaseous
phase, while emulsions consist of liquid droplets in a continuous liquid phase. Since
both disperse phase and continuous phase can occur solid, liquid, or gaseous state,
there are numerous possible combinations. Table 3.1 shows examples of disperse
systems. Many foods, pharmaceuticals, and cosmetics are disperse systems,
e.g. emulsions, suspensions, or foams.

Definition
Disperse means “finely distributed,” a dispersion is thus a fine distribution of
one phase in another phase. Another meaning of dispersion in physics is the
wavelength dependence on optical quantities.

Example
Fog and mist are disperse systems. By increasing the temperature, the disperse
system passes into a single-phase, non-disperse system: air, transparent.

Disperse systems are always multiphase systems. The disperse phase is some-
times called the inner phase. The continuous phase is then called the outer phase. If
the continuous phase has a solubility for the disperse phase, it may happen that the

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 101
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_3
102 3 Disperse Systems: Particle Characterization

Fig. 3.1 Disperse system,

schematic, gray: disperse
phase, white: continuous
phase

Table 3.1 Examples of disperse systems

State of
State of continuous Generic term for
disperse phase phase disperse system Examples
Solid Gaseous Powder, bulk solids Cornstarch, sugar, grain
Solid Liquid Suspension Vegetable puree, cocoa drink,
rapeseed honey
Liquid Liquid Emulsion Milk, mayonnaise, salad dressing,
butter, margarine
Liquid Gaseous Aerosol Spray, fog
Gaseous Liquid Foam Whipped cream, ice cream
Gaseous Solid Solid foam Marshmallows, toast, meringue

disperse phase dissolves. Then the disperse system passes into a non-disperse,
single-phase system. Conversely, disperse phases can form due to precipitation,
insolubility, or crystallization.

Example
Add a little water to ouzo or pastis, the clear liquid becomes a solution with a
milky appearance (Louche effect). The change in concentration has produced
an emulsion. It is a disperse system of droplets of the essential oils of the anise
seed in an aqueous solution.

For the identification of disperse systems, information on the size and shape of the
particles is required. The particle size of an emulsion, suspension, or foam strongly
influences the physical properties of such systems, e.g. the creaming tendency of
cow’s milk is lowered when the fat droplets are brought to smaller diameters. The
process of homogenization therefore increases the stability of the emulsion against
creaming.
The particles of a disperse system do not necessarily have to be spherical. While
liquid droplets and small gas bubbles are usually spherical due to the capillary
pressure (see Chap. 6), this is hardly true for solid particles. Solid particles can
have a fibrous shape or be, e.g., edgy or crystalline. Unground starch grains often
have an ellipsoidal shape, particles in a powder created by crushing processes have
edgy shapes, which depend on the crushing process used.
3.1 Particle Size Distributions 103

In addition, the particle size and shape of a real product are not uniform. We
cannot simply tell the particle size of a disperse system, but must specify the particle
size distribution with the help of suitable distribution parameters. The same applies
to the particle shape. For these reasons, we will start this chapter with an introduction
to distribution functions.

3.1 Particle Size Distributions

Distributions play an important role for biological and technical quantities. Variables
such as the age of the population, the spatially different harvest quality, or the data
streams of the Internet can be described by distribution functions and associated
statistical parameters.
First, let us imagine a bowl full of glass balls (marbles) of different diameters. The
total number of balls is the population (No). We determine the diameter of each of the
spheres and assign the balls to different size classes designated by having diameters
above diameter x. If we count the number of balls N in a size class above diameter x,
we get a number distribution of our sample. Often, the relative number is plotted,
which is the quotient of the number in the category over the total number of spheres
in the population, N/No. This gives the percentage of the population in each size
category (Fig. 3.2).
By summing up the relative number in the individual size classes, we get the
distribution sum function Q. It runs from 0 to 1 (which is equivalent to 0% to 100%)
and represents the distribution of the sphere diameter in a sigmoid curve. In Fig. 3.3,
the distribution sum function Q (left) and the distribution density function q (right) of
a distributed quantity x are shown. The distribution sum has a sigmoid curve, while
the distribution density shows a bell-shaped curve. The figure illustrates also that the
distribution density function q is the derivative of the distribution sum function Q.
A distribution can be characterized by its central value and its spread. Figure 3.4
shows two distributions that differ in central value and spread for populations of two
different particulate materials. The combination of the distribution density function,
q, and the distribution sum function, Q, characterizes the differences between these
two distributions.

Fig. 3.2 Distribution

function: Plotted is the
number N of particles
occurring in a size class x
104 3 Disperse Systems: Particle Characterization

Fig. 3.3 Distribution sum function Q(x) (left) and the distribution density function q(x) (right) of a
property x, e.g. the sphere diameter x

Fig. 3.4 Median and statistical spread of two different distributions

Fig. 3.5 Mode xh, median x50, and arithmetic mean x of a distribution in comparison

Various expressions can be used as central values. Figure 3.5 shows the mode,
median, and arithmetic mean as common central values.

Median
The median of a set of numbers is the value that divides the ordered series into two
equal halves. For odd numbers n of measured values, the median value is found in
the middle, for even numbers n of measured values is calculated as the average value
of the two inner measured values.
3.1 Particle Size Distributions 105

n odd n even
x50,r = xnþ1 xn þxnþ1
2 x50,r = 2
2
2

Example
The median of the series of numbers 48 51 55 60 66 75 87 is
60 (
xarith = 63:1).
The median of the series of numbers 48 51 55 60 66 75 76 87
is 63 (
xarith = 64:8).
The median of the series of numbers 48 51 55 60 66 75 76 187
is 63 (
xarith = 77:3).

The example shows that the median is different from the arithmetic mean xarith .
While the arithmetic mean takes all values into account equally, the median is
insensitive to extreme values. This is shown in the example above by the fact that
the median value of the last two series of numbers is the same, although the values of
the number series differ. The median value is therefore referred to as a characteristic
value with a greater robustness than the arithmetic mean. In distribution functions,
the median is read by reading the value at the abscissa (x) coinciding with the
midpoint value on the ordinate axis (Q = 0.5). In the case of a particle size
distribution, the value of x at this point would be the median of the distribution.

Example
From the sum distribution of our glass sphere example, we read the median:
x50 = 12 mm. This means that 50% of the sample has a particle size below
12 mm and 50% of the sample has a particle size above 12 mm.

Mode
The maximum of the distribution density function q(x) is called mode of distribution.
The corresponding value on the abscissa is the mode value xh,r . The mode of a
distribution indicates the value that is most represented in terms of quantity. In our
example distribution of the number of glass spheres, the mode would be the glass
ball diameter that has occurred most often. It is therefore also the most likely
diameter found when sampling the glass balls. In addition to monomodal (one
maximum) distribution density functions, bimodal (two maxima) or multi-modal
distributions can occur. Disperse systems are then called mono-disperse, bi-disperse,
tri-disperse systems, etc. When a variety of modes occur, it is called a polydisperse
system (Fig. 3.6).
106 3 Disperse Systems: Particle Characterization

Fig. 3.6 Monomodal distribution (left) and bimodal distribution (right)

Quantiles
In Fig. 3.4, we have already read the x50 value (the median) from the sum distribu-
tion: It was the particle size for Q = 0.50. Similarly, we can also read the particle
sizes for other values of Q, e.g. for 25% or for 75%:

x25 is the particle size at Q = 0.25.

x75 is the particle size at Q = 0.75.

The parameters x25, x50, x75 are also referred to as quartiles (first, second, third
quartile) of a distribution. Similarly, other parameters are in use, such as

Terciles: x33, x66, x99.

Quartiles: x25, x50, x75.
Quintiles: x20, x40, x60, x80.
Sextiles: x17, x34, x50, x67, x84.
...
Deciles: x10, x20, x30, x40, x50, x60, x70, x80, x90.
Percentiles: x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13, etc.

These parameters are summarized under the generic term quantiles. The quantiles
are robust characteristic values. Common synonyms for commonly used quantiles
are [1]:

x25 Lower quartile, first quartile

x75 Upper quartile, third quartile
x50 Second quartile
x10 Lower decile
x90 Upper decile

With the help of quantiles, parameters for the width of distributions can be
formed, e.g.
3.1 Particle Size Distributions 107

Calculation Name
x75 þx25 Quartile mean
2
x75 - x25 Interquartile range, middle fifty
x75 - x25
2 Quartile distance
x90 þx10 Decile mean
2
x90 þx10 Span
x50
1 x84 - x16
2  x50  100% Coefficient of variation
x5,0 x5,3
x90,0 resp: x90,3
Uniformity index
x60,0 x60,3
x10,0 resp: x10,3
Non-uniformity index

Up to this point, we have learned some parameters for the central value and the
spread of distributions. For symmetric distribution functions, mode, median, and
arithmetic mean coincide. The normal distribution according to Gauss is an example
of such a function. However, most real distribution functions are not such symmetric
distributions and therefore require the differentiation of these central values. There
are also other parameters for characterizing the form of distribution functions or for
deviations from the normal distribution like the so-called skewness and the kurtosis
of a distribution, which can also be formed with the help of quantiles [1].

Types of Quantities
We started the chapter with the example of glass balls. The spheres had been put into
categories based on their diameter and then we had used the number of spheres per
category to create a number distribution. These categories often are called classes. If
we are dealing with a large number of particles, it saves time to weigh the balls
instead of counting them. The difference is that as quantity we now use not the
number but the mass (or volume). The different types of quantities which can be
used are listed in Table 3.2, index r is used to indicate the type of quantity used.
Distribution functions of a sample, which are based on different types of
quantities, differ in appearance and thus by central and spread parameters. The
following example helps to understand that very different curves can be created
even when the sample is the same. For this reason, it is essential to indicate in each
distribution on which type of quantity it is based.

Table 3.2 Types of Type of quantity Index r Use with distribution functions
quantities
Number r=0 Very common
Length r=1 Rare
Area r=2 Frequently
Volume r=3 Frequently
Mass r = 3a Very common
a
Mass and volume have the same dimension, they can be converted
into each other with the help of density and therefore have the same
index
108 3 Disperse Systems: Particle Characterization

Example
A simple particle sample consists of 9 spheres: 3 balls with 1 mm diameter,
3 balls with 2 mm diameter, and 3 balls with 3 mm diameter. We calculate the
number distribution and the volume distribution. For this purpose, the balls are
first put into size classes. Then the quantities in the classes are determined.
From the number the relative number N/Ntotal and the relative volume V/Vtotal
are calculated. Cumulation of the data gives the sum distribution Q. Q0 is the
sum distribution to the quantity type number (r = 0), and Q3 is the sum
distribution to the quantity type volume (r = 3) (Fig. 3.7).

Index i Class (mm) Number N

N
N total Q0 = N
N total
i
1 0.5–1.5 3 3/9 = 0.33 0.33
2 1.5–2.5 3 3/9 = 0.33 0.66
3 2.5–3.5 3 3/9 = 0.33 1.0
Ntotal = 9

Index i Class (mm) Volume V = N  πd6

3 V
V total Q3 = V
V total
i
= 0:03
3 1:6
1 0.5–1.5 V = 3  π ð1 6mmÞ = 1:6 mm3 56:6
0.03
= 0:22
3
V = 3  π ð2 6mmÞ = 12:6 mm3
2 1.5–2.5 12:6 0.25
56:6

3 2.5–3.5 3
V = 3  π ð3 6mmÞ = 42:4 mm3
42:4
56:6 = 0:75 1.0
Vtotal = 56.6 mm3

In the field of food engineering particle size distributions often are determined by
dynamic laser diffraction (refer Sect. 1.2 and Chap. 12). By this measurement
method the particle volume is recorded, i.e. volume distributions (r = 3) are
obtained. In classical sieve analysis, the particle fractions are weighed, i.e. mass
distributions (r = 3) are obtained. However, there are also electrical counting
methods for particle size analysis, so that number distributions (r = 0) are obtained.

Fig. 3.7 Number distribution

and volume distribution
(dashed) of the same sample,
here 9 spherical particles with
3 different diameters, have a
completely different
appearance
3.1 Particle Size Distributions 109

Image analysis methods in which the (shadow) area of particles is determined can be
used for area distributions. In addition, it is possible to convert different distributions
into each other (see below).

Example
A sieve analysis of a sum distribution shows that the median x50,3 = 75 μm.
This means that 50% of the sample has a particle size below 75 μm and 50% of
the sample has a particle size above 75 μm. The index 3 in x50,3 stands for r = 3
and reminds us that the sieve analysis provides a mass distribution. So 50% of
the sample here means 50% of the sample mass.
Another laboratory has obtained the median value x50,0 = 75 μm from a
particle size measurement. The index r = 0 indicates that it is a number
distribution, i.e. 50% of the particles have a particle size below 75 μm and
50% of the particles have a particle size above 75 μm. With 50% of the
particles here is meant “50% of the total number of particles.” Have you
noticed the difference?

Bottom Line
A graphical presentation of a particle size distribution can be represented as a
number distribution or volume distribution. The particle size is on the abscissa,
the quantities such as number or volume are plotted on the ordinate. The type
of quantity selected determines the measuring method and vice versa. The
particle size distribution can be represented graphically as a distribution sum
or as a distribution density.

Terminology of Particle Size Distributions

To get particle size distributions, the particles are classified according to their size.
Then the quantities in the individual size classes are determined. The amount of
particles in a class is called fraction. Each fraction contains particles which are larger
than the lower interval limit xi - 1 and less than or equal to the upper interval limit xi .
Fraction i is often characterized by the arithmetic mean xi of these interval limits. The
fraction i therefore includes all particles with the diameter x, for which applies:
xi ≥ x > xi-1 (Fig. 3.8).

Δxi = xi - xi-1
xi = xi þx2i - 1
110 3 Disperse Systems: Particle Characterization

Fig. 3.8 Illustration of

interval limits, interval (class),
and interval width (class
width)

xi-1 lower interval limit

xi upper interval limit
Δxi interval width (class width)
xi arithmetic mean of interval
i interval index

The particle size sum distribution curve is now created by a diagram in which the
particle size x is plotted on the abscissa and the quantity fraction is plotted on the
ordinate, which is less than (or equal to) this particle size Qr, i(xi) xi. This fraction
Q(xi) is obtained by subsuming the quantities of the individual fractions for which is
x ≤ xi. This is the reason for the function being called distribution sum function.
General

particle fraction ðxmin ⋯xi Þ

Qr,i ðxi Þ = ð3:1Þ
all particles ðxmin ⋯xmax Þ

in the case of laser diffraction (quantity type: volume r = 3)

particle volume fraction ðxmin ⋯xi Þ

Q3,i ðxi Þ = ð3:2Þ
particle total volume ðxmin ⋯xmax Þ

with ΔVi as the volume of the individual classes

i
1
Q3,i = ΔV i ð3:3Þ
V i=1

in the case of or sieve analysis (quantity type: mass, r = 3)

particle mass fraction ðxmin ⋯xi Þ
Q3,i ðxi Þ = ð3:4Þ
particle total mass ðxmin ⋯xmax Þ

with Δmi as mass of the individual classes

3.1 Particle Size Distributions 111

i
1
Q3,i = Δmi ð3:5Þ
m i=1

r index for type of quantity

i index for class
x particle size in m
m mass of particles in kg
V volume of particles in m3
Q distribution sum
q distribution density in m-1

The difference quotient ΔQΔxr ðixi Þ or the differential dQ

dx is referred to as distribution
r

density q(xi). The application of q(xi) above the particle size x is called distribution
density function [2].
dQr ðxi Þ
qr ðxi Þ = : ð3:6Þ
dx

ΔQr ðxi Þ
qr ðxi Þ ffi : ð3:7Þ
Δxi

Attention
In the real world of particle size analysis, the class width is not infinitesimal
small (dx) but has finite values specified by the measurement method (e.g.,
laser diffraction or sieve analysis). So we work here with

ΔQr ðxi Þ
ð3:8Þ
Δxi

Particle Dimensions
When working with particle size distributions, the question soon arises what is
particle size. To understand we first consider a single particle as in Fig. 3.9. For

Fig. 3.9 Platon’s bodies: I tetrahedron, II hexahedron, III octahedron, IV pentagon dodecahedron,
and V icosahedron
112 3 Disperse Systems: Particle Characterization

Fig. 3.10 Examples of bodies with different main dimensions: cylinder, rotational ellipsoid, and
pyramid

Fig. 3.11 Projection of a

cuboid particle to determine
the particle size. Depending b b
on the measuring direction,
different dimensions are a c a
obtained

c c
b a
a

b
b a c
c

this purpose, clear geometric dimensions such as the main dimensions of a particle
are suitable. These are the lengths of a particle in the x, y, and z directions. Highly
symmetrical bodies such as spheres, in which the length is the same in all three
spatial directions, cause fewer problems here than those bodies that are of different
lengths in x, y, and z directions. These include, e.g., cuboids and cylinders, see
Fig. 3.10.
If we look at a cuboid body under a microscope from different viewing angles, we
find that the observed dimensions depend on the direction of measurement, ref.
Fig. 3.11. Depending on the measuring direction, we get a different value for the
projection surface, the circumference, the longest or shortest tendon, the maximum
diameter, etc. This is because the cuboid is an anisometric body, in contrast to the
sphere, which as an isometric body having the same dimensions in all spatial
3.1 Particle Size Distributions 113

Table 3.3 Geometric Equivalent diameter Calculation

equivalent diameters
Volume equivalent sphere diameter dV = 3 6V
π

Surface equivalent sphere diameter dA = A

π

Projection area equivalent diameter dP = 4S

π

Projection perimeter equivalent diameter d Pe = U

π
Sieve mesh size

directions. The sphere represents the body with the highest symmetry, the symmetry
is even higher than that of Platonic bodies (Fig. 3.9). The mathematical treatment of
spherical particles therefore is easy. For this reason, the spherical shape is often used
as an approximation for real particle shapes. The sphere is also used as a reference
body to indicate the particle size by a so-called equivalent diameter.

Equivalent Diameter
An equivalent diameter is the diameter of a sphere that has the same volume as the
actual particle under consideration. With the help of an equivalent diameter, we can
quantify the particle size e.g. of an ellipsoid particle (Fig. 3.10) as the diameter of a
spherical particle that has the same volume as our ellipsoid particle. This equivalent
diameter is called the diameter of the sphere with the same volume. Analogously we
can define particle surface or circumferential equivalent diameters (see Table 3.3).
The advantage of using equivalent diameters is the mathematic simplification. Now
you have one clearly defined diameter for a particle, although the particle has
different lengths in all spatial directions. The use of equivalent diameters is a
mathematic approximation and, as with all approximations, we have to be aware
that this approximation will cause a certain error.

Attention
Using equivalent diameters like “diameter of a sphere with the same volume”
is an approximation which is well or less well fulfilled depending on the
application. At, e.g., ellipsoid starch grains this approximation is better ful-
filled than with needle-shaped crystals.

The sieve mesh size is also an equivalent diameter. In particle size analysis by
sieving (cf. Sect. 1.2) it is assumed that a particle which has passed through a certain
sieve opening has a diameter that is less than or equal to this sieve mesh size. Most
sieve openings are the square-shaped spaces on a screen mesh, but some can be
round or slit-like. Therefore, it becomes clear that we have an approximation here,
too. Its validity has to be checked product by product.
114 3 Disperse Systems: Particle Characterization

Table 3.4 Physical equivalent diameters

To be determined
Equivalent diameter Calculation experimentally:
Diameter of a sphere with the same descent d ST = 18η
c Rate of descent c
ðρS - ρF Þg
rate in the Stokes range
Diameter of a sphere with the same descent d N = 0:33 ðρ -ρFρ Þg  c2 Rate of descent c
S F
rate in the Newton range
Diameter of a sphere with the same light Chap. 12 Scattered light
scattering intensity
Sieve mesh size Particle fraction on
sieve

Table 3.5 Examples of equivalent diameters

Measurement Equivalent diameter used Symbol in Fig. 3.12
Sieving Sieve mesh size dsieve
Sedimentation Diameter of stokes equivalent sphere dStokes
Laser diffraction Diameter of light scattering equivalent sphere dV
Flow resistance Diameter of surface equivalent sphere dA
Weighing Diameter of mass equivalent sphere dW

Fig. 3.12 Symbols of

equivalent diameters used in
Table 3.5

In addition to geometric equivalent diameters, there are physical equivalent

diameters. Here, we compare the particle shape with a spherical particle that behaves
the same as the real particle with regard to a certain physical property. Thus, we can
specify the particle size of, e.g., our ellipsoidal particle as the diameter of a sphere
that shows the same rate of descent to a specified experiment. Some physical
equivalent diameters are listed in Table 3.4.
V volume in m3
A area in m2
U circumference in m
d diameter in m
(continued)
3.1 Particle Size Distributions 115

η dynamic viscosity of the fluid phase in Pa∙s

g gravitational acceleration m∙s-2
c rate of descent in m∙s-1
ρF density of fluid phase in kg∙m-3
ρS density of particles in kg∙m-3

If we want to use the volume equivalent sphere to represent size of a particle, we

have to determine the volume of the particle and use the corresponding determina-
tion equation in Table 3.3 to get the diameter of the equivalent sphere. If we want to
calculate the surface equivalent diameter, we have to determine the surface area of
the particle, and if we want to determine the Stokes equivalent diameter, we need to
determine the rate of descent c of the particle, etc. Conversely, the preferred
equivalent diameter should be selected according to which measurement method is
available. If we have a way to measure Stokes descent rates of particles, we will opt
for Stokes equivalent diameter (refer Table 3.4).

Example
The equivalent diameter of a sphere with the same volume shall be specified:

(a) for an ideal cube with edge length a = 0.7 μm

(b) for an ideal tetrahedron with edge length a = 0.7 μm
(c) for an ideal octahedron with edge length a = 0.7 μm

Solution:

π 3
V= d
6 V

3 6V
dV =
π

(a) The volume is V = a3

3 6V 3 6  a3 3 6
dV = = =a 
π π π
3 6
dV = 0:7 μm  = 0:87 μm
π

(continued)
116 3 Disperse Systems: Particle Characterization

p
a3  2
(b) The volume is V = 12

p p
3 6V 36  a3  2 3 2
dV = = =a 
π 12π 2π
p
3 2
d V = 0:7 μm  = 0:43 μm
2π

p
a3  2
(c) The volume is V = 3

p p
3 6V 36  a3  2 3 2  2
dV = = =a 
π 3π π
p
3 2 2
dV = 0:7 μm  = 0:68 μm
π

Example
The equivalent diameter of a surface equivalent sphere shall be specified:

1. For an ideal cube with edge length a = 0.7 μm

2. For an ideal tetrahedron with edge length a = 0.7 μm
3. For an ideal octahedron with edge length a = 0.7 μm:

Solution:

A = π  d2A

A
dA =
π

(continued)
3.1 Particle Size Distributions 117

(a) The area is A = 6  a2

A 6  a2 6
dA = = =a 
π π π
6
d A = 0:7 μm  = 0:97 μm
π

p
(b) The area is A = 3  a2

p p
A 3  a2 3
dA = = =a 
π π π
p
3
dA = 0:7 μm  = 0:52 μm
π

p
(c) The area is A = 2  3  a2

p p
A 2 3  a2 2 3
dA = = =a 
π π π
p
2 3
d A = 0:7 μm  = 0:74 μm
π

Particles are often analyzed using imaging techniques. One common method is to
take numerous snapshots of the particle surface in different orientations to obtain
images of the particle surface from different directions of projection. The projection
area can be imagined as the area of shadow caused by the particle. Image analysis
software derives the particle size from these projected areas. The objective is to
determine how closely the particle shape can approach that of a sphere because
spherical shapes are the simplest to use for mathematical calculations. This can be
done by finding the diameter of the smallest circle that completely encloses the
projected area of the particle and comparing the particle projected area with the total
area enclosed by the circle (projected area of a sphere). As would be expected, the
difference between these areas will be relatively small for rounded-body objects
compared to particles with projected areas that deviate widely from a circular shape
(see example in Fig. 3.13). Here image analysis can use a number of equivalent
diameters too [3] (Table 3.6).
118 3 Disperse Systems: Particle Characterization

Fig. 3.13 Projection of a

single particle: The Martin
diameter xMa divides the
projection surface into two
equal halves, Feret diameter:
xF, longest chord: xC,max

Table 3.6 Lengths from projection surfaces [4]

Definition Symbol
Feret diameter Vertical distance between two tangents xFe
Martin diameter Length of the segment that halves the projection area xMa
Longest chord Largest length of projection area xC,max

Particle Form
So far, we have dealt with regularly shaped particles. However, we often must deal
with particles having irregular shapes. Powders produced by spray drying often
consist of ellipsoidal particles. Powders produced by crushing processes often have
very irregularly shaped particles. Regularly shaped particles similar to Platon bodies
(Fig. 3.9) are obtained by crystallization of solids. In the case of crystals, a distinc-
tion is made between shape and habitus. The shape depends on the structure of the
solid-state lattice (e.g., octahedral, tetrahedral, prismatic, cubic), while the habitus is
dominated by the manufacturing process. A cubic crystal (the shape is cubic) can
have a cuboid, platelet-shaped or needle-like habitus.
To characterize particle shape, indicators such as aspect ratio (ratio of length to
width or ratio of largest to smallest dimension) or form factors are used as a basis
upon which to compare the particle shape with the shape of a sphere. These are taken
from projected areas of the body in all three dimensions (Fig. 3.14). The ratio of
specific surface area AV of a particle and the specific surface area of the surface
equivalent sphere is the form factor used in DIN 66141 [5]. It is for spheres and
cubes fr = 1, for elongated shapes fr > 1 and runs toward infinity for long thin
particles such as needles (Table 3.7).
3.1 Particle Size Distributions 119

Table 3.7 Sphericity and form factor examples

Sphericity φWA Form factor fr
[6] [5] Examples
Isometric particles 1 1 Sphere, cube, octahedron,
icosahedron
Oblong particles <1 >1 Ellipsoids, short cuboids, short
cylinders
Needle-shaped →0 →1 Elongated cuboids or cylinders
particles

AV
fr = : ð3:9Þ
π  d 2V

The reciprocal of fr is called sphericity φWA after Wadell [6]. φWA has a value
range from 0 to 1. Spherical particles have φWA = 1, other particle form has φWA < 1,
the minimum value of φWA = 0 is reached for needle-shaped particle with infinite
length, refer Table 3.7.

π  d 2V
φWA = : ð3:10Þ
AV
With the help of equivalent diameters, further form characteristics can be derived
to describe the particle shape, e.g. circularity, concavity, or convexity [7]
(Table 3.8).
Since particle shape and size are not uniform for all particles of a collective, form
factors are derived from particle form factor distributions (Table 3.7). For a compar-
ison of sphericities for different shapes and forms see Table 3.8.

Statistical Moments
The so-called statistical moments are suitable for characterizing distribution
functions. They are formally structured analogous to moments of mechanics.
Examples from mechanics are torque, mass moment of inertia, geometric moment
of inertia, etc. In order to understand the concept of statistical moments, we first want
to familiarize ourselves with integral mean values.
A weighted mean is formed by weighting the incoming values according to their
occurrence. For particle sizing, we use intervals with the width Δxi and the middle of
the interval xi . The relative amount of particles in an interval now determines the
weighting of that interval. In the case of sieve analysis, this is easy to imagine
because the relative mass ΔQi = mmi is actually determined by weighing. Now xi is
multiplied by its “relative weight” ΔQi:

xi  ΔQi ð3:11Þ

By subsuming these products for all intervals used, we get the so-called weighted
mean value x.
120 3 Disperse Systems: Particle Characterization

Fig. 3.14 Projection areas of

a soybean. Quantities such as
sphericity and particle size
depend on the direction of
measurement

Table 3.8 Sphericity of Particle form Sphericity after Wadell

particles, examples [8, 9]
Sphere 1.000
Cylinder (h = d ) 0.874
Octahedron 0.846
Cylinder (h = 2d ) 0.832
Cube 0.806
Tetrahedron 0.670
Needle →0
Maize kernel 0.655
Sucrose crystal 0.848
Wheat kernel 0.833
Oat kernel 0.555
Soy bean 0.860
Potatoes 0.780

i i
x= xi  ΔQi = xi  q  Δxi ð3:12Þ
i=1 i=1

In the case of infinitesimal small intervals, this is the integral mean

xmax

x= x  qðxÞ  dx ð3:13Þ
xmin

With an index for the type of quantity r = 3 or r = 0, we get

3.1 Particle Size Distributions 121

Type of quantity Integral mean

r=3 xmax
x3 = x  q3 ðxÞ  dx
xmin
r=0 xmax
x0 = x  q0 ðxÞ  dx
xmin
r=r xmax
xr = x  qr ðxÞ  dx
xmin

Let us first consider the case of the number distribution (r = 0):

xmax
x0 = x  q0 ðxÞ  dx is the integral mean of length x from a number distribution
xmin
q0. It is the same what we know as arithmetic mean of x. This expression is called the
statistical moment M1, 0 of a number distribution. Such integral means can be formed
not only for lengths (k = 1), but also for the squares (k = 2), cubes (k = 3), etc.
Table 3.9 lists the equations for the calculation of such statistical moments Mk,r.
Parameter r indicates the type of quantity in the distribution function (Table 3.2) and
parameter k indicates the dimension of property x.
The table does not list the statistical moments for k = 0 because of x0 = 1 and they
all have the same value, i.e. Mk, r = M0, r = M0, 1 = M0, 2 = M0, 3 = 1
A general formulation of the kth moment of a qr-distribution is

xmax

M k,r = xk  qr ðxÞ  dx ð3:14Þ

xmin

and provides the integral mean value xk of a given distribution.

Example
Known statistic moments from a number distribution are:

The statistical moment M1,0 is identical to the arithmetic mean of x1

The statistical moment M2,0 is identical to the arithmetic mean of x2
The statistical moment M3,0 is identical to the arithmetic mean of x3

The variance (= square of the standard deviation σ) of a number distribution can

also be represented with the help of statistical moments:

xmax

σ =
2
ðx - xÞ2 q0 ðxÞ  dx = M 2,0 - M 21,0 ð3:15Þ
xmin

for distributions with r > 0 the variance is [7]:

122 3 Disperse Systems: Particle Characterization

Table 3.9 Nomenclature Integral mean Statistical moment

of integral mean values and xmax M1, 0 r=0
statistical moments x0 = x  q0 ðxÞ  dx
xmin
k=1
xmax M2, 0 r=0
x20 = x2  q0 ðxÞ  dx
xmin
k=2
xmax M3, 0 r=0
x30 = x3  q0 ðxÞ  dx
xmin
k=3
xmax M4, 0 r=0
x40 = x4  q0 ðxÞ  dx
xmin
k=4
xmax M1, 1 r=1
x1 = x  q1 ðxÞ  dx
xmin
k=1
xmax M2, 1 r=1
x21 = x2  q1 ðxÞ  dx
xmin
k=2
xmax M3, 1 r=1
x31 = x3  q1 ðxÞ  dx
xmin
k=3
xmax M1, 2 r=2
x2 = x  q2 ðxÞ  dx
xmin
k=1
xmax M2, 2 r=2
x2 2 = x2  q2 ðxÞ  dx
xmin
k=2
xmax M3, 2 r=2
x2 3 = x3  q2 ðxÞ  dx
xmin
k=3
xmax M1, 3 r=3
x3 = x  q3 ðxÞ  dx
xmin
k=1
xmax M2, 3 r=3
x23 = x2  q3 ðxÞ  dx
xmin
k=2
xmax M3, 3 r=3
x33 = x3  q3 ðxÞ  dx
xmin
k=3
General: General:
xmax xmax
xkr = x  qr ðxÞ  dx
k
M k,r = xk  qr ðxÞ  dx
xmin xmin

xmax

σ 2r = ðx - xÞ2 qr ðxÞ  dx = M 2,r - M 21,r ð3:16Þ

xmin

Statistical moments are useful during converting a distribution of a given quantity

type into another quantity type, or when converting characteristic values from such
distributions. If we want to convert an existing distribution qr into the distribution qs,
we use:
3.1 Particle Size Distributions 123

x s - r  qr ð xÞ
qs ð x Þ = : ð3:17Þ
M s - r,r

The required statistic moment M can be obtained from the moment relationship

M kþs - r,r
M k,s = : ð3:18Þ
M s - r,r

with

k dimension of x
r type of quantity of the given distribution
s type of quantity of the wanted distribution

The moment relationship allows the calculation of the kth moment of the desired
distribution with quantity type s from two moments of a given distribution with
quantity type r. So with two known moments from one distribution, each integral
mean of a different distribution can be calculated.
For r = 0 (number distribution), the moment relationship is

M kþs,0
M k,s = : ð3:19Þ
M s,0

Example
A number distribution q0 shall be converted into a mass distribution q3.
with

r=0
s=3

from

xs - r  qr ð xÞ
qs ð x Þ =
M s - r,r

we get

x3 - 0  q0 ðxÞ
qs ð x Þ =
M 3 - 0,0

and the desired distribution is

(continued)
124 3 Disperse Systems: Particle Characterization

x3  q0 ð xÞ
q3 ð x Þ =
M 3,0

Example
A mass distribution q3 shall be converted into a number distribution q0.
with

r=3
s=0

from

xs - r  qr ð xÞ
qs ð x Þ =
M s - r,r

we get

x0 - 3  q3 ðxÞ
qs ð x Þ =
M 0 - 3,3

and

x - 3  q3 ð xÞ
q0 ð x Þ =
M - 3,3

using the moments relation

M kþs - r,r
M k,s =
M s - r,r

with

k= -3
s=3

we get

M - 3þ3 - 0,0 M 0,0 1

M - 3,3 = = =
M 3 - 0,0 M 3,0 M 3,0

so the desired function is

(continued)
3.1 Particle Size Distributions 125

M 3,0  q3 ðxÞ
q0 ð x Þ = :
x3

There are a number of parameters for particle distributions that can be expressed
by statistical moments. Table 3.10 lists some examples.
From the graphical representation of the distribution in Fig. 3.15, the mode and
the median of a particle size distribution can be read: The mode is the particle
diameter belonging to the maximum of the bell-shaped distribution density function
q(x). The median is the particle diameter at Q(x) = 0.5. The first moment M1,r of the
distribution is the “balance point” of the distribution density function qr. Figure 3.15
illustrates the coexistence of these different parameters of a distribution.

Specific Surface, Sauter Diameter, De Brouckere Diameter

In addition to particle size, the specific surface area is another important parameter
for characterizing particle fineness. A specific surface is the quotient of surface area

Table 3.10 Moment nota- p

xk,r = k M k,r Mean particle size
tion for particle p
xk,0 = k M k,0 Arithmetic mean particle size
characterization
x1,0 = M 1,0
p
x2,0 = 2 M 2,0
p
x3,0 = 3 M 3,0
M rþ1,0
x1,r = M 1,r = M r,0
Weighed mean particle size
x1,1 = M 1,,1 = M 2,0
M 1,0
M 3,0
x1,2 = M 1,2 = M 2,0
x1,3 = M 1,3 = M 4,0
M 3,0

Fig. 3.15 Median (x50,r),

mode (xmod), and 1.moment
( ) of a distribution density
function [4]

( )
( )
126 3 Disperse Systems: Particle Characterization

to mass of a particle. Occasionally, the quotient of surface area to volume of a

particle is also called a specific surface. The following table compares both
definitions. With the density of the particles, ρS both properties can be converted
into each other:

AV = ρS  Am ð3:20Þ

Specific surface area of bodies

Designation Definition SI unit

Specific surface area based on volume AV = A
V
m-1
Specific surface area based on mass Am = A
m
m2∙kg-1

For a spherical particle with the diameter d it is

π  d2
AV = π
ð3:21Þ
6d
3

so
6
AV = ð3:22Þ
d
It can be seen that the specific surface area grows inversely proportional to the
particle diameter and can be a suitable property for characterizing the particle size.

Example
Assuming sugar icing (particle density) (ρS = 1500 kg  m-3) consists of
uniform, spherical particles with diameter d = 10 μm the specific surface area
can be estimated:

6 6
AV = = = 6  105 m - 1
d 10 μm

AV 6  105 m - 1
Am = = = 400 m - 2  kg - 1
ρS 1500 kg  m - 3

If the specific surface area is related to the bulk density of the sugar icing
ρB = 700 kg  m-3 the result is

AV 6  105 m - 1
Am = = = 857 m - 2  kg - 1
ρB 700 kg  m - 3

For non-spherical shapes, the Heywood factor f is introduced

3.1 Particle Size Distributions 127

A 6
AV = = f
V x
f Heywood factor
A particle surface
V particle volume
x particle size

Thus, the specific surface area for a particle collective is

xmax

x2  q0 ðxÞ  dx
xmin
AV = xmax : ð3:23Þ
x  q0 ðxÞ  dx
3

xmin

In moment notation it is

M 2,0
AV = ð3:24Þ
M 3,0

Sauter Diameter
The Sauter diameter d32 is again an equivalent diameter. It is the diameter of a
spherical particle that has the same specific surface area AV based on volume as the
particle collective under consideration.

6
d32 = ð3:25Þ
AV
in moment notation

6 6
d3,2 = = ð3:26Þ
AV 6M 2,0
M 3,0

so

M 3,0
d3,2 = ð3:27Þ
M 2,0

Sometimes the Sauter diameter is called the surface weighted mean.

128 3 Disperse Systems: Particle Characterization

De Brouckere Diameter
The De Brouckere diameter d4,3 is the equivalent diameter of a sphere with the same
volume as the particle under consideration. Sometimes it is called volume weighted
mean. In moment notation it is

M 4,0
d 4,3 = : ð3:28Þ
M 3,0

Example
A particle collective consists of 3 spheres with a diameter of 1 μm, 3 spheres
with a diameter of 2 μm, and 3 spheres with a diameter of 3 μm. The diameters
d4, 3 and d3, 2 and the specific surface AV can be calclated with the help of
statistic moments.
p
The arithmetic mean values xk,0 = k M k,0 are:

1 1 1
x1,0 = M 1,0 =  ð1 μmÞ1 þ  ð2 μmÞ1 þ  ð3 μmÞ1 = 2:00 μm
3 3 3
1 1 1
x2,0 = 2
M 2,0 =  ð1 μmÞ2 þ  ð2 μmÞ2 þ  ð3 μmÞ2 = 2:16 μm
3 3 3
1 1 1
x3,0 = 3
M 3,0 =  ð1 μmÞ3 þ  ð2 μmÞ3 þ  ð3 μmÞ3 = 2:29 μm
3 3 3
1 1 1
x4,0 = 4
M 4,0 =  ð1 μmÞ4 þ  ð2 μmÞ4 þ  ð3 μmÞ4 = 2:39 μm
3 3 3
The moments with r = 0 are

M 1,0 = 2:00 μm

M 2,0 = ð2:16 μmÞ2

M 3,0 = ð2:29 μmÞ3

M 4,0 = ð2:39 μmÞ4

M rþk,0
with xk,r = M k,r = M r,0 we get

M 4,0 ð2:39Þ4
d 4,3 = = μm = 2:72 μm
M 3,0 ð2:29Þ3

(continued)
3.1 Particle Size Distributions 129

M 3,0 ð2:29Þ3
d 3,2 = = μm = 2:57 μm
M 2,0 ð2:16Þ2

M 2,0 ð2:16 μmÞ2

AV = 6  =6  = 2:33  106 m - 1
M 3,0 ð2:29 μmÞ3

Attention
For diameters calculated from moments different nomenclatures are used:
The Sauter diameters are abbreviated x1,2 and also d3, 2. Both are identic:
x1,2 = M 1,2 = M
M 2,0 = d 3,2 .
3,0

M 4,0
Analog for the De Brouckere diameter: x1,3 = M 1,3 = M 3,0 = d4,3 .

Empirical Distribution Functions

Particle size distributions of disperse systems have to be determined experimentally.
The resulting distribution curves Qr(x) over x can be modeled with empirically found
mathematical relationships. For many process engineering applications, it is impor-
tant to know “what is the mean” and “how broad” is the distribution. Options used to
statistically quantify these characteristics about distributions are statistical
parameters, such as the “norm” of a distribution and the standard deviation from
the norm. For every day work in engineering it is useful to have linear functions
instead of non-linear curves that require more complicated mathematical equations.
Particle size distributions can be transformed into linear functions by using appro-
priate transformation model functions. In the following subsections, we will learn
about some options to represent complete distribution functions by two
parameters only.

GGS Distribution
The function after Gates, Gaudin, and Schuhmann is a power-law function. By
plotting values onto logarithmic paper, the distribution function appears as a straight
line which can be easily quantified by a slope and an intercept:
m
x
Q r ð xÞ = ð3:29Þ
xmax

lgQr ðxÞ = m  lgx - m  lgxmax ð3:30Þ

x ≤ xmax
130 3 Disperse Systems: Particle Characterization

When a particle size distribution function can be fitted with a GGS distribution
only two parameters are sufficient to represent the total curve. These are the slope m,
representing the width of the original distribution and xmax which locates the point of
the curve where Qr (x) = 1.

Log-Normal-Distribution

1 1 x-x 2
qr ð x Þ = p  exp - ð3:31Þ
σ 2π 2 σ

with the median x50, r taken for x, σ as standard deviation of x and

x-x
c=
σ
it is

1 c2
qr ð cÞ = p  e - 2 ð3:32Þ
2π
Using special grid paper a log-normal-distribution appears as a straight line in the
diagram. Also, a particle size distribution can be plotted as a straight line which is
characterized by its median x50, r and the standard deviation σ of the distribution.

RRSB Distribution
The RRSB distribution (after Rosin, Rammler, Sperling, and Benett) is a
two-parameter exponential function:

x n
1 - Qr ðxÞ = exp - ð3:33Þ
x0
with

R r ð xÞ = 1 - Q r ð xÞ ð3:34Þ

it is

1
lg lg = n  lgx - n  lgx0 þ lg ðlgeÞ = n  lgx þ d ð3:35Þ
R
Using special grid paper (RRSB net) our particle size distribution appears as a
straight line in the diagram which can be characterized by x′ and the slope n only.
There is no guarantee that an experimental particle size distribution can be
represented with one of the models mentioned. Whether a model can be used or
choosing what model will best fit the data depends on the food sample properties that
result from the process of its preparation, e.g. milling process or crystallization
process.
3.2 Measurement of Particle Size Distributions 131

3.2 Measurement of Particle Size Distributions

To determine the particle size, there are a number of different physical measurement
methods, e.g. electrical impulse methods, optical methods such as microscopy or
laser scattering or gravimetric size classification. In order to obtain reproducible size
distributions, sampling technology and sample division are of great importance.
Disperse systems such as emulsions, foams, and suspensions can undergo changes
due to contact with diluents (coalescence, flocculation, dissolution, etc.). Powders
can change due to contact with humidity or solvents. The method of sampling shall
ensure that the sample is representative of the disperse system under consideration.
There are special methods for dividing the sample into similar analytical samples
[4]. It cannot be emphasized enough that despite all the progress in computing power
and automatic evaluation of particle size distributions, the manual errors in the
handling of the samples must not be ignored. In order to create more safety here,
standard samples and comparison with different measuring methods are advisable.

Optical Methods
Optical methods of particle size measurement include microscopic methods and
camera methods, which are operated with automatic image processing system.
Microscopic methods, both light microscopic and electron microscopic, have the
advantage that the particles can be visually assessed in terms of their shape and size.
Their main disadvantage is that comparatively few particles of a sample can be
considered at one time. When we realize that a powder sample can consist of billions
of particles and we look at a few 10 or 100 of them only, it becomes clear that the risk
of misinterpretation is high here.

Example
A spherical particle with a diameter of 10 μm has a volume of

-6 3
π  d3 π  10  10 m
VK = = = 5:2  10 - 16 m3
6 6
1 g of a powder with a solids density of 1300 kgm-3 has a volume of
10 - 3 kg -7 3
V = mρ = 1300 kgm‐3 = 7:7  10 m
This corresponds to
7.7  10-7m3/5.2  10-16m3 = 1.5  109 particles with a diameter of 10 μm.

Scattered light methods have become widely used in which monochromatic laser
light is scattered on the particles to be determined and the scattered light distribution
is analyzed [10, 11]. Figure 3.16 illustrates that scattering of light on a particle can be
caused by reflection, refraction, or diffraction. In laser diffraction (low angle laser
light scattering, LALLS) the effect is used that the diffraction angle increases with
decreasing particle diameter. This diffraction angle is measured and the particle size
132 3 Disperse Systems: Particle Characterization

Fig. 3.16 Scattering of light

waves (1) on a single particle
(2): diffraction (3), absorption
and refraction (4), and 1 3
reflection (5) 2
4

is calculated with the help of the theory of Fraunhofer. The light intensity of a certain
deflection angle is a measure of the number of particles of this diameter.
Fraunhofer’s theory only accounts for the contour of the particle, but not to the
material of which the particle is made. This results in good validity in cases where
the wavelength of the light used is much greater than the particle size. When using
visible light, e.g., a He-Ne laser with a wavelength of 680 nm, Fraunhofer’s theory
provides good results for particle sizes above about 50 μm.
For smaller particle sizes, Mie´s theory is used, and the refractive properties of the
particle, i.e. h. the ability to refract and absorb, are required as input variable. This
can be achieved by using the complex refractive index. The real part of the complex
refractive index (see Chap. 12) characterizes the refraction of light in the particle
while the imaginary part of the complex refractive index describes the light absorp-
tion of the particle. Mie’s theory assumes the scattering particles to be spherical,
smooth, and homogeneous in terms of refractive index. Particle size analysis using
laser diffraction based on the Mie theory provides good results for small particles up
to a size of 0.1 μm. Smaller particles can be examined, e.g., by photon correlation
spectroscopy (PCS) or dynamic light scattering (DLS).
Flow cells in the beam allow the characterization of particles while they flow
through the device by a carrier liquid. Liquid dispersants are used as carrier fluids for
emulsions and suspensions, while powders, dusts, and aerosols are moved into the
laser beam by an air jet. In this way, particle size distributions based on very large
particle numbers can be created in less than a minute. Figure 3.17 shows the
principle of a laser diffractometer schematically. By installing several detectors
(forward and backward scattering) and the use of several wavelengths, the perfor-
mance of such devices can be further increased. The particle size distributions output
by laser diffractometers are volume-based (r = 3) size distributions. Conversion into
distribution curves of other types of quantities r is also possible.
A limitation of laser diffraction is that, like all optical particle measurement
methods, the concentration of the disperse phase must not be too high. In order for
the diffraction angle to be assigned to the particle diameter, multiple diffraction must
be avoided. This is best achieved by sufficient dilution of the sample. If irregularly
shaped particles are examined by laser diffraction, it should be noted that the particle
3.2 Measurement of Particle Size Distributions 133

Fig. 3.17 Principle of laser scattering, schematic: I light source, II sample, III diffraction light, IV
lens, V focal plane, VI detector V, and VII light intensity distribution at the detector

diameter output is an equivalent diameter, namely the diameter of a spherical particle

with the same scattering light behavior.

Dynamic Light Scattering (DLS): Photon Correlation Spectroscopy (PCS)

When observing dormant samples with very small particles, it is found that the
particles perform slight movements that counteract sedimentation. This Brownian
molecular movement depends on the diffusion coefficient of the particle. Assuming
spherical particles subjected to a Stokes law flow, the diffusion coefficient is
inversely proportional to the particle radius. Stokes–Einstein relationship is
given by:
kB  T
D= ð3:36Þ
6π  r  η
D diffusion coefficient in m2 s-1
kB Boltzmann’s constant
T temperature in K
r particle radius in m
η dynamic viscosity in Pa s

By evaluating the laser light scattering caused by Brown´s motion of the particles,
particle diameters up to the order of 1 nm can be determined from the Stokes–
Einstein relationship. This technique is called dynamic light scattering (DLS),photon
correlation spectroscopy (PCS), or quasi-elastic light scattering (QELS) and is used
for suspended colloids, micelles, emulsions, i.e. proteins, fats, and carbohydrates.
The particle diameter determined here for a particle collective again is an equivalent
134 3 Disperse Systems: Particle Characterization

Fig. 3.18 Electrozone

counting method, schematic

diameter: it is the diameter of a spherical particle with the same diffusion behavior as
the particles examined [12].

Electro-Resistive Techniques
If particles suspended in an electrolyte solution are flowing through a narrow
opening, the electrical resistance changes during particle passage. The change in
resistance correlates with the volume of the particle in the counting opening. If we
count the change in resistance and classify it by size, particle size number
distributions can be created very quickly. In Fig. 3.18, the measuring principle of
such electrozone counters is shown. The counting orifice is realized by a narrow
glass capillary with a defined diameter. In order to cover a larger measuring range,
several capillaries with different diameters are used.
Electrozone counters are widely used, originally developed as blood cell counting
devices, they can now also be used for nanoparticles [13]. A technical limitation is
that the particles must be dispersed in an electrical conductive liquid.

Size Classification and Weighing: Sieve Analysis

For powders and bulk solids, sieving is the simplest method for size classification.
The quantity of particles in each size class is generally determined by weighing. It is
therefore not a measurement of the particle size, but a gravimetric determination of
the quantities resulting from size classification. The equivalent diameter used here is
the sieve opening width. When specifying the screen opening width, it must be
indicated whether it is a wire mesh sieve (DIN ISO 3310 [14]) or perforated sheet
screen with square holes, round holes, or slotted holes.
Analytical sieving is carried out by placing the sample on a set of sieves (see
Fig. 3.19). The sieve set consists of different screens, the mesh size of which
decreases from top to bottom. By a suitable type of sieve movement, the particles
fall through those sieves whose sieve opening width is greater than the particle size
and remain on the upper sieve of those sieves whose sieve opening width is below
the particle size. Among the various types of screen movement, a distinction is made
between plan sieving (horizontal movement), throwing sieving (vertical movement),
and combined movements (tumbling sieving). The sieve movement serves to give
each particle opportunities to pass through a free sieve opening and it counteracts
blockage of the sieve openings. Vibration sieving is often used, in which a set of
3.2 Measurement of Particle Size Distributions 135

Fig. 3.19 Sieving machine. 7

1 engine housing, 2 vibrating
plate, 3 set of sieves (sieve 4
tower), 4 straps for mounting
of sieve tower, 5 pan, 6 sieve
and 7 cover
6
3

5
2

screens performs vertical movements with selectable frequency and amplitude. In

addition to the sieve movement, spherical or cube-shaped solids can be used as
screening aids. They are intended to support the screening of difficult-to-flow
powders or agglomerating particles. However, the screening aids must not change
the particle size sought, i.e. they must not lead to the crushing of the primary
particles. In the so-called air jet screen, compressed air serves as a screening aid.
Those particles that are on a sieve with mesh size xi-1 after sieving are greater
than xi–1 and less than or equal to xi . The quantity of particles that remains on a sieve
is called residue. The amount of particle that has passed through a sieve is called
passage. Particles in the size category xi - 1 ≤ x < xi have a size between xi-1 and xi.
To characterize the size category by a diameter the arithmetic mean of the interval
xi = xi þx2i - 1 is used.
Although it would be desirable to use as many narrow intervals as possible,
i.e. numerous screens, the effort is kept within limits. A set of sieves usually consists
of only 5–10 screens. It can be seen that this is a comparatively simple method with
limited particle size resolution.
In addition to the dry sieving described here, there is wet screening. It is preferred
at very small particle sizes using suitable liquid dispersants, if necessary with
ultrasonic support.
136 3 Disperse Systems: Particle Characterization

Attention
A dispersant is used for dispersing, while a solvent is used to dissolve a
substance. In particle size analysis, particles must not be dissolved, but
dispersed. For this reason a liquid is needed in which the sample is not soluble.

Sieve Analysis
Sieving of particles is a process for classifying the particles into size classes. This
process is used in the production of powders and bulk granular solids as well as on a
laboratory scale to analyze an existing powder (analytical sieving, sieve analysis,
sizing by sieving). For this purpose, a weighed sample quantity is successively
applied to sieving with decreasing mesh size. Let us consider one of these sieves:
some of the particles cannot pass through the sieve openings due to their size and
form a residue on the sieve, which is easy to determine by weighing. The part of the
particles that have passed through the sieve due to their size (the passage) falls on a
subsequent sieve, on which a separation into residue and passage takes place again.
A sieve analysis is usually carried out in such a way that the sample is placed on a set
of sieves and, after sieving each sieve is weighed individually. From the difference
between the weight and the weight of the empty sieve, the fraction of the sample is
obtained, containing particles whose size lies between the upper and lower sieve
mesh size.
In this method of particle size analysis, the sieve mesh size is used as the
equivalent diameter. The amount of particle fractions is determined by weighing,
i.e. we create a distribution of the quantity type “mass,” i.e. r = 3.
With Eq. (3.5), we get the distribution sum Q3,i and from that we calculate
distribution density for each by

ΔQ3 ðxi Þ
q3 ð x i Þ = : ð3:37Þ
Δxi
Table 3.11 shows an example of how to get from the values of a sieve analysis to
the values of Q3 and q3. In Table 3.12 this is shown in tabular form according to DIN
ISO 9276 [5].
In Table 3.12, Q3 is also referred to as D (passage). The mathematical comple-
ment to the passage is the residue R.

R=D-1 ð3:38Þ

because of

D þ R=1 ð3:39Þ
3.2 Measurement of Particle Size Distributions 137

Table 3.11 Sieve analysis, example. In the left column, the sieve set, the index i indicates the mesh
size xi and the associated particle fraction with a particle size of xi ≥ x > xi - 1
i xi/μm Δxi/μm mi/g mi
m Q3,i q3,i/μm-1
8 500 100 0.6 0.006 1.000 0.6∙10-4
7 400 85 2.2 0.022 0.994 2.6∙10-4
6 315 115 12.2 0.122 0.972 10.6∙10-4
5 200 75 34.8 0.348 0.850 46.4∙10-4
4 125 25 17.0 0.170 0.502 68.0∙10-4
3 100 37 25.2 0.252 0.333 68.1∙10-4
2 63 13 4.9 0.049 0.080 37.7∙10-4
1 50 50 3.1 0.031 0.031 6.2∙10-4
0 0,000

Table 3.12 Sieve analysis, table display according to DIN ISO 9276 [15]
i xi/μm Δxi/μm xi /μm Δmi/g Δmi/m Q3,i = D R q3,i/μm-1
1 0 50 25 3.1 0.031 0 1.0 6.2∙10-4
2 50 13 57 4.9 0.049 0.031 0.969 37.7∙10-4
3 63 37 82 25.2 0.252 0.080 0.920 68.1∙10-4
4 100 25 113 17.0 0.170 0.333 0.667 68.0∙10-4
5 125 75 163 34.8 0.348 0.502 0.498 46.4∙10-4
6 200 115 258 12.2 0.122 0.850 0.150 10.6∙10-4
7 315 85 358 2.2 0.022 0.972 0.028 2.6∙10-4
8 400 100 450 0.6 0.006 0.994 0.006 0.6∙10-4
500 0 1.000 0.000

mi mass fraction in class i in kg

mi
m relative mass fraction in class i
i sum of mass fractions with x ≤ xi
m= Δmi
i=1
ges total mass (sample mass)
m= Δmi
i=1
i
Δmi passage (distribution sum)
Q3,i ðxi Þ = m = Di
i=1
Ri = 1 - Di residue sum
mi
q3,i = m distribution density
Δxi
i class index
138 3 Disperse Systems: Particle Characterization

Example
Let us look at the passage and residue on sieve 6 in Table 3.12. It has a mesh
size of 200 μm (x6 = 200 μm). Through this sieve, 85% of the sample has
passed. The reason is that 85% of the sample has a particle size less than or
equal to 200 μm (Q3,6 = 0.85). So 15% of the sample with a size larger than
200 μm lies on the sieve 6, i.e. the residue is R = 0.15.

In Fig. 3.20, the corresponding graphical representation of the distribution sum is

shown schematically.
The diagram in Fig. 3.20 shows the representation Q above the particle size, the
distribution sum function Q. The function value is dimensionless and runs from
Q = 0⋯1, i.e. Q = 0 % ⋯100%. Reason for this is the standardization

xmax

Qr ðxmax Þ = qr  dx = 1: ð3:40Þ
xmin

The value of Q = 0.5 on the abscissa, divides the distribution into two equal
halves, 50% above and 50% is below. The corresponding particle size on the
abscissa is called the median (cf. Fig. 3.20). The distribution density function q(x)
is the first derivative of the function Q(x). The maximum of the q(x)-function
corresponds to the inflection point of the Q(x)-function. The abscissa value
associated with the maximum of the q(x)-function is called the mode. The figure
shows that the mode value and median value are not identical (there is a special case
when both have the same value). It should be noted that q is created by division with
Δx (thus the unit is m-1 or μm-1). Therefore, the value of q depends on the selected
class width Δx of the sieve analysis. The ordinate value of the q(x)-distribution can
thus vary from laboratory to laboratory and must not be misinterpreted. In some
particle size software instead of the distribution density q over x the representation ΔV
V
over x is preferred, then the ordinate is scaled in % (V/V) [16].

Fig. 3.20 Distribution sum

function with central values
3.2 Measurement of Particle Size Distributions 139

Sedimentation
Sedimentation is a particle size distribution method in which the particles are
allowed to free fall in a fluid of lesser density than that of the particles, much like
falling through air in a parachute. During free fall, the particles are exposed to
competing forces. These forces are the acceleration force of gravity (particle weight),
acting downward in direction of free fall and the buoyancy force and drag force
caused by friction acting upward in the opposite direction in which the sample falls.
According to Stokes’ law, drag force increases with relative velocity. At the begin-
ning of free fall, the particles will accelerate until they reach their terminal velocity
when the drag force becomes equal to particle weight force. At that point, there is no
longer any acceleration, and the particles will continue free fall at their constant
terminal velocity. Eventually, the particles reach the bottom as a sediment. At low
speeds (laminar case), the frictional force according to Stokes is:
F R = 6π  r  η  v: ð3:41Þ

In balance with weight force

F G = ðρK - ρF Þ  V K  g: ð3:42Þ

the rate of descent is

ð ρK - ρ F Þ  V K  g
v= : ð3:43Þ
6π  r  η
with the volume of a sphere

4
V= π  r3 ð3:44Þ
3
the rate of descent is

2  g  r 2 ð ρK - ρF Þ
v= : ð3:45Þ
9η
with radius replaces by the particle diameter

d=2  r ð3:46Þ

the rate of descent is

g  d 2 ð ρK - ρF Þ
v= : ð3:47Þ
18  η
FG weight force in N
FR friction force in N
r radius of spherical particle in m
(continued)
140 3 Disperse Systems: Particle Characterization

Fig. 3.21 Sedimentation of

spherical particles in a liquid

d diameter of spherical particle in m

V volume of spherical particle in m3
g gravitational acceleration in ms-2
v velocity of particle in ms-1
ρK density of particle in kgm-3
ρF density of fluid in kgm-3
η dynamic viscosity of fluid in Pas

Since friction force is affected by particle diameter, particles with different

diameters will fall (sink) at different velocities and reach the bottom at different
times. A classification according to particle sizes takes place with increasing sedi-
mentation time [4]. From the rate of descent, the particle size can be concluded
(Fig. 3.21).
Using the sedimentation method particle size measurements can be carried out by
placing the analytical sample in a cylinder filled with liquid to allow the particles to
sink and form the sediment. After suitable times, the sedimentation pattern in the
stationary cylinder is determined. The determination can be carried out optically
(turbidity measurement, X-ray absorption) or by gravimetrically determining the
solids or the density of samples taken.
The method can also be applied to non-spherical particles. The determined
equivalent diameter then is the diameter of a sphere with the same Stokes behavior.

Air Classification
Separation by air classification is commonly used in the process industries to
separate particles having different terminal velocities (hence, particle size diameters)
in air. The scientific principle is the same as sedimentation because it is also based on
Stokes´ law. However, it works backward from sedimentation. Instead of allowing
particles to free fall in still air, the mixture of particles is placed to rest upon a
perforated screen surface. Then, air is blown upward from beneath the perforated
screen at a controlled velocity. As soon as the velocity exceeds the terminal velocity
of the smallest size particles, the drag force acting upward will exceed the weight of
the particles, and they will become airborne, rising upward to be captured by an
appropriate particle collection system. Samples of these particles can then be taken
for analysis to measure particle size. When higher air flow velocities are used, then
3.2 Measurement of Particle Size Distributions 141

larger size particles will become airborne, and so on. In the oil seed crushing and
grain processing industries, air classification is widely used to automatically separate
the lightweight (low density) hulls, husks, and shells from the relatively higher
density seeds, grains, and kernels. In this way, the seeds continue down-stream to
the oil extraction or other process operations free of contamination from the
unwanted shells or hulls.

Other Techniques
Particle size determination is possible by measuring the specific surface area of the
particles by nitrogen adsorption. Due to the known size of N2 molecules, the increase
in mass during adsorption (see Chap. 1, Sorption) can be used to calculate the BET
surface [17] of a solid sample. Also, the specific surface area of a compressed
powder like pharmaceutical tablets can be measured that way. Another way to
determine the specific surface area is to determine the gas flow resistance of a
powder bed according to Blaine [18]. Particle sizes can also be determined by
acoustic methods since sound waves are attenuated by the particles of a suspension
[19]. Particle sizing by ultrasonic attenuation spectroscopy (UAS) is possible [20–
23]. An advantage of ultrasonic methods is the possibility to measure emulsions in
an undiluted state. For online particle sizing sensors, optical sensors and MEMS
(micro electronic mechanical systems) can be used (see Chap. 16).
Looking back at the techniques of particle size determination presented, we can
divide these into three groups.

Group Examples
Measurement of collectives Laser diffraction, dynamic light scattering, ultrasound
spectroscopic
Counting methods Electrozone and optical counters, microscopy with image
analysis, time-of-flight counter
Classification with Air classification, sieving, sedimentation, centrifugation with
subsequent measurement subsequent measurement of mass, turbidity, etc.

These different methods lead to different equivalent diameters for the particle size
under consideration. Even when directly observing particles with a camera or
microscope, it is necessary to determine in advance what is meant by particle size.
In addition, when particle sizes and shapes are calculated using statistical methods, it
is necessary to define which type of quantity is used, e.g. whether there is a number
distribution or a volume distribution. Main parameters for the characterization of
distributions are center of the distribution and its width. These information can be
displayed graphically, by quantiles from the distribution sum or by statistical
moments. While the statistical moments offer advantages from a mathematical
point of view, different nomenclatures in use can be confusing for beginners. For
this purpose, study of the presented examples here is recommended.
142 3 Disperse Systems: Particle Characterization

Further Reading

Functional nanoparticles in food [24, 25]

Natural nanoparticles [26]
Particle sizing by acoustic spectroscopy [27]
Chocolate: Particle size and sensory quality [28–30]
Coconut milk emulsion: Particle size distribution modeling [31]
Structural investigation of Pickering emulsions [32, 33]
Brownies: Influence the sugar particle size on sensory properties [34]
Characterization of double emulsions [35]
Inline particle size measurement during drying [36]
Particle size flow properties of pharmaceutical powders [37]
Lactose: Influence of particle size on dehydration and caking [38, 39]
Edible packaging with nanotechnology [40]

Summary
In bulk solids, powders, emulsions, suspensions, and other disperse systems,
the geometric dimensions of the particles play a decisive role in the product
properties. In this chapter, the basic concepts of particle size distributions are
explained step by step and explained with examples. The associated measure-
ment methods are presented comparatively and the classic analysis sieving is
calculated using an example. At the end of the chapter, application examples
are listed, which can be used for further studies and as suggestions for your
own investigations

References
1. Ferschl F (1980) Deskriptive Statistik. Springer, Heidelberg
2. Leschonski K (1984) Representation and evaluation of particle size analysis data. Part Part Syst
Charact 1:89. https://doi.org/10.1002/ppsc.19840010115
3. Bhandari B, Bansal N, Zhang M, Schuck P (2013) Handbook of food powders: processes and
properties. Woodhead Publishing, Cambridge, UK. https://doi.org/10.1533/9780857098672
4. Stieß M (1995) Mechanische Verfahrentechnik 1. Springer, Berlin. https://doi.org/10.1007/978-
3-662-08600-1
5. DIN ISO 9276 (2004) Representation of results of particle size analysis - Part 1: Graphical
representation. Beuth, Berlin. https://doi.org/10.31030/9560648
6. Wadell H (1933) Volume, shape and roundness of rock particles. J Geol 41:310
7. Rumpf H (1975) Particle technology. Springer, Berlin
8. Kurzhals H-A (2003) Lexikon Lebensmitteltechnik. Behr’s Verlag, Hamburg
9. Moshenin NN (1986) Physical properties of plant and animal material. Gordon and Breach
Science Publishers, New York
References 143

10. Xu R (2015) Light scattering: a review of particle characterization applications. Particuology

18:11. https://doi.org/10.1016/j.partic.2014.05.002
11. Rhodes M (2008) Introduction to particle technology. Wiley, Hoboken, NJ. https://doi.org/10.
1002/9780470727102
12. ISO 22412 particle size analysis – dynamic light scattering (DLS) (2017) Beuth, Berlin
13. Figueiredo M, Ferreira P, Campos E (2015) Electrozone sensing goes Nano. Wiley, Hoboken,
NJ, pp 1–19. https://doi.org/10.1002/9780470027318.a9521
14. DIN ISO 3310 (2017) Test sieves - Technical requirements and testing. Beuth, Berlin. https://
doi.org/10.31030/2743363
15. DIN ISO 9276 (2004) Darstellung der Ergebnisse von Partikelgrößenanalysen - Teil 1. Beuth,
Berlin. https://doi.org/10.31030/9560648
16. Sommer K (2000) 40 Jahre Darstellung von Partikelgrößenverteilungen—und immer noch
falsch? Chem Ing Technol 72(8):809. https://doi.org/10.1002/1522-2640(200008)72:8<809::
AID-CITE809>3.0.CO;2-K
17. DIN ISO 9277 (2014) Bestimmung der spezifischen Oberfläche von Festkörpern mittels
Gasadsorption - BET-Verfahren. Beuth, Berlin. https://doi.org/10.31030/2066286
18. DIN 66126 (2015) Bestimmung der spezifischen Oberfläche disperser Feststoffe mittels
Gasdurchströmung - Blaineverfahren. Beuth, Berlin. https://doi.org/10.31030/2313103
19. Babick F, Ripperger S (2004) Schallspektroskopische Charakterisierung konzentrierter
Emulsionen. Chem Ing Technol 76(1-2):30. https://doi.org/10.1002/cite.200406145
20. Povey MJW (2017) Applications of ultrasonics in food science - novel control of fat crystalli-
zation and structuring. Curr Opin Colloid Interface Sci 28:1. https://doi.org/10.1016/j.cocis.
2016.12.001
21. Yang H, Su M, Wang X, Gu J, Cai X (2016) Particle sizing with improved genetic algorithm by
ultrasound attenuation spectroscopy. Powder Technol 304:20. https://doi.org/10.1016/j.powtec.
2016.08.027
22. McClements DJ (2009) Ultrasonic characterization of foods and drinks: principles, methods,
and applications. Crit Rev Food Sci Nutr 37:1. https://doi.org/10.1080/10408399709527766
23. Holmes MJ, Povey MJW (2017) Ultrasonic particle sizing in emulsions. Wiley, Hoboken,
NJ. https://doi.org/10.1002/9781118964156.ch2
24. Szakal C, Roberts SM, Westerhoff P, Bartholomaeus A, Buck N, Illuminato I, Canady R,
Rogers M (2014) Measurement of nanomaterials in foods: integrative consideration of
challenges and future prospects. ACS Nano 8(4):3128. https://doi.org/10.1021/nn501108g
25. Krishna VD, Wu K, Su D, Cheeran MCJ, Wang J-P, Perez A (2018) Nanotechnology: review of
concepts and potential application of sensing platforms in food safety. Food Microbiol 75:47.
https://doi.org/10.1016/j.fm.2018.01.025
26. Rogers MA (2016) Naturally occurring nanoparticles in food. Curr Opin Food Sci 7:14. https://
doi.org/10.1016/j.cofs.2015.08.005
27. Bonacucina G, Perinelli DR, Cespi M, Casettari L, Cossi R, Blasi P, Palmieri GF (2016)
Acoustic spectroscopy: a powerful analytical method for the pharmaceutical field? Int J
Pharm 503(1–2):174. https://doi.org/10.1016/j.ijpharm.2016.03.009
28. Dahlenborg H, Millqvist-Fureby A, Bergenstahl B (2015) Effect of particle size in chocolate
shell on oil migration and fat bloom development. J Food Eng 146(0):172. https://doi.org/10.
1016/j.jfoodeng.2014.09.008
29. Renshaw RC, Dimitrakis GA, Robinson JP, Kingman SW (2019) The relationship of dielectric
response and water activity in food. J Food Eng 244:80. https://doi.org/10.1016/j.jfoodeng.
2018.08.037
30. Rohm H, Böhme B, Skorka J (2018) The impact of grinding intensity on particle properties and
rheology of dark chocolate. LWT 92:564. https://doi.org/10.1016/j.lwt.2018.03.006
31. Jena S, Das H (2006) Modeling of particle size distribution of sonicated coconut milk emulsion:
effect of emulsifiers and sonication time. Food Res Int 39(5):606. https://doi.org/10.1016/j.
foodres.2005.12.005
144 3 Disperse Systems: Particle Characterization

32. Wei ZH, Huang QR (2019) Edible Pickering emulsions stabilized by ovotransferrin-gum
Arabic particles. Food Hydrocoll 89:590. https://doi.org/10.1016/j.foodhyd.2018.11.037
33. Berton-Carabin CC, Schroen K (2015) Pickering emulsions for food applications: background,
trends, and challenges. Annu Rev Food Sci Technol 6:263. https://doi.org/10.1146/annurev-
food-081114-110822
34. Richardson AM, Tyuftin AA, Kilcawley KN, Gallagher E, O’Sullivan MG, Kerry JP (2018)
The impact of sugar particle size manipulation on the physical and sensory properties of
chocolate brownies. LWT 95:51. https://doi.org/10.1016/j.lwt.2018.04.038
35. Pulatsu ET, Sahin S, Sumnu G (2018) Characterization of different double-emulsion
formulations based on food-grade emulsifiers and stabilizers. J Dispers Sci Technol 39(7):
996. https://doi.org/10.1080/01932691.2017.1379021
36. Reimers T, Thies J, Dietrich S, Quodbach J, Pein-Hackelbusch M (2019) Evaluation of in-line
particle measurement with an SFT-probe as monitoring tool for process automation using a new
time-based buffer approach. Eur J Pharm Sci 128:162. https://doi.org/10.1016/j.ejps.2018.
11.026
37. Tay JYS, Liew CV, Heng PWS (2017) Powder flow testing: judicious choice of test methods.
AAPS PharmSciTech 18(5):1843. https://doi.org/10.1208/s12249-016-0655-3
38. Crisp JL, Dann SE, Edgar M, Blatchford CG (2010) The effect of particle size on the dehydra-
tion/rehydration behaviour of lactose. Int J Pharm 391(1-2):38. https://doi.org/10.1016/j.
ijpharm.2010.02.012
39. Carpin M, Bertelsen H, Dalberg A, Bech JK, Risbo J, Schuck P, Jeantet R (2017) How does
particle size influence caking in lactose powder? J Food Eng 209:61. https://doi.org/10.1016/j.
jfoodeng.2017.04.006
40. Katiyar V, Ghosh T (2021) Nanotechnology in edible food packaging. Springer, Singapore.
https://doi.org/10.1007/978-981-33-6169-0
Rheological Properties
4

Rheology is the branch of physics in which we study the way in which materials
deform or flow in response to applied forces or stresses. The material properties that
govern the specific way in which these deformation or flow behaviors occur are
called rheological properties. The Greek philosopher and scholar, Heraclitus
(550–480 BC) once said “πάντα ῥεῖ ” (“everything flows”). In the context of physics,
“flow” can be defined as continuous deformation over time, and it can be said that all
materials can flow. Therefore, the ability to flow is not only possessed by gases and
liquids, but also by solids to a varying degree. Indeed, we all know examples of
solids which are capable of continuous deformation over time (flow), like asphalt on
a road surface after long-term usage. It is also evident that temperature can have a
strong influence on the ability of materials to flow. For example, the asphalt road
surface will deform at a faster rate when carrying traffic during time periods of
elevated temperatures.
In this chapter, the rheology of both solids and liquids will be studied, but in the
context of being at opposite ends of a continuum (continuous spectrum) of rheologi-
cal behavior exhibited by matter in different forms. Because liquids and gases are the
forms in which matter can flow most easily, the flow of liquids and gases will be
covered in greater depth. After an introduction of ideal rheological behavior in solids
and liquids, non-ideal behavior is treated, and concepts like viscoelasticity will be
discussed. At the end of the chapter, texture measurement and quantification in food
will be discussed as an important practical application of rheology of solids to food
technology.

4.1 Elastic Properties

Solid bodies can deform elastically as a result of stress. With a uniaxial load, the
body shows a change in length, when load is applied from all directions, such as
under pressure, a body will show a change in volume. In the case of a shear load, a
change in shape occurs. Figure 4.1 shows the types of load that can be applied to a

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 145
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_4
146 4 Rheological Properties

Fig. 4.1 Load types of

technical mechanics: I
tension, II compression, III
shear, IV bending, and V
torsion

Fn
Fz

Ft

Fy Fy
Fx J
Ft
Fn Fz

I II III

Fig. 4.2 Deformations in response to load types: I uniaxial, II pressure, and III shear

body. If the deformation is reversible, i.e. the body returns to its original dimensions
and to its thermodynamic initial state after the load has been removed, this is referred
to as elastic behavior.
In uniaxial loading, the force acts perpendicular to the loaded surface and is
therefore called normal force or tensile force when being stretched or compressive
force when being compressed (I and II in Fig. 4.2). In shearing, the force acts parallel
to the surface, it is also called tangential force (III in Fig. 4.2).
Figure 4.1 in addition to these types of stress, bending and torsion are known in
technical mechanics. Bending can be understood as a combination of uniaxial tensile
and compressive loads, torsion as a combination of load types I, II, and III.

4.1.1 Stress, Pressure, Uniaxial Tension, Young’s Modulus

A one-dimensional tensile stress is caused by a one-dimensional force on a surface

directed away from the body. The quotient of force and the size of the surface area on
which it acts is called mechanical stress σ. The unit of stress is 1 N∙m-2 = 1 Pa. The
reaction of a body to a tensile or compressive stress is a change in length. The ratio of
observed change in length Δl to the original length l is called strain ε. With an ideal
elastic body, the strain increases linearly with the tension. The stress–strain curve of
4.1 Elastic Properties 147

Fig. 4.3 The stress F/A

causes a decrease in length Δl
F A
of the sample. The materials F A
modulus of elasticity E rules
the extent of deformation Δl

l Δl
l

an ideally elastic body is thus a straight line that runs through the coordinate origin.
When the tension is reduced, the ideally elastic body releases its elastic energy
without delay and completely. The stress–strain curve of an ideally elastic body thus
applies to both increasing and decreasing load or stretching. The linear relationship
between stress and strain is called Hooke‘s law (Fig. 4.2).
σ = E
ε: ð4:1Þ

dF
σ= : ð4:2Þ
dA

Δl
ε= : ð4:3Þ
l
F force in N
A area in m2
α stress in N
m-2
ε strain
l length in m
Δl change in length in m
E Young’s modulus in N
m-2

The proportionality constant in Hooke’s (Eq. 4.1) law is called Young s modulus,
also known as modulus of elasticity. The modulus of elasticity corresponds mathe-
matically to the slope of the linear curve on the stress–strain graph (ref. Fig. 4.4)
dσ
E= = tan α: ð4:4Þ
dε
When a stress–strain diagram is constructed to illustrate shear loading, the slope
of the straight line or modulus of elasticity is called Shear modulus, G. In the case of
bulk loading from all directions (pressure), the modulus of elasticity is called bulk
modulus, K.
If the stress is directed away from the body, it is called tension, and when directed
toward the body, it is called compression (unit N∙m-2 or Pa) or often simply
pressure. The change in length Δl that the body shows in response to a compressive
stress now has a negative value (decrease in length) (Fig. 4.3).
148 4 Rheological Properties

Fig. 4.4 Stress–strain

diagram of a material with a
clear yield strength Re Re H
Re L

N
V/
mm 2 dV
E
dH
D

H /%

A Hookean body is an ideal model body because its stress–strain behavior

follows Hooke’s law. However, most solid bodies show deviations from ideal elastic
behavior, e.g. non-linear behavior or a non-completely reversible behavior. Fig-
ure 4.4 shows the stress–strain diagram of a material that behaves like a Hookean
body at low tensile stresses, but shows non-linear and non-reversible behavior above
a certain load ReL. A tensile stress above ReL leads to a non-reversible, i.e. permanent
deformation of the body. Such a deformation is called a plastic deformation. As the
tensile load continues to rise, a point is reached beyond which the body can no longer
respond to any further stress and is ruptured. This is called the rupture point. The
curve section between ReL and the rupture point is the area of plastic deformation.
ReH is known as yield strength or yield point. It is the point at which the stress does
not rise or begins to fall with increasing strain. The point of maximum tension is also
called tensile strength of the body. Occasionally, ReH and ReL are also referred to as
the upper and lower yield strength.

Definition
An ideally elastic body, a so-called Hooke body, shows a linear stress–strain
curve passing through the coordinate origin. The strain is completely revers-
ible, i.e. it decreases along the linear curve without delay and completely when
the load is removed.

Figure 4.5 shows the stress–strain diagram of a material with a less clear
transition from elastic to plastic behavior. Such a smooth curve makes it difficult
to specify a precise yield stress. A rule of convention has been adopted for this
purpose which specifies the yield stress as the stress at which a straight line
originating from a designated offset of strain and drawn parallel to the original
straight-line portion of the stress–strain diagram, intersects the actual stress–strain
curve. This is illustrated in Fig. 4.5 for a designated offset strain of 0.2%. Analo-
gously, the stress at 0.1% offset can be taken as the technical limit of elasticity. In
this material example, the rupture stress, B, is lower than the measured maximum
tensile stress. Such small strain response is typical of metals such as aluminum,
which are stiff and exhibit very small strain in response to applied stresses. In
4.1 Elastic Properties 149

Fig. 4.5 Stress–strain Rm

diagram of a solid with hard-
to-detect yield strength Rp0.2

N
V/
mm 2
dV
E
dH
D
0.2% H /%

Table 4.1 Terms for interpretation of σ - ε-diagrams

Criterion
Elastic range σ - ε-function is a line through origin dσ
dε = constant
Strain and energy storage reversible dσ
≠0
dε
Non-elastic Strain and energy storage not or not completely dσ
dε ≠ constant
range reversible dσ
≠0
dε
Yield strength Maximum stress in σ - ε-function dσ
dε =0
0.2% yield Permanent strain reaches 0.2%
strength
0.1%-yield Permanent strain reaches 0.1% Technical elastic limit
strength
Rupture point Stress or strain at which rupture occurs Rupture strain, rupture
stress
Young’s Slope of the σ - ε-function in the linear range Modulus of elasticity,
modulus E-modulus

contrast, food materials are relatively soft, and a much larger value of strain would
need to be chosen as the designated offset strain. Table 4.1 compiles the terms.
The stress–strain behavior of a material can also be represented as a function of
“force over change in length (deformation)” and shown on a force-deformation
diagram.
σ = E
ε: ð4:5Þ

F Δl
=E
: ð4:6Þ
A l

A
F=E

Δl: ð4:7Þ
l
F = D
Δl: ð4:8Þ
150 4 Rheological Properties

Table 4.2 Magnitude of mechanical moduli

G-modulus E-modulus
Material in MPa in MPa Poisson’s ratio References
Silicon elastomer 0.003 0.01 0.5 [5]
Gummy bears 0.07 [6]
Carrot 7 [6]
Almond 20 [6]
Wood (beech, parallel) – 15 – [7]
Ice (-4 °C) 3.7 9.9 0.33 [7]
Quartz glass 33 76 0.17 [7]
Stainless steel X5CrNi18-10 77 200 0.3 [5]
Diamond 500 1100 0.1 [5]
Cherry pit 190 [6]

α stress in N
m-2
E Young’s modulus in Pa
ε strain
F force in N
A area in m2
Δl change in length in m
l length in m
D Hooke’s constant in N
m-1

Hooke´s constant is proportional to Young´s modulus but depends on the geo-

metric dimensions (l and A) of the sample:
A dσ A
D=E
=
ð4:9Þ
l dε l
In the case of severe deformations, it must be taken into account that the surface
A of the sample is no longer constant. This means that the quotient of force and area
(the stress) is also not constant during a load. The mechanical examination of solid
foods or “soft solids” often leads to such strong deformations, which lead to a change
in the geometry of the sample [1]. In Chap. 5 (texture), we will come back to this.
The elasticity of materials is routinely determined in materials science by means
of compression tests or tensile tests. Table 4.1 shows some terms used to describe
stress–strain curves.
The magnitude of some E-modulus can be taken from Table 4.2.

Remark
In 1675, Robert Hooke announced some inventions in the postscript of his work A
description of helioscopes, and some other instruments, Hooke’s Law was initially
published in this encrypted form: ceiiinosssttuv [2].
4.1 Elastic Properties 151

In the case of natural products and foodstuffs, it should be noted that the
mechanical properties may depend on the direction of the load. Such material
properties are called anisotropic properties. Material properties that are the same in
all spatial directions are called isotropic properties.

Example
The modulus of elasticity of wood is anisotropic. Parallel to the fiber, the
modulus of elasticity is much higher than perpendicular to the fiber. Similarly,
fibrous foods (plant stems or leaves, meat and fish) show anisotropic material
properties.

4.1.2 Three-Dimensional Stress, Bulk Compression

By applying a compressive pressure from three directions (x,y,z), the volume of a

solid, liquid, or gaseous material is reduced (Fig. 4.6). Hereby the system absorbs
energy, the so-called expansion work. An ideal elastic material shows
proportionality between load and deformation, i.e. between volume change and
pressure and returns to its initial state after the load has been removed.
The volume reduction of a body is called bulk compression. The proportionality
variable pressure dp and relative deformation is called compressibility κ. The
reciprocal compressibility is the bulk modulus of elasticity K:
dV 1
- = κ
dp =
dp ð4:10Þ
V K

1 dV
κ= - ð4:11Þ
V dp

V
dp
K= - ð4:12Þ
dV

Fig. 4.6 Compression of a Fz

cube by three-dimensional
compressive stress

Fy Fy
Fx

Fz
152 4 Rheological Properties

Fig. 4.7 Compression of a

solid by isotropic
pressure. Left: isotropic
compressibility. Right:
Example of a change in shape
due to anisotropic
compressibility

V volume in m3
p0 pressure in Pa
K bulk modulus in Pa
κ compressibility in Pa-1

Similarly, a body shows a volume increase (volume dilation) when applying

all-sided tensile stress. This occurs, e.g., during vacuum applications or during
relaxation (decompression) after a pressure treatment.
At the molecular level, the elastic volume change consists of a change in the
molecular distances or lattice spacings. Real materials show deviations from the
ideal elastic compression or dilatation behavior. These deviations can consist of
non-linear behavior or non-reversible behavior. For example, materials can show
plastic deformation or fracture as a result of three-dimensional stress.
Solid materials with anisotropic mechanical properties have K-values which are
dependent on the direction of stress. To express this mathematically, the bulk
module K is written as a tensor. The tensor has 21 linear-independent components.
At higher symmetry, as in the case of orthotropic or transverse isotropic material
properties, it may have less, e.g. nine linear-independent components Kij [3].
K xx K xy K xz
K= K yx K yy K yz ð4:13Þ
K zx K zy K zz

In the special case of isotropic material properties, all components Kij have the
same value and the tensor notation can be omitted. In the case of anisotropic
compressibility, the dimensions of a body change differently in different spatial
directions under the same pressure on all sides. The result is a change in shape of the
body (Fig. 4.7). In the case of isotropic compressibility, the shape of the solid does
not change as a result of a pressure application.
During high-pressure treatment of food [4] significant volume changes can occur.

4.1.3 Shear, Shear Modulus

If a tangential force is applied to the surface of a body, a deformation of the body

occurs, which is called shear strain. In Fig. 4.8, this type of deformation is illustrated
4.1 Elastic Properties 153

Fig. 4.8 Shear deformation

of a body by angle γ under A x Ft
influence of the tangential
force Ft
y
J

on a cube-shaped body. The extent of the deformation is expressed by the shear

angle γ.
The quotient of tangential force and area is called shear stress (τ). It has like a
perpendicular stress the physical unit N∙m-2 or Pa.
Ft
τ= : ð4:14Þ
A
In the case of an ideally elastic body, there is a linear relationship between the
shear stress τ and the tangent of the shear angle. In the ideal elastic body, the shear is
completely reversible, i.e. when the shear stress is removed, the body returns to its
geometric and thermodynamic initial state. The proportionality quantity of this law
of elasticity (Eq. 4.15) is called shear modulus G.
τ = G
tan γ: ð4:15Þ

For small angles, the approximation applies

tan γ = γ ð4:16Þ

so
τ=G
γ ð4:17Þ

Ft tangential force in N
A area in m2
τ shear stress in N
m-2
G shear modulus in N
m-2
γ angle of deformation in rad

For materials with anisotropic properties, the G-modulus can be written as a

tensor analogous to the E-modulus and K-modulus. For isotropic materials, the shear
deformation is independent of the spatial direction. Real materials show deviations
from the ideal elastic shear behavior. These deviations can consist of non-linear
behavior or non-reversible behavior. For example, materials can show plastic defor-
mation or breakage in response to an applied shear stress. The shear modulus is also
known as modulus of rigidity.
154 4 Rheological Properties

4.1.4 Transverse Strain, Poisson’s Ratio

When a material shows an elastic change in length Δl simultaneously a material

specific change in thickness Δd can be observed. The relative change in thickness Δd d
is called transverse strain. Transverse strain is proportional to longitudinal strain.
The proportionality factor is called Poisson´s ratio μ. Poisson’s ratio is a dimension-
less quantity that indicates the ratio of relative transverse strain to relative longitudi-
nal elongation.

εq
μ= - : ð4:18Þ
ε

Δd
εq
μ= - =- d
: ð4:19Þ
ε Δl
l

d initial thickness in m
Δd change in thickness in m-1
l initial length in m
Δl change in length in m
ε strain, longitudinal strain
εq transverse strain
μ Poisson’s ratio

The elastic and compressible properties of a material are related to each other via
that Poisson’s ratio μ. For the relation between μ and the elastic modulus (G-, E-, and
K-modulus) it is [5].

E
G= ð4:20Þ
2ð 1 þ μ Þ

and

E
K= : ð4:21Þ
3ð1 - 2μÞ

For positive stress Poisson’s ratio is between 0 ≤ μ ≤ 0.5. Let us look at the two
extreme values: Solids in which the elastic properties dominate over compressibility
(case A) have a high Poisson’s ratio up to μ = 0.5. Solids, on the other hand, in
which the elastic properties are subordinate to compressibility (case B), have a very
small value for Poisson’s ratio (see table below). Thus, materials can be
characterized on the basis of Poisson’s ratio.

K E G μ
Case A High Small Small μ = 0.5
Case B Small High High μ=0
4.2 Rheological Models 155

In addition, the Poisson number of a material can be used to calculate missing

moduli [8].
E 3E
K 3K ð1 - 2μÞ
G= = = : ð4:22Þ
2ð1 þ μÞ 9 K - E 2ð1 þ μÞ

E E
G 2ð1 þ μÞ
K= = =G : ð4:23Þ
3ð1 - 2μÞ 9G - 3E 3ð1 þ 2μÞ

9G
K
E= = 2Gð1 þ μÞ = 3K ð1- 2μÞ: ð4:24Þ
3K þ G

E - 2G 1 - 3K
E
3K - 2G
μ= = = : ð4:25Þ
2G 2 2ð3K þ GÞ

For cases A and B, this results in

μ=0 μ = 0.5
E 3 E E
G= = K G= =
2 2 2ð 1 þ μÞ 3
E E E
K= = K= =1
3ð1 - 2μÞ 3 3ð1 - 2μÞ

E=3
K E=3
G

4.2 Rheological Models

For a simpler mathematical description of the rheological behavior of materials,

idealized model materials are often used. We have already used the ideally elastic
material (Hooke´s body) to get to know the terms stress, stain, and shear. There is
also a model for the ideal rigid, totally incompressible body, a model body for the
ideal viscous material, a model body for the ideal plastic material, and a model body
for the ideal, frictionless flowing fluid. Table 4.3 lists those five ideal model
materials.

Table 4.3 Model Model body Rheological properties

materials with ideal
Pascal Ideal frictionless flowing
mechanical properties
Newton Ideal viscous flowing
St. Venant Ideal plastic
Hooke Ideal elastic
Euclid Ideal rigid (incompressible)
156 4 Rheological Properties

W W W W

W W W W

I II III IV
Fig. 4.9 Element symbols for the basic model materials: I Hooke element, II fraction element, III
Newton element, and IV St. Venant element

Less simple and non-ideal behavior of materials can be modeled by

interconnecting these ideal materials. Similar to electrical engineering, elements
with switching symbols are used for this purpose, which are connected in series or
in parallel to each other. Figure 4.9 shows the symbols of basic elements.
Figure 4.9 shows examples of some model materials composed by ideal elements.
For example, a Maxwell body consists of a Hooke element and a Newton element
connected in series. The Kelvin body consists of a Hooke element and a Newton
element connected in parallel. A Kelvin body can be used to model the behavior of a
viscoelastic material (Table 4.4).

Table 4.4 Examples of composed models

Model Symbol With fracture element added
Hook

Newton

(continued)
4.3 Viscous Behavior, Flow 157

Table 4.4 (continued)

Model Symbol With fracture element added
St. Venant

Maxwell

Prandtl

Kelvin

4.3 Viscous Behavior, Flow

When a sample of material is subjected to a shear stress there is a shear deformation

which we can measure by the shear angle γ. When under a constant shear stress the
shear angle grows continuously—instead of taking a fixed value—this is called
viscous behavior or flow.
While ideally elastic materials show a temporally constant shear angle under a
given shear stress, ideally viscous bodies react with a constantly increasing shear
158 4 Rheological Properties

Fig. 4.10 Two-platen model:

Flow of a material between a A x Ft
stationary and a moving plate

y
J

angle, i.e. with a constant shear rate γ_ . Figure 4.10 shows the flow of a substance
between a stationary plate and a plate moved at a constant speed.
While the elastic behavior is described by Hooke’s law τ = G
γ, the ideal viscous
behavior is a proportionality between shear stress and shear rate τ = η
γ_ . The
proportionality constant is called dynamic viscosity η. Later we will get to know
further terms such as Newtonian viscosity, Bingham viscosity, relative viscosity,
apparent viscosity, etc. (Sects. 4.3.9–4.3.12).
With shear stress

Ft
τ= ð4:26Þ
A
it is ideal elastic behavior:

τ=G
γ ð4:27Þ

and ideal viscous behavior:

τ = η
γ_ ð4:28Þ

Ft tangential force in N
A area in m2
τ shear stress in N
m-2
G shear modulus in N
m-2
γ angle of deformation in rad
γ_ shear rate in s-1
η viscosity in N
m-2
s

Solids generally exhibit elastic behavior due to the strong interaction forces
between their atoms or molecules. In contrast, fluids such as liquids and gases in
which the intermolecular interactions are much weaker than in solids, show viscous
behavior. However, some materials exhibit a transitional behavior in which they may
gradually change from solid to liquid behavior. In this way, solids can also be made
to flow if the shear stress exceeds the elastic limit (see Sect. 4.1.1). Conversely,
flowable systems can also show elastic properties (see Sect. 4.4). In the following,
simplest case, ideal viscous flow is treated first. Initially, we will treat only the
vortex-free so-called laminar flow.
4.3 Viscous Behavior, Flow 159

4.3.1 Shear Rate

When a shear stress τ is applied to a viscous body, the body shows a constantly
increasing shear angle γ over time, i.e. a constant shear rate γ_ . The shear rate is
defined as:

dγ
γ_ =
dt
and thus has the SI unit s-1.
In this case, the shear rate is calculated as the quotient of velocity of the moving
plate in Fig. 4.10 over the distance between the moving and resting plates.
Shear rate can also be calculated with reference to angular velocities and rota-
tional movement as shown in Fig. 4.11.
If we look at the angle γ in Fig. 4.11, we can write

ds
tan γ =
r
for small angles, the approximation applies

tan γ = γ

i.e.,

ds
γ= ð4:29Þ
r
so

dγ ds
= ð4:30Þ
dt dt
r

Fig. 4.11 Angular velocity

and shear angle
ds

r
J
160 4 Rheological Properties

with the circumferential velocity v in Fig. 4.11

ds
v= ð4:31Þ
dt
can be written for the shear rate:
v
γ_ = : ð4:32Þ
r
The shear rate γ_ therefore is the quotient of the circumferential velocity and the
distance between the circumference and the center of the rotational movement. At
the same time, one can consider it as the angular velocity of a rotational movement.
With a look on Fig. 4.10, we can also write
dx
tan γ = ð4:33Þ
y
then with approximation tanγ = γ
it is

dx v
γ_ = = ð4:34Þ
y
dt y
That is, here, too, the shear rate as the quotient results from the velocity of the
moving plate in Fig. 4.10 and the distance between the moving and resting plates.
For this reason, the shear rate is sometimes referred to as the velocity gradient in
the two-platen model. The lateral shear velocity gradient is the quotient of the change
in velocity vx and the path perpendicular to it, in Fig. 4.10 that means
dvx Δvx v2 - v1 v2 - 0 v
γ_ = ≈ = = = ð4:35Þ
dy Δy y2 - y1 y2 - 0 y

If we consider the material between the plates in Fig. 4.10 as a point on the
circumference located a large distance from the center of the circle in the rotational
movement illustrated in Fig. 4.11, both approaches come together. Because of the
different ways of looking at movement, there are a number of synonymous terms for
the shear rate. Here is a summary of these terms.

Definition
Shear rate: synonymous terms used
shear rate
rate of shear
shear velocity, shear speed
shear velocity gradient, velocity gradient
lateral angular velocity gradient, angular velocity
4.3 Viscous Behavior, Flow 161

Because the shear rate has the SI unit s-1 terms ending with velocity or speed are
confusing and should be avoided. However, in many fields of engineering, the term
“cycles per second or minutes, etc.” often is used.
Different names are also used for the shear angle and the shear stress or force.
Here are some of them.

Definition
Shear angle (in rad): Angle of deformation that a material shows under shear
stress. Synonym: angular deformation.

Definition
Shear force (in N): The tangential force acting on a surface. Synonymous
terms: tangential force, transverse force.

Definition
Shear stress τ in N
m‐2. The quotient of shear force and area. Synonyms used:
tangential stress.

Shear rates occur in nature and technology in a wide range of orders of magni-
tude. Examples of typical shear rate orders of magnitude are compiled in Table 4.5.
Some examples for calculating approximate values for shear rates in the field of
food technology are listed below.

Example
Shear rate when spreading or painting: With value from Table 4.5, we get

Spreading speed Thickness Shear rate

v = 0.5 m
s-1 y = 2 mm γ_ = dv
dy = 250 s - 1
-1
v=1m
s y = 0.2 mm γ_ = dv
dy = 5000 s - 1
v = 0.1 m
s-1 y = 1 mm γ_ = dv
dy = 100 s - 1
162 4 Rheological Properties

Table 4.5 Magnitudes and examples of shear rates [9, 10]

Shear rate
Process γ_ =s - 1 Example
Sedimentation of small solid 10-6–10-4 Fruit juice, medical suspensions
particles
Sedimentation of larger solid 10-4–10-1 Spice particles in dressings, suspensions,
particles paint pigments
Flow caused by surface tension 10-2–10-1 Glazes, paintings, coatings, print colors
Flow caused by gravity 10-1–101 Dripping off coated materials
Extrusion processing 100–103 Snacks, cereals, tooth paste, noodles, pet
food, polymers
Roller drum processing 101–102 Rolling of dough
Pouring 101–102 Food, cosmetics pouring out of containers
Dipping 101–102 Dipping of food in brine, of confectionary in
chocolate
Chewing/swallowing 101–102 Food, feed, pharmaceutical products
Flow-through pipe 100–103 Pumping of liquid food
Stirring/mixing 101–103 Stirring of liquid food
Painting, brushing 102–104 Paint, lip gloss, nail polish, spreads
Surface smearing 102–104 Skin creams, salves, bath gels, hand lotion
Spraying (aerosols) 103–106 Spray drying, spray painting, fuel injection
Wet milling 103–105 Food suspensions, mustard
Roller (pressing) 104–106 Printing of newspapers
Homogenization (high- 105–106 Milk homogenization
pressure nozzle)
Lubrication 103–107 Machine mount
Calendering 103–107 Polymer film coating

Example
Shear rate when rollering: With value from Table 4.5, we get

Circumferential
Roller Thickness velocity v = ω
r Shear rate
Radius 50 cm y = 100 μm 20
2π γ_ = dv
= 1 m
s - 1
= 10000 s - 1
v=
0:5 m dy 10 - 4
Rotational 60 s
frequency v = 1 m
s-1
20 RPM
Radius 20 cm y = 25 μm 100
2π γ_ = dv
= 2:6 m
s - 1
≈ 105 s - 1
v=
0:25 m dy 25
10 - 6
Rotational 60 s
-1
frequency v = 2:6 m
s
100 RPM
4.3 Viscous Behavior, Flow 163

Example
Shear rate for liquid flow in a pipe:
During laminar flow of an incompressible, ideal viscous fluid in pipe, the
shear stress is given by τW = r
Δp
2l :
and the shear rate

4
V_
γ_ W = :
π
r3
In tubular flow in a tube with a diameter of 10 cm and a flow rate of 10 l/
min, the shear rate at the wall of the tube is

4
V_ 4
10
10 - 3 m3 4
10 - 3 4
10 - 3
γ_ W = = = =
π
r 3
60 s
π
ð0:05 mÞ 3
6 s
π
5
10
3 - 6
6 s
π
53
= 1:7 s - 1 :

Example
Shear rate during sedimentation:
According to Stokes law, for Reynolds number < 1 and without
interactions between particles and fluid, the terminal velocity of a particle
d2
g
falling through a fluid can be given by v = 18
η ρK - ρfl .
For sand particles with

d = 0:5 μm:

in water, there is a difference between fluid and particle density of

ρK - ρfl = 1500 kg
m - 3 - 1000 kg
m - 3 = 500 kg
m - 3 :

With the dynamic viscosity of water of about

η = 1 mPa
s

there will be a terminal velocity of

5
10 - 7 m
9:81 m
s - 2
2
v=
500 kg
m - 3 = 6:8
10 - 8 m
s - 1
18
10 - 3 N
m - 2
s
so the shear rate is

(continued)
164 4 Rheological Properties

dv 6:8
10 - 8 m
γ_ = ffi ffi 0:1 s - 1
d_γ 5
10 - 7 s
m

4.3.2 Newtonian Flow Behavior

When there is a linear dependence between the shear stress and the shear rate of a
material, we are observing ideal viscous behavior or Newtonian flow behavior.
Fluids that behave in this way are called Newtonian fluids.
If the shear stress is applied above the shear rate in a diagram, the so-called flow
curve is obtained (Fig. 4.12 top diagram). In Newtonian fluids, the flow curve is a
straight line passing through the origin. The Newtonian flow law is
τ = η
γ_ ð4:36Þ

The slope of the flow curve is called dynamic viscosity η

dτ Δτ
η= bzw:η ≈
d_γ Δ_γ

Fig. 4.12 Flow curve (top)

and viscosity curve (bottom)
of a Newtonian fluid. With a
Newtonian fluid, the slope,
i.e. the viscosity, is the same at
all shear rates W
'W
'J•

J•

J•
4.3 Viscous Behavior, Flow 165

Table 4.6 Newtonian Material η20°C/mPa
s

fluids, examples
Carbon dioxide 0.0148
Nitrogen 0.0177
Water 1.002
Ethanol 1.20
Milk 2
Olive oil 84
Glycerol 1490

The SI unit of dynamic viscosity η is N
m-2
s-1 or Pa
s.

In order to determine the dynamic viscosity of a fluid, the ratio of shear stress and
shear rate must be measured and quotient η = d_ γ formed. For Newtonian fluids, this
dτ

quotient is the same for all shear rates, i.e. it is sufficient to determine the quotient at
a single shear rate. With the help of this one point you draw a line from the origin in a
Δτ
τ versus γ_ diagram. The application of the quotient η = d_ dτ
γ ≈ Δ_γ over the shear rate is
called the viscosity function of the fluid.
Table 4.6 lists some fluids, which in good approximation behave like Newtonian
fluids.
A distinction must be made between dynamic viscosity η and kinematic viscosity
ν, which can be calculated from dynamic viscosity using density (see Chap. 2).

η
ν= : ð4:37Þ
ρ
η dynamic viscosity in Pa∙s-1
ν kinematic viscosity in m2
s-1
ρ density of fluid in kg∙m-3
ϕ fluidity in Pa-1∙s-1

The reciprocal quantity of dynamic viscosity is called fluidity.

1
ϕ= ð4:38Þ
η

Example
The dynamic viscosity η of water at room temperature is approximately
1 mPa
s.
The density ρ of water at room temperature is about 1000 kg∙m-3.
Thus, the kinematic viscosity of water at room temperature is

(continued)
166 4 Rheological Properties

η 1 mPa
s 10 - 3 Pa
s N
m-2
s
ν= = = = 10 - 6
ρ 1000 kg
m ‐3 1000 kg
m ‐3 kg
m‐3
kg
m
s - 2
m - 2
s - 1
= 10 - 6 = 10 - 6 m2
s - 1 :
kg
m‐3

The conversion of the formerly used unit Poise (P) for the dynamic viscosity is

10 - 5 N
s
1 P = 1 dyn
s
cm - 2 = = 10 - 1 Pa
s
10 - 4 m2
1 cP = 10 - 2 P = 10 - 3 Pa
s = 1 mPa
s
1 cP = 1 mPa
s

Attention
The unit Poise (P) should not be confused with the unit Poiseuille (Pl).
1 Pa
s = 1 Pl.
The conversions of the formerly used unit Stokes (St) for the kinematic
viscosity is

1 St = 1cm2
s - 1 = 10 - 4 m2
s - 1
1 cSt = 10 - 6 m2
s - 1

In Newtonian fluids, the flow curve is a linear function starting in the origin,
i.e. the slope which is the dynamic viscosity is the same for all shear rates. This
simple case does not apply to most biological materials and foods. Here, the flow
curve is not a linear function starting in the origin, i.e. the viscosity depends on the
shear rate or the applied shear stress. This is called non-Newtonian behavior.

Definition
Laminar flow: In the two-platen model, it is assumed that all liquid layers
move laterally and glide past each other like plates. This flow without trans-
verse movement, without flow vortices is called laminar flow. A fluid flow that
involves transverse movements, i.e. vortices, is called turbulent flow.

4.3.3 Non-Newtonian Flow Behavior

A typical characteristic of a non-Newtonian fluid is that the flow curve is not a

straight line through the origin but depends upon shear rate. With such fluids, the
4.3 Viscous Behavior, Flow 167

Fig. 4.13 Flow functions:

Newtonian (2), pseudoplastic
(1), dilatant (3), Bingham
Plastic (5), and mixed
behavior (4)

Fig. 4.14 Viscosity

functions relating to figure
above, schematic: Newtonian
(2), pseudoplastic (1), dilatant
(3), Bingham plastic (5), and
mixed behavior
(pseudoplastic with yield
stress) (4)

viscosity depends on the shear rate. Typical deviations from the linear course (curve
2) are shown in Fig. 4.13. A pseudoplastic fluid exhibits a downward or convex
shape of the flow curve (curve 1) in which the slope decreases with increasing shear
rate. A dilatant fluid is reversed and exhibits an upward or concave shape of the flow
curve (curve 3). Here, the slope increases with increasing shear rate. Curve 5 is
linear, but not through the coordinate origin. In order for a measurable shear rate to
occur here, a minimum initial shear stress must be applied. This minimum shear
stress is called the yield stress of the fluid. Fluids exhibiting this type of behavior are
known as Bingham plastic fluids. In addition, there are numerous more complicated
curves that can be thought of as superimpositions or combinations of these basic
types. For example, curve 4 can be interpreted as a mixed flow behavior fluid,
exhibiting pseudoplastic behavior, but only after an initial yield stress has been
applied (combination of pseudoplastic and Bingham plastic) (Fig. 4.14).

4.3.4 Comparison: Newtonian and Non-Newtonian Fluids

While Newtonian fluids have the same viscosity at different loads (shear rates), the
viscosity of non-Newtonian fluids depends on the strength of the load or the
load time.
If we compare three different fluids at the low shear rate γ_ 1 in Fig. 4.15 so fluid
3 has the highest viscosity. If we compare them at shear rate γ_ 2 , fluid 1 has a higher
viscosity than fluid 3, and at the highest shear rate, fluid 1 has greater viscosity than
both fluids 2 and 3. The viscosity of fluid 1, on the other hand, is the same at all shear
rates. Fluid 1 is a Newtonian fluid, while fluids 2 and 3 are different pseudoplastic
168 4 Rheological Properties

Fig. 4.15 Newtonian

(1) fluid and non-Newtonian
fluid (2,3) in comparison

K 1
2

J&1 J&2 J&3

J&

fluids. This example shows that in the case of non-Newtonian fluids, the specifica-
tion of a single viscosity alone is not sufficient to describe the flow behavior. The
viscosity must be specified at a certain shear rate. Advantageous is the indication of
the entire viscosity function instead of the single value for η.

Attention
Simple viscosity meters which do not allow to control the shear rate are
suitable for Newtonian fluids only not for non-Newtonian fluids.

4.3.5 Pseudoplastic Flow Behavior

Pseudoplastic flow behavior occurs when we observe shear stress increasing at a

diminishing rate with increasing shear rate. On a shear stress–shear rate diagram, the
flow behavior curve has a convex profile in which the tangential slope is decreasing
with increasing shear rate. This means the viscosity is decreasing with increasing
shear rate. Sometimes this observation is referred to as “shear-thinning” behavior.
Figures 4.13 and 4.14 show the flow function and the viscosity function of a
pseudoplastic fluid.
It is a characteristic of pseudoplastic fluids that the increasing flow resistance
(shear stress) seems to decrease when the fluid is subjected to higher shear rates. This
behavior is assumed to be caused by decreasing molecular interactions within the
molecular structure of the fluid during flow. In the case of flowing macromolecules,
the additional effects of unfolding and reorientation of molecules may be occurring.
When these effects are completely reversible, we call this “true” pseudoplasticity.
This means that if we first increase shear rate to a given value, and then decrease it
back again, the viscosity will be fully restored to the same value as before.
An aqueous xanthan gum solution shows a reversible structural degradation
during shear, i.e. true pseudoplasticity while, e.g., the protein gel of a set yogurt is
irreversibly degraded by shear.
4.3 Viscous Behavior, Flow 169

Definition
True pseudoplasticity is when the viscosity decreases reversibly as a result of
structural degradation due to shear, i.e. after the load has been removed, the
material returns to its initial viscosity. Non-true pseudoplasticity is when the
viscosity decreases irreversibly as a result of permanent structural degradation
due to shear.

4.3.6 Thixotropic Flow Behavior

Thixotropic flow behavior occurs when the viscosity of a fluid decreases over time
while subjected to a constant shear rate. The viscosity therefore does not decrease
with increasing shear rate—as with pseudoplasticity—but with increasing time
during which the shear load takes place. A distinction is also made here between
true thixotropy and non-true thixotropy. If, after removing the shear load, the
viscosity gradually rises back to the initial value, it is true thixotropy. If the viscosity
remains permanently low, it is an irreversible structural change of the material and
one speaks of non-true thixotropy.

4.3.7 Dilatant Flow Behavior

Dilatant flow behavior occurs when we observe shear stress increasing at an increas-
ing shear rate. On a shear stress–shear rate diagram, the flow behavior curve has a
concave profile in which the tangential slope is increasing with increasing shear rate.
This means the viscosity is also increasing with increasing shear rate. Sometimes this
observation is referred to as “shear-thickening” behavior. This flow behavior is seen
in highly concentrated suspensions. Figures 4.13 and 4.14 show examples of the
flow behavior curve and viscosity curve for a dilatant fluid. It is assumed that under
increasing shear rate the liquid between the solid particles is squeezed out and the
friction between the particles increases. This increasing friction accounts for the
increasing shear stress and increasing viscosity experienced as shear rate increases.
Also, the squeezing causes a dilatation (an increase in volume) of the suspension,
and therefore the phenomenon is called dilatant flow behavior. When these effects
are completely reversible, we call this “true” dilatancy. This means that if we first
increase shear rate to a given value, and then decrease it back again, the viscosity will
be fully restored to the same value as before. Sometimes the effect of shear-
thickening is not completely reversible. This can be explained by a permanent
damage to the molecular structure in the material under investigation. In that case
we have no “true” dilatancy, but an apparent dilatancy (Fig. 4.16).
170 4 Rheological Properties

Fig. 4.16 Time dependency

of viscosity, schematic:
Thixotropy (1), Rheopexy (3).
Newtonian Fluids (2) have a
constant viscosity

4.3.8 Rheopectic Flow Behavior

Rheopectic flow behavior is observed when shear stress or viscosity increases over
time at a constant shear rate. The reason for the increasing viscosity (or shear stress)
is also assumed to stem from the intermolecular interactions causing friction to
increase with time at constant shear rate within the molecular structure of the
material. When shear rate stops, the original structure is restored, as well as the
initial viscosity. This is called “true” rheopexy. If the viscosity is not restored
because of irreversible structure damage, the behavior is called apparent rheopexy.
When observing what appears to be rheopectic behavior, it is important that we
interpret correctly what we observe. For example, when whipping liquid heavy
cream into whipped cream, we observe the need for more and more effort to maintain
the same whipping speed (viscosity increasing over time at constant shear rate). This
would appear to be rheopectic behavior. But, that is not the case here. In whipping
cream, we start out with a liquid of relatively low viscosity that becomes transformed
into a soft-solid foam as a result of incorporating air into the cream producing a
gas-liquid emulsion. We now have a totally new material with different rheological
properties, and not a liquid with rheopectic behavior.

4.3.9 Plastic Flow Behavior

If a material only begins to flow when a yield stress is exceeded, it is called plastic
flow behavior. When the material is stressed below the yield point, the substance
behaves like a solid, i.e. it can be deformed a little elastically. If the shear stress
exceeds the yield point, the viscous deformation begins, i.e. the material begins to
flow. Think of ketchup, which requires an initial impulse of shear stress induced by a
shake of the bottle in order to flow. If the shear load is reduced again and the yield
point falls below the yield point, the flow process stops and the body retains its
current shape and shape. The shape of the material has been permanently deformed
by this process, which is why it is called plastic deformation.
The yield stress (yield point) is therefore a minimum shear stress for the flow of
the material in question. In the flow curve, the yield point can be read as the yield
stress on the ordinate axis beyond which the flow occurs (see Fig. 4.13), in the
viscosity curve (see Fig. 4.14) the material shows a transition from initially infinitely
4.3 Viscous Behavior, Flow 171

high viscosity to a finite viscosity. Examples of plastic materials are short pastry,
butter, clay, chocolate melt.
But also e.g. metals can be plastically deformed. They are materials with a very
high yield point. Below the yield point, they behave like elastic solids, above them
like highly viscous fluids. The production of aluminum beverage cans is based on the
plastic deformation of a flat aluminum sheet. The deformation (the so-called deep
drawing) can take place at ambient temperature (cold forming) or at elevated
temperature. Like viscosity, the yield point decreases with increasing temperature.
Steel forging processes are therefore carried out at temperatures of, e.g., 800 °C, at
higher temperatures steel can also be cast.
The plasticity or the yield point is caused by structure-forming interactions
between the molecular components of the material. If these interactions are strong,
the system has a high yield point and high shear forces or high temperatures are
required to get the material flowing. With low structure-forming forces, lower shear
stresses are sufficient to exceed the yield point. Whipped cream is an example of a
material with a low but clearly recognizable yield point: The whipped cream changes
its shape only by external action (shear load), after which the shape is preserved. The
yield stress is low means it can be exceeded by a relatively low shear stress.
Other materials have such a low yield point that they already begin to flow under
the influence of gravity, apparently “by themselves.” Examples are unwhipped
cream, oil, water. It looks like that materials do not have a yield point. To be precise,
we must have to say that the yield point is very low, or imperceptible without
sensitive instruments.

Definition
The yield point is the minimum shear stress that must be exceeded for flowing
to take place. Thus, a portion of mayonnaise on a plate retains its shape, i.e. it
does not flow. If the yield point is exceeded, e.g. by spreading the mayonnaise
with a spoon, then it changes its shape, i.e. it flows. A drop of vegetable oil on
a flat surface, on the other hand, seems to spread “by itself.” It does not seem to
have a yield point. With a sensitive rheometer, however, the value of the yield
point can also be determined for such materials. It is so low that a flow can be
caused by gravity alone.

Example
Thickness of a chocolate coating.
What is the thickness of a coating layer on a food made with liquid milk
chocolate with a yield stress of 30 Pa?
The weight force FG of a layer with the layer thickness d is equivalent to the
tangential force acting at the surface A (shear stress). As long as this shear
stress is greater than the yield point of the liquid chocolate, the chocolate flows

(continued)
172 4 Rheological Properties

off and the layer thickness decreases, along with the weight of the layer. As
soon as the layer thickness decreases to the point where the weight of the
remaining coating leaves a shear stress equal to the value of the yield point τ0,
the flow stops, i.e. a layer of thickness d0 remains on surface A.

FG = m
g = ρ
A
d
g
FG = τ
A
ρ
A
d 0
g = τ0
A
ρ
d 0
g = τ0

τ0 yield point in Pa
A area in m2
d0 thickness in m
ρ density in kg∙m-3
g gravitational acceleration in m
s-2
FG weight force in N
m mass in kg

From this balance of forces, the thickness of the remaining layer can be
roughly calculated: The shear stress when the chocolate layer flows down with
the layer thickness d0 is:

τ0
d0 = :
ρ
g
At a yield point of τ0 = 30 Pa this results in a layer thickness of

30 Pa 30 N
m - 2
d0 = - -
=
1520 kg
m
9:81 m
s
3 2 1520 kg
m - 3
9:81 m
s - 2
30 kg
m
s - 2
m - 2
=
1520 kg
m - 3
9:81 m
s - 2
30
d0 =
10 - 3 m = 2
10 - 3 m = 2 mm
1:52
9:81
d0 = 2 mm

With this simple approach and the assumption of a constant temperature, the layer
thickness is proportional to the yield point. Figure 4.17 shows the estimated magni-
tude of achievable layer thicknesses during brushing, dripping, coating, etc. In
practice, a constant temperature is usually not given. Especially in industrial pro-
cesses, the temperature dependence of the material properties is used to shorten the
time of the processes (Table 4.7).
4.3 Viscous Behavior, Flow 173

Fig. 4.17 Thickness d0 of a d0

chocolate coating

Table 4.7 Yield points and layer thicknesses, examples of simple estimates
τ0/Pa 3 30 150
Thickness d0 200 μm 2 mm 1 cm

4.3.10 Overview: Non-Newtonian Flow Behavior

Non-Newtonian behavior is when the flow curve of a material is not a straight line
through the origin (see Fig. 4.13) or is dependent on time. That is, these are all those
cases in which the flow curve is curved or does not pass through the coordinate
origin due to the presence of a yield point or if the appearance of the flow curve
changes with the duration of the shear load. Figure 4.18 shows the subdivision of the
different cases of non-Newtonian behavior and the related terminology. In cases
where the viscosity changes with the duration of the stress (i.e., the viscosity is time-
dependent), a distinction is made between thixotropy and rheopexy. If the viscosity
is not dependent on the duration of the stress but on the shear rate, a distinction is
made between pseudoplasticity and dilatancy.
Materials which show plastic flow are also non-Newtonian materials, even when
the viscosity is not dependent on shear stress and time. Flow behavior like this is
called Bingham plastic flow, and the material is called a Bingham plastic fluid.
Flowable materials with elastic properties, the so-called viscoelastic substances, are
not ideal fluids, i.e. non-Newtonian fluids.

non-NEWTONian

non time dependant time dependant

pseudoplastic dilatant plastic visco-elastic thixotropic rheopectic

A B C D E F

Fig. 4.18 Non-Newtonian flow behavior, schematic of behavior categories and their terminology.
Examples are: (a) paint, tomato paste, (b) crystallized honey, (c) whipped egg white, solid butter,
lipstick, (d) wheat dough, and (e) ketchup
174 4 Rheological Properties

Examples
Pseudoplastic materials: ketchup, tomato paste, starch paste
Dilatant materials: starch suspension, crystallized honey
Plastic materials: cream cheese, butter, short paste
Viscoelastic materials: wheat flour yeast dough, camembert
Thixotropic materials: pudding, ketchup
Rheopectic materials: gel-forming materials

Bottom Line
Terms for the characterization of flow behavior.
Newtonian the flow curve is a straight line through the origin
Non-Newtonian the flow curve is not a straight line through the origin
Pseudoplastic viscosity decreases with increasing shear rate (shear-
thinning)
Dilatant viscosity increases with increasing shear rate (shear-
thickening)
Plastic the fluid has a yield point
Thixotropic viscosity decreases over time at a constant shear rate
Rheopectic viscosity increases over time at a constant shear rate

Examples for Misinterpretations

Apparent dilatancy or apparent rheopexy:
Vortex formation (increase in flow resistance) in the rotational rheometer
Air input during the examination in the rotational rheometer
Solvent evaporation
Chemical reaction
Gel formation during the examination
Apparent pseudoplasticity or apparent thixotropy:
By particle size change in a dispersion during the examination in the
rotational rheometer

Real materials normally have a more complicated flow behavior which can be
explained with various mixtures of the different ideal behaviors described up to here.
Sometimes approximations can help, like treating the Bingham plastic fluid as a
Newtonian fluid once the initial yield stress has been discounted.
For many food science and engineering applications, it is convenient to transform
complicated flow behavior curves and viscosity curves into mathematical equations
with constant parameters. Then only a few numbers (parameters) are sufficient to
represent and to specify the complete curves. For this purpose, model functions have
to be derived, and tested and challenged to be sure of their validity.
4.3 Viscous Behavior, Flow 175

4.3.11 Model Functions

Rheological model functions are nothing more than the mathematical equations
derived to describe the various flow behavior curves on shear stress–shear rate
diagrams. First we will focus on model functions for fluids without yield stress.
Then, we will address fluids with yield stress.
Pseudoplastic fluids are assumed to have an initial viscosity at the origin
η ðγ_ = 0Þ = η0 which then decreases with increasing shear rate. The decrease is
assumed to be the consequence of (reversible) loss of structure/network in the
material (in the German language pseudoplastic is called “strukturviskos” because
of this interpretation). When the molecular structure/network in the fluid has reached
a steady state (where the intermolecular forces acting to build the structure/network
and those acting to break it down by shear are in equilibrium) no further decrease of
viscosity is observed. This viscosity is called equilibrium viscosity η ðγ_ = 1Þ = η1 :
Figure 4.19 shows a typical profile for a flow curve of this type. The viscosity curve
(which is the derivative of the flow behavior curve) can also be constructed and is
given in Fig. 4.20. In this type of pseudoplastic material, the viscosity is changing

Fig. 4.19 Flow function

during structure loss in a
pseudoplastic fluid, schematic
Kf
K0
W

J&
Fig. 4.20 Viscosity function
during structure loss in a
pseudoplastic fluid, K0
schematic: the viscosity
decreases from η0 to η1

K
Kf

J&
176 4 Rheological Properties

Table 4.8 Model functions for fluids without yield stress, examples
Viscosity function
Name Flow function model model
Newton τ = η
γ_ η = γτ_ τ: Shear stress in Pa
γ_ : Viscosity in Pa
s
γ_ : Shear rate in s-1
η0 η0
Ferry τ= 1þC
τ
γ_ ηs = 1þC
τ
C: Constant in Pa-1
η0: Initial viscosity in Pa
s
(for C = 0) Newton)
Steiner–Steiger– τ= 1
C þA
τ2 γ_ ηs = 1
C þA
τ2 C = η1 : Constant in
0
Ory (Pa
s)-1
A: Constant in Pa-1
η0: Initial viscosity in Pa
s
η0 η0
De haven τ= 1þC
τn γ_ ηs = 1þC
τn
C: Constant in Pa-n
n: Flow behavior index
(for n = 1) Ferry)
Ostwald–de τ = K ow
γ_ n ηs = K ow
γ_ n - 1 Kow: Consistency
Waele coefficient in Pa
sn
(power law) n: Flow behavior index
(for n = 1) Newton)
Sisko τ = η1
γ_ þ b
γ_ n ηs = η1 þ b
γ_ n - 1 b: Constant in Pa
sn
η1: Equilibrium viscosity
in Pa
s
Ellis I τ = η0 þ K
γ_ n - 1 γ_ ηs = η0 þ K
γ_ n - 1 η0: Initial viscosity in Pa
s
K: Constant in Pa
sn
η0 η0
Ellis II τ= A-1
γ_ ηs = A-1 η0: Initial viscosity in Pa
s
1þτ τ 1þ τ
τ1= A: Constant
1=2 2
τðη1 Þ - τðη0 Þ
τ1=2 = 2
η1: Equilibrium viscosity
in Pa
s
η0 þη1
Peek–McLean– τ = η1 þ η01þ
- η1
τ γ_ ηs = η0 - η 1
τ þ η1 τm = τ 2
1þτm
Williamson τm

Reiner– η0 - η 1 η0 - η 1
τ = η1 þ γ_ ηs = þ η1
Philippoff 1þðττm Þ 1þðττm Þ
2 2

η1 φ1 - φ0
Reiner τ= η0 - η1 - τ 2
γ_ κ= dφ : Coefficient of
1- η0
e
κ d ðτ2 Þ

pseudoplasticity
ϕ = 1η: Fluidity

from η0 to η1 with increasing shear rate. For a mathematical description of these

sigmoid shaped curves, a number of model functions have been developed.
In general, most of these functions require two or three parameters. These
parameters are constants in the model function, whose numerical values make the
function specific for a given material. Therefore, in order to use these model
functions, these parameters have to be known. Normally, model parameters are
measured experimentally, or they be obtained from other investigators working on
the same material, who report their findings in the scientific literature. Table 4.8 lists
4.3 Viscous Behavior, Flow 177

some model functions like these. There are cases where a model function takes on
the same form as that of a simpler more easily recognized model function, like
Newton’s law. In other words, these model functions can be thought as more or less
sophisticated modifications of basic functions like those which describes Newtonian
flow behavior.
A classic question posed by investigators working on rheology of fluids is,
“Which model is the right one for my material and/ or application?” The answer
will depend on factors like:

• For what purpose is the viscosity needed?

• In what range of shear rates will the work be done?
• How precise must the data be?

These models differ in their ability for preciseness or “goodness of fit” in

approximating experimental data. For example, the model functions after Ferry,
Steiner–Steiger–Ory, De Haven, Ostwald–de Waele, Ellis I, and Sisko are valid only
in the region η < η1 where the fluid structure is not completely broken down.
For synthetic polymers (but not those in the form of dispersions or gels), the
models after Carreau, Cross, Ellis-Sisko, Phillips-Deutsch, Reiner-Philippoff,
Krieger–Dougherty are recommended [10].

4.3.12 Ostwald–de Waele Law

A very useful and not too complicated model function is the power-law equation
after Ostwald–de Waele. It is a power law with the exponential parameter n, which is
called flow exponent. It reads

τ = K ow
γ_ n ð4:39Þ

Kow consistency coefficient in Pa
sn

γ_ shear rate in s-1
τ shear stress in pa
n flow exponent

The flow behavior index n indicates the deviation of the flow curve from a straight
line, i.e. from Newtonian behavior. When the flow behavior index is less than unity
(n < 1) the function describes a pseudoplastic flow behavior curve (convex profile).
When the flow behavior index is greater than unity (n > 1) the function describes a
dilatant flow behavior curve (concave profile). For Newtonian fluids, the flow
behavior index becomes 1, and the consistency coefficient becomes the viscosity.
In other words, for Newtonian fluids is n = 1 and Kow = η.
178 4 Rheological Properties

Definition
The flow exponent of a material characterizes the curvature of the flow curve.

Definition
The consistency coefficient Kow also is called Ostwald–Faktor Kow in the
Ostwald–de Waele law. The higher the consistency factor, the tougher the
fluid.

For non-Newtonian fluids, the consistency coefficient Kow has the meaning of a
mathematical parameter with the physical unit Pa
sn. From the unit it can be seen
that Kow it is not same as viscosity, which has the unit Pa
s. Its value is determined
experimentally. In case of a Newtonian fluid we have Kow = η.
The Ostwald–de Waele law allows the calculation of ηs of a material at a given
shear rate or a given shear stress.
For this purpose, the Ostwald–de Waele law is used in the logarithmic form:
From τ = K ow
γ_ n we get lgτ = lgK ow þ n
lg_γ .
In the double-logarithmic (log-log) graphical representation, a straight line
results. The slope of the straight line is the flow behavior index (exponent) n. The
value of Kow we get from
lgK ow = lgτ - n
lg_γ : ð4:40Þ

At a shear rate of γ_ = 1 s - 1 this simplifies to

lgK ow = lgτ ð4:41Þ

i.e. by reading the ordinate value for γ_ = 1 s - 1 we get Kow (refer Fig. 4.21).
This results in the dynamic viscosity

Fig. 4.21 Evaluation of an

Ostwald–de Waele flow curve

.
4.3 Viscous Behavior, Flow 179

Table 4.9 Examples of flow exponents and consistency coefficients for selected materials
Material ϑ/ ° C Kow/Pa
sn n Source
Tomato juice 12.8% (m/m) dm 32 2.0 0.43 [11]
Tomato sauce 25.0% (m/m) dm 32 12.9 0.4 [11]
Tomato paste 30.0% (m/m) dm 32 18.7 0.4 [11]
Condensed milk, 10% (m/m) fat 10 0.123 0.861 [12]
UF concentrate (protein concentrate from 5–50 20–16,000 1.13–0.17 [13]
cheese whey or whole milk) 1–99% (m/m)
Xanthan solution 1% (m/m) in water – 10 0.18 [13]
Xanthan solution 0.5% (m/m) in water – 3 0.24 [14]
Xanthan solution 0.25% (m/m) in water – 0.4 0.35 [14]
Xanthan solution 0.125% (m/m) in water – 0.14 0.5 [14]
Pure water 20 0.001 1 [15]

dτ K ow
γ_ n
ηs = = ð4:42Þ
d_γ γ_
so

ηs = K ow
γ_ n - 1 ð4:43Þ

With knowledge of Kow and n, the dynamic viscosity of a material can be

calculated for any shear rate. Table 4.9 shows some examples of Ostwald–de
Waele values reported in the literature.

Attention
In the case of non-Newtonian fluids, the viscosity depends on the load, i.e. the
shear rate. A recommended notation for the viscosity of non-Newtonian fluids
is the co-indication of the shear rate, e.g. in one of these forms:

η γ_ = 50 s - 1 = 24 Pa
s
η50 s - 1 = 24 Pa
s
η50 = 24 Pa
s

Definition
Apparent viscosity is a term used to distinguish between the constant viscosity
of Newtonian fluids (slope of a straight line), and some corresponding value
that can be obtained when the flow behavior curve is not straight, and there is
no constant slope. Apparent viscosity is taken to be the slope of a straight line
through the origin intersecting the curve at a specified constant shear rate. Such

(continued)
180 4 Rheological Properties

a value, when taken at a specified shear rate, is often referred to as “apparent

viscosity,” ηs. In other words, apparent viscosity ηs is the slope of the τ - γ_ -
graph at γ_ , i.e. ηs = d_
γ ðγ_ Þ.
dτ

4.3.13 Model Functions for Plastic Fluids

The models best suited to describe plastic flow behavior (those with a yield stress)
are the models after Bingham, Casson, Heinz, Herschel–Bulkley, Schulmann–
Haroske–Reher, Tscheuschner and Windhab (Table 4.10). In the use of models
like these, the model parameters are often given special names. For example, the
parameter τ0 in the Bingham model is given the name Bingham yield stress, and the
parameter η of this model is named the Bingham viscosity. By use of this naming
convention, we know to which model the parameter belongs. Otherwise, the
parameters per se would not give any indication as to what model they belonged.
In many of these models that describe flow behavior with a yield stress, the
viscosity term η sometimes is called plastic viscosity, and the parameter K often is

Table 4.10 Model functions for plastic fluids, examples

with τ0 = 0:
Bingham τ = τ0 þ η
γ_ τ0: Bingham yield stress in Pa τ = η
γ_
η: Bingham viscosity in Pa
s (plastic Newton
viscosity)
Casson 1 1 1
τ2 = τ0 2 þ ðη
γ_ Þ2 τ0: Casson yield stress in Pa τ = η
γ_
η: Casson viscosity in Pa
s Newton
Heinz 2 2 2
τ3 = τ0 3 þ ðη
γ_ Þ3 τ0: Heinz yield stress in Pa τ = η
γ_
η: Heinz viscosity in Pa
s Newton
Casson (general) 1 1 1
τn = τ0 n þ ðη
γ_ Þn τ0: Casson yield stress in Pa τ = η
γ_
Newton
η: Casson viscosity in Pa
s For
n = 1):
Bingham
n = 2):
Casson
n = 3):
Heinz
Herschel–Bulkley τ = τ0 þ K
γ_ n τ0: Yield stress in Pa τ = K
γ_ n
K: Plastic viscosity in Pa
s Ostwald–de
(consistency coefficient) Waele
n: Flow exponent
Schulmann– 1 1
τn = τ0 n þ ðK
γ_ Þm
1
τ0: Yield stress in Pa τ = K 0
γ_ k
Haroske–Reher K: Plastic viscosity in Pa
s Ostwald–de
(consistency coefficient) Waele
m, n: Constants
4.3 Viscous Behavior, Flow 181

Table 4.11 Herschel–Bulkley parameters for dairy products, examples

Material ϑ/ ° C τ0/Pa Kow/Pa
sn n Quelle
Buttermilk 0.4% (m/m) fat 10 1.299 1.607 0.410 [16]
Yogurt, stirred 1.5% (m/m) fat 10 4.34 11.05 0.285 [16]
Yogurt, stirred 13.5% (m/m) fat 10 11.03 6.91 0.408 [16]
Sour cream % (m/m) fat 10 25.17 21.53 0.532 [16]
Cream cheese 20% (m/m) fat in dry matter 10 67.47 14.65 0.688 [16]

called consistency coefficient, refer Table 4.10. For a very small yield stress τ0 the
approximation τ0 = 0 leads to simpler model function. So, for τ0 = 0 the models after
Bingham, Casson, and Heinz transform into the basic Newtonian model. In the same
way, the models after Herschel–Bulkley and Schulmann–Haroske–Reher, for the case
where τ0 = 0 would transform into the Ostwald–de Waele model, refer Table 4.10.
Casson´s model (the general Casson’s model) represents stronger dependency on
the flow behavior index n than other models like Bingham, Casson, or Heinz. In the
case of no yield stress, where τ0 = 0, Casson’s general model will transform into the
basic Newtonian model. The models after Bingham, Casson, and Heinz can be seen
as special cases of the Casson’s general model. In the same way, the Ostwald–de
Waele model obviously is a special case of the Herschel–Bulkley model.
The model function after Schulmann–Haroske–Reher is a power-law function
like Ostwald–de Waele, but with 4 parameters instead of two. Table 4.10 gives an
overview of these model functions.
If we compare Bingham, Casson, and Heinz model functions, we can see that
these can be understood as special cases of the general Casson model function. The
general Casson model function is thus very versatile and can even take the form of
the Newton model function with n = 1 as flow exponent and yield point τ0 = 0.
Likewise, the Ostwald–de Waele model function can be understood as a special case
of the Herschel–Bulkley model function. Table 4.11 gives numerical values of the
specific flow behavior constants for several viscous foods.
The model functions of Tscheuschner and Windhab, specially developed for
melted chocolate, use further parameters (ηstr1 and τ1) to increase the quality of
approximation between experimental and mathematical curves (Table 4.12).

Table 4.12 Models of Tscheuschner and Windhab

Tscheuschner τ = τ0 þ η1
γ_ þ ηStr1
γ_γ_n τ0: Yield stress in Pa
r
η1: Final viscosity in Pa
s
γ_ Str1 = 1 s - 1
γ_
γ_ r = γ_ Str1
ηStr1 = ηðγ_ Str1 Þ
Windhab τ = τ0 þ η1
γ_ þ ðτ1 - τ0 Þ
1- e - γ_γ_ γ_  = γ_ ðτ Þ
τ = τ0 þ ðτ1 - τ0 Þ
1- 1e
τ1: extrapolated yield stress in Pa
182 4 Rheological Properties

For molten chocolate the International Office of Cocoa, Chocolate and Sugar
Confectionery (IOCCC) recommended in 1973 Casson’s model for shear rates from
5 to 60 s-1. However, since the year 2000, the Windhab model has been
recommended for shear rates in the range γ_ = 2 . . . 5 s - 1 at ϑ = 40 ° C [17]:
γ_
τ = τ0 þ η1
γ_ þ ðτ1 - τ0 Þ
1- e - γ_  ð4:44Þ

γ_ shear rate in s-1

τ shear stress in Pa
η1 equilibrium viscosity in Pa∙s
γ_  constant in s-1
τ1 constant in Pa

According to Windhab, a shear-induced restructuring takes place in the shearing

of chocolate with increasing shear rate. This restructuring manifests itself in a
transition from an initial viscosity to a point of equilibrium viscosity η1, refer
Fig. 4.22. At this point with the shear stress τ = τðγ_  Þ the flow curve transits into
a linear flow function, the extrapolation of which to the ordinate provides the
so-called extrapolated yield point τ1 (refer Fig. 4.22).
With the shear rate γ_  at this point we get
γ_
τ = τ0 þ η1
γ_ þ ðτ1 - τ0 Þ
1- e - γ_  : ð4:45Þ

τ = τ0 þ η1
γ_ þ ðτ1 - τ0 Þ
1- e - 1 : ð4:46Þ

so

τ = τ0 þ η1
γ_ þ ðτ1 - τ0 Þ
0:632: ð4:47Þ

Above this shear stress or shear rate, chocolate melts show a linear flow curve,
i.e. they behave like a Bingham fluid from here on [18].

Fig. 4.22 Determining the

extrapolated yield stress τ1 for
the Windhab model
Kf

W
W1

W0
.
J
4.4 Temperature Dependency of Viscosity 183

Bottom Line
The characteristics of ideal materials, which we got to know at the beginning
of the chapter can be expressed very shortly on the basis of their material
properties:

Pascal material η=0

Newton material η = const.
St. Venant material τ0 > 0
Hook material E = const.
Euklid material G→1

4.4 Temperature Dependency of Viscosity

Since the viscosity of fluids is based on intermolecular interactions, it is a

temperature-dependent quantity. As the temperature rises, the thermal motion of
the molecules increases, which reduces the temporal mean of the bond strength to the
neighboring molecule. For this reason, the viscosity of liquids decreases with
increasing temperature and increases with decreasing temperature.
For low-viscosity substances, this behavior can be represented with an Arrhenius-
analogous approximation, the so-called Andrade equation

B Eo
η = A
eT = A
eRT ð4:48Þ

or

η E 1 1
ln = a - ð4:49Þ
ηr R T Tr

η dynamic viscosity in Pa∙s at T

ηr dynamic viscosity in Pa∙s at Tr
T temperature in K
Tr reference temperature in K
Ea activation energy in J∙mol-1
R universal gas constant in J
K-1
mol-1
A, B, C constants

If we draw viscosities η measured logarithmically above the reciprocal absolute

temperature T1 , a straight line with the slope b = ERa is obtained.
The quantity EA is called activation energy or shear activation energy in analogy
to the Arrhenius approach. In systems with unknown molecular weight, EA is not
determined, instead we work with slope b as a material parameter for the
184 4 Rheological Properties

temperature-related change in viscosity. If the viscosity ηr is known at any reference

temperature Tr, the viscosity η can be calculated at a different temperature T in
this way.

η 1 1
ln =b - : ð4:50Þ
ηr T Tr

η = ηr
eb
ðT - T r Þ :
1 1
ð4:51Þ

Since this is an approximation and the temperature dependence of the viscosity of

real substances deviates from the linear behavior of Arrhenius behavior, the smaller
the temperature difference between T and Tr the better the result.

Example
The dynamic viscosity of water at 33.4 °C is required. With the reference value

ηr = 1:0 mPas
ϑr = 20 ° C
at
T r = 293:15 K
and

ϑ = 33:4 ° C
T = 306:55 K
η
with the help of Andrade’s equation we get ln ηr =b
1
T - 1
Tr = 2036:8 K

-4 -1
1
306:55 K - 1
293:15 Kr = 2036:8 K
- 1:49
10 K = - 0:037

η
= eb
ðT - T r Þ = e - 0:3037 = 0:738
1 1

ηr

η = ηr
0:738 = 1:0 mPas
0:738 = 0:738 mPas

In Vogel’s equation [19] an additional parameter the so-called Vogel temperature

C is used.
B
η = A
eTþC ð4:52Þ

or

η 1 1
ln =B - ð4:53Þ
ηr T þ C Tr þ C
4.4 Temperature Dependency of Viscosity 185

η dynamic viscosity in Pa∙s at T

ηr dynamic viscosity in Pa∙s at Tr
T temperature in K
Tr reference temperature in K
Ea activation energy in J∙mol-1
R universal gas constant in J
K-1
mol-1
A, B, C constants

In addition, there are over 100 other, mostly empirically equations for calculating
viscosity as a function of temperature [20–22]. Grigull [23] uses the following
empirical equation (Vogel type) for water at ϑ in ° C
b
ηðϑÞ = a
eϑþc ð4:54Þ

with

a = 0:0318 mPa
s
b = 484:3726 ° C
c = 120:2202 ° C

Example
The dynamic viscosity of water at 33.4 °C is required. With
b
ηðϑÞ = a
eϑþc

and the above constants for water are

484:3726 ° C
ηð33:4 ° CÞ = 0:0318 mPa
s
e33:4 ° Cþ120:2202 ° C = 0:0318 mPa
s
e3:153
= 0:0318 mPa
s
23:407
ηð33:4 ° CÞ = 0:744 mPa
s

For many polymers and highly viscous substances, the Williams–Landel–Ferry

equation (WLF equation) is used instead of the Arrhenius relationship or Vogel’s
equation [24]:

η C
ðT - T r Þ
log =- 1 : ð4:55Þ
ηr C 2 þ ðT - T r Þ

The material constants C1, C2 depend on the selected reference temperature, often
the glass transition temperature Tg of a material is chosen as the reference
temperature.
186 4 Rheological Properties

η C0
T - T g
log = - 01 ð4:56Þ
ηg C2 þ T - T g

η dynamic viscosity in Pa∙s at T

ηr dynamic viscosity in Pa∙s at Tr
T temperature in K
Tr reference temperature in K
Tg glass transition temperature in K
C 1 , C 01 constants
C 2 , C 02 material constants in K

If a fluid is cooled, its viscosity increases. When the glass transition temperature
Tg is reached, the material has such a high viscosity that it behaves as a solid. This
non-crystalline state of a solid is called the glassy state. If the temperature is
increased again, a sharp decline in viscosity takes place at the glass transition
temperature: the glass softens into a viscous flowing material.
Glass transitions are examined with the help of thermal analysis. Therefore, in
Chap. 8 on thermal properties we will hear again about the glass transition
temperature.
Since numerous physical, chemical, and biological transformations in a food are
related to molecular mobility, many temperature correlated phenomena can be
described more precisely with WLF kinetics than with activation kinetics according
to Arrhenius. These include the viscosity of carbohydrate melts, crystallization rates,
browning reactions, oxidations [25]. Even the stability of food in the sense of
temperature-dependent shelf life can be predicted with the WLF equation [26]. On
the other hand, the WLF concept has limits, e.g. it must not be overlooked that the
WLF material constants are not “universal,” as in the field of synthetic polymers, but
depend on the composition of the food, e.g. on water content and water activity
[27, 28].

4.5 Viscosity of Solutions

If macromolecular substances are dissolved in a solvent, e.g. water, the viscosity of

the solution increases with increasing concentration of the solution. The ratio of
viscosity of the solution and viscosity of the pure solvent is called relative viscosity
or viscosity ratio.

η
ηrel = : ð4:57Þ
ηsolvent

It is a ratio without a unit. If the viscosity of the solution is reduced by the value of
the pure solvent, we get
4.6 Viscoelasticity 187

η - ηsolvent
: ð4:58Þ
ηsolvent

or

η
-1 ð4:59Þ
ηsolvent

which is called viscosity relative increment. It is a measure for the viscosity effect
of the solute. The former term “specific viscosity“for this size should be avoided. If
the viscosity relative increment is applied above the concentration of the dissolved
material and the function extrapolated to the hypothetical value c = 0, the limit value
for an infinite dilution, the so-called Staudinger index of the dissolved material, is
obtained.

1 η
Staudinger index = lim -1 : ð4:60Þ
c→0 c ηsolvent

The former term “intrinsic viscosity“for the Staudinger index shall be avoided
because it is not a viscosity in the unit Pa
s. The Staudinger index is related to the
hydrodynamic volume of the dissolved substance. Therefore, the Staudinger index
can be used to give information about the molar mass of the dissolved substance, for
this purpose the Kuhn–Mark–Houwink relationship is used, cf. [29]. If the solute is
not a pure substance, average molecular weights of the dissolved substance are
obtained in this way.

Definition
Staudinger Index: Extrapolated viscosity of a solution at infinite dilution.
Although this is unusual for an index, the Staudinger index has a unit, namely
the unit of an inverse concentration, e.g. (g/100 cm3 solvent)-1.

4.6 Viscoelasticity

To understand viscoelastic materials, we first consider a soft, elastic solid,

e.g. camembert, marshmallows, yeast dough. The material reacts to a shear stress
with deformation. Typically the deformation does not occur instantly, but gradually.
Likewise, when the stress is removed, the material gradually returns to its initial
form. Figure 4.23 shows such behavior schematically. Here, a shear stress τ is
applied abruptly, kept constant for some time and then suddenly set back to zero
(rectangle shear signal). The deformation γ is the response signal of the body.
An ideal elastic material would react to the shear stress with an instantaneous
deformation and return to the initial state just as instantaneously when the stress is
removed. A material that as in Fig. 4.23 follows a shear stress with a delayed
188 4 Rheological Properties

Fig. 4.23 Viscoelasticity: A

material loaded with a
rectangular shear stress signal
(τ - t-curve) shows a time-
delayed deformation γ

Fig. 4.24 Maxwell element:

shear stress curve in response
to a rectangular deformation
signal

deformation, but returns to its initial state, is called viscoelastic [30]. At the begin-
ning of the chapter, we got to know simple mechanical models with which the
rheological behavior of materials can be mapped. For viscoelastic behavior, the
Maxwell element and the Kelvin element are suitable in the simplest case each
consisting of a spring and a damper (Fig. 4.9). The spring models the elastic behavior
of the body, the damper the viscous flow resistance, which slows down the elastic
deformation [9]. If a Maxwell element is loaded with a rectangular deformation
signal, it shows a typical, time-delayed course of the shear stress in response (see
Fig. 4.24), which can be calculated as follows:
The applied stress is the same for the Hook element and the Newton element, so

τ = τHooke = τNewton

the deformation of the Maxwell element we get by addition

4.6 Viscoelasticity 189

γ = γ Hooke þ γ Newton
dγ d
γ_ = = ðγ Hooke þ γ Newton Þ
dt dt
with G = const:
d 1 1 dτ τ
γ_ =
τ þ γ Newton =
þ
dt G G dt η
or
η dτ
τ þ
= η
γ_
G dt
η
with λrel =
G
dτ
τ þ λrel

dt
with γ = const:or γ_ = 0 we get

dτ
τ þ λrel
=0
dt
τ 1
=- dt
dτ λrel
τ t
τ 1
=- dt
dτ λrel
τ0 0
τ t
ln = -
τ0 λrel
-λt
τ = τ0
e rel

This means that the shear stress drops exponentially over time, the stress relaxa-
tion is a first-order decay function. From the relaxation curve we can determine the
relaxation time λrel of the material. It is the period of time after which the stress has
fallen to 1/e of the initial value. Materials that relax slowly have a greater relaxation
time. Materials that relax immediately have a shorter relaxation time.
Assuming γ = const., that means that the shear modulus of the viscoelastic
material changes in time, we can therefore write:

τ0 -λt -λt
G=
e rel = G0
e rel :
γ const

This formulation shows that the shear modulus of a Maxwell material is not
constant (as with a Hookean material), but changes over time. Note the indices: here
τ0 is the maximum stress, it must not be confused with the yield point of the material.
In physics, the symbol τ for the relaxation time is often used. To avoid confusion
with the shear stress τ, in rheology the symbol λrel for the relaxation time is used.
Viscoelastic materials that look solid at first glance can flow slowly, this is called
creep. Let us consider a Kelvin element (Fig. 4.25) that is loaded with a constant
strain. The applied strain is the same for the Hook element and the Newton element
190 4 Rheological Properties

Fig. 4.25 Kelvin element:

reaction to a rectangle stress
signal

γ = γ Hooke = γ Newton

the stress of results additively:

τ = τHooke þ τNewton

i.e.,

τ = G
γ þ η
γ_
1 dτ d_γ

= γ_ þ λret

G dt dt
with γ = const:, or γ_ = 0 is

1 dτ

= γ_
G dt
integration results to
0 t
1

dτ = γ_
dt
G
τ0 0
1 γ
- τ0 = - t
G e ret - 1
λ

τ0
γ =
1 - e - λret
t

G
The deformation of the Kelvin element thus proceeds in the form of a (1 - e-time)-
function asymptotic to a maximum value. Figure 4.25 illustrates the behavior.
This behavior is often described with a reciprocal quantity to the shear modulus,
the so-called compliance J.
With
4.6 Viscoelasticity 191

1
J= ð4:61Þ
G
we have

1 - t
J=
1- e λred : ð4:62Þ
G
Here λred is the retardation time. This is the time it takes to reach (1 - e-1)-fold of
the final value, i.e. 63.2% of the final value.

Attention
Relaxation time λrel and retardation time λred are not the same.

To improve the approximation between experimental data and model, the Kelvin
element can be supplemented by a Maxwell element in series (see Fig. 4.26) which is
called a Burgers model.
The Burgers model can be further optimized by using several Kelvin elements in
parallel. Finally, a fracture element and a St. Venant element (for the yield point) can
be added to model the behavior of real materials.

Definition
The differential equations for the Kelvin model were formulated by
O.E. Meyer in 1874 and extended by W. Vogt in 1892. W. Thomson (Lord
Kelvin) described the behavior of viscoelastic bodies as early as 1865. There-
fore the Kelvin/Vogt model (German standard DIN 1342) is the same as the
Kelvin model [10].

In addition, there are viscoelastic materials that do not completely return to their
initial state after deformation, but show a permanent deformation. The behavior of
such plasto-viscoelastic materials can be modeled by adding a St. Venant element for

Fig. 4.26 Burgers model for

viscoelastic materials
192 4 Rheological Properties

Table 4.13 Systematics of Rheology of Solids and Fluids

Material state Solid Liquid Gaseous
Designation Solid Fluid
Ideal body Euklid, Hooke Newton, Pascal
Material properties Elasticity, plasticity Viscosity
Real materials Non-ideal Newtonian, non-ideal elastic, viscoelastic, elastoviscous,
viscoplastoelastic, etc.

Fig. 4.27 Real behavior as mixed cases of ideal behavior

the yield point to the Maxwell model, the Kelvin model, or the Burger model. In the
chapter about texture, we will come back to these mechanical models (Table 4.13).
Rheology as a theory of flow generally describes the deformation behavior of
bodies under the influence of forces or stresses. In order to be able to describe the
often complicated deformation behavior of real bodies analytically, one tries to
present the real behavior as mixed cases of ideal bodies [9]. In Table 4.3, we have
seen the basic bodies for ideal elastic behavior, ideal viscous behavior, and ideal
plastic behavior.
For illustration, Fig. 4.27 shows the ideal cases as vertices of a triangle. The
mixed cases, i.e. real materials, lie within this triangle, depending on the degree of
their elasticity, viscosity, and plasticity.
The experimental study of the deformation behavior of viscoelastic foods is dealt
in the following Sect. 4.7.
4.7 Rheological Measurement 193

Fig. 4.28 Rheological systematics of ideal and non-ideal materials

Figure 4.28 shows a different view of the rheological classification of materials.

Here, mixed cases are also distinguished, which lie between “ideal-elastic” and
“ideal-viscous.” With the plasticity on its own axis, we get from this tabular
representation to the triangle representation in Fig. 4.27.

4.7 Rheological Measurement

There is any number of different types of laboratory instrumentation by which to

measure rheological properties of materials. Perhaps the most commonly used type
of instruments for studying the rheology of liquids are rotational viscometers also
called rheometers [31]. They will be described first in this section. Other types of
instruments, based on different principles like capillary types and falling sphere type,
will be described later.

4.7.1 Rotational Rheometers

All rotational devices consist of a rotating body (rotor) and a stationary body (stator).
The liquid sample is always placed between the rotor and stator, which causes the
sample to experience shear during operation of the instrument. Within the broad
category of rotational instruments, the coaxial cylinder system is perhaps most
widely used. It consists of a smaller cylinder (bob) within a slightly larger cylinder
(cup), such that an annular gap is left between the two cylinders, into which the
liquid sample is placed. This system is called coaxial cylinder measuring
system [32].
194 4 Rheological Properties

Fig. 4.29 Coaxial cylinder

systems after Searle (left) and M
Couette (right). The sample S S S
is located in the annular gap

Fig. 4.30 Rotational

rheometer, Searle-type. 1
1 motor, 2 torque meter,
3 temperature-controlled
cylindrical beaker, 4 rotating 2
cylinder (bob), 5 annular
space for liquid sample

4
5

Fig. 4.31 Mooney–Ewart

system: Rotating bob with
conical frontal surface in a
double-walled, temperature-
controlled cylindrical cup

When the cup is fixed and the bob is rotated, this is called a Searle-type coaxial
cylinder system. When the cup is rotated around a fixed bob, this is called a Couette-
type coaxial cylinder system, refer Fig. 4.29. The Searle-type is technically easy to
implement and widely used, see Fig. 4.30.
If we want to achieve that the same shear takes place on the lower frontal surface
of the rotating cylinder as on the cylinder jacket, we can make the lower cylindrical
frontal surface conical, see Fig. 4.31. If we then omit the cylinder jacket surface, we
4.7 Rheological Measurement 195

get a cone–plate measuring system, see Fig. 4.35. Here, the flat cone rotates over a
stationary plate. If a circular plate is used instead of the rotating cone, this is referred
to as the plate–plate measuring system.
Rotational rheometers allow to vary the speed of the rotor, and by this the shear
rate applied to the sample. By electronic measurement of the torque needed to
achieve a given angular velocity, the shear stress can be measured. From the
rotational speed and the geometry, the shear rate can be calculated. From the torque
and the geometry the shear stress can also be calculated. A plot of shear stress over
shear rate gives the flow curve, as described previously (Fig. 4.12).
Modern instruments allow a choice between controlled shear rate mode (CSR)
and controlled shear stress mode (CSS). In the CSR mode, the shear rate is
programmed and presented to the sample, and the resulting shear stress is measured.
In the CSS mode, the shear stress is programmed and presented to the sample, and
the resulting shear rate is measured.
Many instruments also allow a means of operation to perform measurements in an
oscillating mode. Under an oscillating mode of operation, the rotor reverses direction
periodically at a programmed frequency, meaning it rotates back and forth, first
clockwise, then counter-clockwise, performing a swinging movement with the
sample. This mode of operation is recommended for materials with viscous, as
well as elastic properties, see Sect. 4.7.2.

Cylindric System Geometry

When using coaxial cylindrical rheometers, it is necessary to know the general
relationships between the parameters of the rotating cylinder and the rheological
quantities of shear stress and shear rate. First let us calculate the shear stress when a
cylinder is rotated as shown like in Fig. 4.32. With the measured torque M

M=r
F

and the area A,

A=2
π
r
h

Fig. 4.32 Left: torque on the M

rotary cylinder, right: jacket 2π r
surface of the cylinder F
r

h
196 4 Rheological Properties

the following can be computed:

F M
τ= = ð4:63Þ
A 2
π
r2
h

γ_ shear rate in s‐1

τ shear stress in Pa
A area of cylinder surface in m2
F tangential force in N
M torque in Nm
r radius of cylinder in m
h height of cylinder in m
v tangential velocity in m∙s-1
ω angular velocity in s-1
η viscosity in Pa∙s

The shear rate within the annular sample space is:

dv dω
γ_ = - =-
r ð4:64Þ
dr dr
dv d dω
= ðω
r Þ =
r ð4:65Þ
dr dr dr
That means, at a constant speed of rotation ω, the shear rate γ_ depends on the
distance r.
From Eq. (4.64) we know

M
τ= ð4:66Þ
2
π
r2
h
so

1=
2
M 1=
r=
τ- 2
ð4:67Þ
2
π
h

and

1=
2
dr M 1 3=
=
-
τ- 2 ð4:68Þ
dτ 2
π
h 2

with (4.66) this results to

4.7 Rheological Measurement 197

1=
dr τ
2
π
r2
h 2
1 3= r
=
-
τ- 2 = - ð4:69Þ
dτ 2
π
h 2 2τ

i.e.,

dr dτ
=- ð4:70Þ
r 2τ
with (4.64) written as

dω
γ_ = f ðτÞ = -
r
dr
follows for the ω(r) function:

dr 1 dτ
dω = -
f ðτ Þ = f ðτ Þ
ð4:71Þ
r 2 τ

Integration over the thickness of the annular shear space in Fig. 4.33 gives:
ωa = 0 τa
1 dτ
dω = f ðτ Þ
ð4:72Þ
2 τ
ωi = Ω τi

this is

τa
1 dτ
Ω= - f ðτ Þ
ð4:73Þ
2 τ
τi

Fig. 4.33 Radii, shear stress, ω

and angular velocity of the M
rotating cylinder

Ri

Ra

ωaτa ωτ
i i
198 4 Rheological Properties

τ shear stress in Pa
R radius of cylinder in m
ω angular velocity in s-1
Ω angular velocity of inner cylinder in s-1
Index i inner cylinder (bob)
Index a outer cylinder (cup)

This is the general function between angular velocity Ω of the inner cylinder
(bob) and the resulting shear rate in a Searle-type rheometer. The equation can be
solved only by substituting the behavior of the fluid f(τ). To do this we first choose a
Newtonian fluid. After that it will be shown how the equation also works with an
Ostwald–de Waele fluid.

For a Newtonian Fluid

τ
γ_ = f ðτÞ = ð4:74Þ
η
So Eq. (4.73) becomes:

τa τa τa
1 dτ 1 τ dτ 1
Ω= - f ðτÞ = -
=- dτ ð4:75Þ
2 τ 2 η τ 2η
τi τi τi

that means

1
Ω= ðτ - τa Þ ð4:76Þ
2η i
with Eq. (4.66) we get a term called Margules’ equation

1 1
Ω= ð4:77Þ
4π
h
η R2i - R2a

Margules´ equation provides the relationship between angular velocity Ω of the

rotor and the resulting torque M (angular moment) when a Newtonian fluid with
viscosity η is in the annular space between Ri and Ra, refer Fig. 4.33.

For an Ostwald-de-Waele Fluid

Here we have instead of Eq. (4.74)
1
τ n
γ_ = f ðτÞ = ð4:78Þ
K OW

so
4.7 Rheological Measurement 199

τa τa 1
1 dτ 1 τ n
dτ
Ω= - f ðτ Þ = - ð4:79Þ
2 τ 2 K OW τ
τi τi

i.e.,

n 1 1
Ω= 1 τni - τna ð4:80Þ
2
K OW n

with Eq. (4.66) we get

1 1
η M n
M n
Ω= - ð4:81Þ
2
π
h
R2i 2
π
h
R2a
1
2
K OW n

or
1
2
η M n
Ri
Ω=
1- ð4:82Þ
2
K OW n 2
π
h
Ri
1 2 Ra

This is a general version of the relationship between angular velocity Ω of the

rotating body and torque M in the concentric cylinder system. It applies to Newto-
nian and non-Newtonian fluids. For n = 1 it converts into Margules equation.
Of course, we can also use more complicated model laws instead of the Ostwald–
de Waele law (refer Tables 4.8 and 4.10) if this is required by the material behavior
(the flow curve).
If the annular gap of the concentric cylinder system is very small, i.e.
(Ra - Ri) << Ri, then the system can be approximately treated as if the wall
surfaces were flat planes (not curved). Considerations as with the two-plate model
then apply (refer Sect. 4.3.1). This is called simple shear approximation.
With the two-plate model

vx
γ_ x = ð4:34Þ
y
then is

Ω
Ri
γ_ i = ð4:83Þ
Ra - Ri
with abbreviation δ for the radius ratio of outer and inner cylinder

Ra
δ= ð4:84Þ
Ri
this is
200 4 Rheological Properties

Ω
γ_ i = ð4:85Þ
δ-1
With the average shear stress

1
τ= ðτ - τ i Þ ð4:86Þ
2 a
and Eq. (4.64) we get

1 M M
τ= - ð4:87Þ
2 2
π
h
R2a 2
π
h
R2i

so

M 1 þ δ2
τ= ð4:88Þ
4
π
h R2a

This is the relationship between the measured torque M and the shear stress τ in a
rotational viscometer with a narrow shear gap. According to DIN 53019 [33] the
radius ratio δ should be δ ≤ 1.2, preferably δ = 1.0847.
The calculations presented apply to laminar shear flows in an annular gap. The
error caused by the simple shear approximation can be calculated, as well as
correction variables for interference effects such as wall sliding effects, vortices,
frontal surface influences, etc. [9].

Cone–Plate Systems
Cone–plate measuring systems (Fig. 4.35) consist of a flat cone that rotates over a
resting flat plate.
The shear rate is

dv
γ_ = ð4:35Þ
dh
with

ω=v
r ð4:89Þ

and
tan α = dh
dr (refer Fig. 4.34).
it is

ω
dr ω
γ_ = = ð4:90Þ
tan α
dr tan α
and with Ω as already used symbol for the angular velocity of the rotor
4.7 Rheological Measurement 201

Fig. 4.34 Cone–plate Z

measuring system, schematic M
r

h
D

Ω
γ_ = ð4:91Þ
tan α
The torque on the cone is

M=r
F=r
τ
A

that means

R
R
1 3 2
M= r
τ
dA = r
τ
2
π
r
dr = r
2
π
τ =
π
R3
τ ð4:92Þ
3 0 3
0

Thus, the shear stress is:

3 M
τ=
ð4:93Þ
2 π
R3
In contrast to the concentric cylinder system, the shear stress of this geometry is
not dependent on the location r. This is the great advantage of the cone–plate system:
All fluid elements in the gap are subjected to the same stress.
If we describe the material to be investigated with the Ostwald–de Waele law
τ = K OW
γ_ n ð4:39Þ

by using

Ω 3 M
γ_ = and τ =
:
tan α 2 π
R3
we get

3 M Ω n

= K OW ð4:94Þ
2 π
R3 tan α
The application of the torque M over the angular velocity Ω of the rotating cone in
double-logarithmic form thus provides the flow exponent n.
202 4 Rheological Properties

In rheological measurement technology, very flat cones are (α = 2 ° . . .5°,

i.e. α = 0.035. . .0.087 rad) used. With the approximation tanα ≈ α, which applies
to small angles, the shear rate is then at all locations of the cone gap γ_ = Ωα .

Example
A cone (R = 5 cm, α = 2°) rotates at 60 RPM. What is the shear rate that a
sample of mayonnaise experiences in the cone gap?

2π
60 2π
60
Ω = 2π
f = 2π
60 min - 1 = = = 2π s - 1 = 6:28 s - 1
min 60 s
2° π
α = 2π
=
360 ° 90
Ω 2π s - 1
γ_ = = π = 2
90 s - 1
α
90
γ_ = 180 s - 1

Plate–Plate System
In the plate–plate system, a round plate rotates over a stationary flat plate (Fig. 4.35).
The sample is located in the space between the rotating plate and the dormant plate.
The shear rate here is:

dv ω
d_γ = =
dr ð4:95Þ
h h
integrated

ω
γ_ =
r ð4:96Þ
h

Fig. 4.35 Plate–plate Z

measuring system M

h
4.7 Rheological Measurement 203

with the angular velocity Ω of the rotor

Ω
γ_ =
r ð4:97Þ
h
at the outer edge of the rotating plate

Ω
R
γ_ = ð4:98Þ
h
The torque is

R R

M= r
τ
dA = r
τ
2
π
r
dr = 2
π
r 2
τ
dr ð4:99Þ
0 0

with

Ω
γ_ =
r ð4:100Þ
h
and

h
r=
γ_ ð4:101Þ
Ω
γ_
h 2
using r 2 = Ω and dr = h
γ.
Ω d_
we get

γ_ R
γ_ 2
h3
M= 2
π

τ
d_γ ð4:102Þ
Ω3
γ_ = 0

Ω 3
expanding both sides of the equation by R13 = h
γ_ R .
results to

γ_ R
M 1
= γ_ 2
τ
d_γ ð4:103Þ
2
π
R 3
ðγ_ R Þ3
0

Replacing equation τ = f ðγ_ Þ we get

γ_ R
M

γ_ 3 = γ_ 2
f ðγ_ Þ
d_γ ð4:104Þ
2
π
R3 R
0
204 4 Rheological Properties

Differentiating the left side of the equation returns:

M
d M d 2
π
R3 M

γ_ 3R = γ_ R 3 þ 3_γ R 2 : ð4:105Þ
d_γ R 2
π
R 3 d_γ R 2
π
R3

Differentiating the right side of the equation provides (Leibnitz’s theorem):

γ_ R
d
γ_ 2
f ðγ_ Þ
d_γ = ðγ_ R Þ2
f ðγ_ R Þ: ð4:106Þ
d_γ R
0

Equating both sides results

M
d 2
π
R3 M
γ_ R 3 þ 3_γ R 2 = ðγ_ R Þ2
f ðγ_ R Þ: ð4:107Þ
d_γ R 2
π
R3

Because of τ = f ðγ_ R Þ for the shear stress on the outer edge of the plate we get

M
d 2
π
R3 3M
τ = γ_ R þ ð4:108Þ
d_γ R 2
π
R3
or

M d ln M
τ= 3þ ð4:109Þ
2
π
R 3 d ln γ_ R

In case of Newtonian behavior

Ω
r
τ = η
γ_ = η
ð4:110Þ
h
so

R
Ω
M= 2π
r 3
η

dr ð4:111Þ
h
0

Integration results to

R
2π
η
Ω 2π
η
Ω 1 4 R
π
η
Ω 4
M= r 3
dr =
r =
R ð4:112Þ
h h 4 0 2h
0
4.7 Rheological Measurement 205

with

Ω
γ_ =
R
h
the expression is simplified to

π
M=
η
γ_
R3 ð4:113Þ
2
That is

π 3
M=
R
τ ð4:114Þ
2
or

2
M
τ= ð4:115Þ
π
R3
with Newton’s law τ = η
γ_ thus

2
M Ω
R
=η
ð4:116Þ
π
R3 h
and the viscosity we are looking for

2
M
h
η= ð4:117Þ
π
R4
Ω
for Ostwald–de Waele fluids results analogously

M ð 3 þ nÞ Ω
R n
= K OW
: ð4:118Þ
2
π
R 3 h
If we compare cone–plate and plate–plate systems, cone–plate systems have the
advantage of a location-independent load on the fluid. The shear rate and the shear
stress are the same everywhere in the gap. With the plate–plate system, on the other
hand, the shear load of the fluid is not the same everywhere, it is greater at the outer
edge (r = R) than further toward the center. In comparative calculations, the shear
rate and shear stress at the outer edge of the rotating plate γ_ R are used or alternatively
an average shear rate is used. The advantage of the plate–plate system is the
adjustable and often larger gap width h. If disperse materials (suspensions,
emulsions, foams) are to be measured, the plate–plate system is advantageous so
that the measuring gap remains larger than the particle size. This can also be
achieved with cone–plate systems by making the cone a truncated cone. The
resulting gap between the truncated cone and the plate also enables the measurement
of disperse materials with a particle size that is not too large. As a rule of thumb, the
206 4 Rheological Properties

particle size should be ≤ 15 gap . Example: The truncated cone lacks 50 μm to a

complete cone (gap width = 50 μm), then disperse systems with particle sizes x ≤ 10
μm can be measured.

Example
In the cone–plate viscometer, R = 25 mm, α = 1°, a Newtonian fluid is
examined. At 16.7 RPM, a torque of M = 10 mNm was measured. What is
the viscosity of the sample?
Calculation of the angular velocity of the rotor:

16:7
ω=2
π
f =2
π
= 1:748 s - 1
60 s
Calculation of the shear rate:

ω ω 2π
16:7 16:7
360

γ_ = ≈ = γ_ = 2π
160° s = = 6
16:7s - 1 = 100 s - 1
tan α α 360 °
60 s

Calculation of shear stress:

3
M 3
10
10 - 3 Nm 3
10 - 3 30
103
τ= = = Pa = Pa
2
π
R3 2
π
ð0:025 mÞ3 2
π
2:53
10 - 6 2
π
2:53
= 305:6 Pa

Calculation of viscosity:

τ 305:6 Pa
η= = = 3:1 Pa
s:
γ_ 100 s‐1

Example
A gel should be deformed in the plate–plate system (H = 1 mm,
R = 37.5 mm). How big is the deformation γ of the sample if the upper
plate is deflected by 0.15°?
Solution:
Sketch: Plate seen from above, deflection α.

(continued)
4.7 Rheological Measurement 207

Sketch: Plate from the side: Deflection α results in deformation of γ.

Calculation of deformation γ:
Because of

ω
R φ
R
γ_ = ist γ =
H H
it is

φ
R 0:15
360
37:5 mm

2π
γ= = = 0:1≙5:7 ° :
H 1 mm

Example
In the plate–plate viscometer (R = 25 mm, H = 1 mm) a Newtonian fluid is
examined. At 30 RPM, a torque of M = 10 mNm was measured. What is the
viscosity of the sample?

(continued)
208 4 Rheological Properties

Calculation of shear stress:

2
M 2
10
10 - 3 Nm
τ= = = 407 Pa:
π
R3
π
ð0:025 mÞ3

Calculation of the shear rate:

R R 30 25 mm
γ_ = ω
= 2π
f
= 2π

= π
25 s - 1 = 78:5 s - 1
H H 60 s 1 mm
Calculation of viscosity:

τ 407 Pa
η= = = 5:2 Pa
s
γ_ 78:5 s‐1

4.7.2 Oscillation Test

Modern rotation rheometers can also be operated in oscillation mode instead of in

rotation mode. Cone–plate or plate–plate systems are often used for this purpose. By
means of oscillating shear stress, solid and semisolid materials can be examined
non-destructively and elastic and viscous material properties can be recorded simul-
taneously. In order for the examination to be actually non-destructive, the oscillation
must be carried out with small shear amplitudes. The maximum permissible ampli-
tude, i.e. maximum deformation of the sample, must be determined in a
preliminary test.
In the case of a sinusoidal oscillating deformation, the deformation is γ

γ = γ 0 sin ωt

and the shear rate

γ_ = γ 0
ω
cos ωt = γ_ 0 cos ωt

synonymous with

γ_ = γ_ 0 sinðωt þ 90 ° Þ

With a sine-function for deformation, the shear rate follows as a cos-function.

This is equivalent to a phase shift of 90° between the two functions.
If a sinusoidal, oscillating deformation of the sample is now carried out, an ideally
elastic material will show a shear stress proportional to the deformation γ. An ideally
viscous material, on the other hand, will show a shear stress proportional to the shear
rate γ_ .
4.7 Rheological Measurement 209

From Sect. 4.3, we know for ideal elastic behavior

τ=G
γ ð4:17Þ

and for ideal viscous behavior

τ = η
γ_ ð4:28Þ

Therefore, the oscillating measured shear stress will be in phase with the defor-
mation for an elastic material. While, for a viscous material, a phase shift of 90°
between deformation and shear stress occurs. Figure 4.36 schematically shows the
phase shifts of these cases.
For this reason, determination of the phase shift δ under oscillating stress is a
simple way of characterizing viscoelastic materials or their elastic and viscous
components (refer Fig. 4.37).
By using the complex shear modulus, it is possible to characterize the elastic and
viscous parts of a substance solely on the basis of the phase shift δ. If you write the
shear modulus as a complex quantity

G = G0 þ G00 ð4:119Þ

ideal elastic visco-elastic ideal viscous

G 0q G G 90q

shear stress

deformation

Fig. 4.36 Phase shift δ between applied (shear) stress and resulting deformation under oscillating
load

Fig. 4.37 Phase shift

between deformation (I) and
G
shear stress (II)
Amplitude

I II

t
210 4 Rheological Properties

then G′ is the storage modulus (this is the elastic component) and G″ for the loss
modulus l (this is the viscous component) of the complex shear modulus. The
appendix contains the most important rules for working with complex quantities.
If we expose a material to a sinusoidal oscillating deformation γ,
i.e. γ = γ 0
sin ωt, we obtain an oscillating shear stress—measurable in an
oscillation rheometer—which has a phase shift δ to the excitatory sine oscillation.

τ = τ0
sinðωt þ δÞ: ð4:120Þ

In the simplest case δ = 0, the excitation τ and response signal γ are in phase.
However, we first want to assume that δ has a value not known to us, between 0° and
90°, i.e. δ = 0⋯ π2. The shear stress results from

τ = K elastic
γ þ K viscous
γ_ ð4:121Þ

we write
τ = G
γ = η
γ_ ð4:122Þ

or

G00
τ = G0
γ þ
γ_ ð4:123Þ
ω
00
The part G′
γ is in phase with the excitation, the part Gω
γ_ has a phase shift of π2 to
the exciting oscillation. It is

τ
G0 = cos δ ð4:124Þ
γ0

and

τ
G00 = sin δ ð4:125Þ
γ0

The absolute value of the complex shear modulus is

τ
G = G0 þ G00
2 2
= ð4:126Þ
γ0

with the phase shift δ

G00
tan δ = ð4:127Þ
G0
4.7 Rheological Measurement 211

With the dynamic viscosity from Newton’s law it can be formulated:

G
η = ð4:128Þ
ω

G00
η0 = ð4:129Þ
ω

G0
η00 = ð4:130Þ
ω

η = η0 2 þ η00 2 ð4:131Þ

With the Ostwald–de Waele constant of a non-Newtonian fluid, the complex

viscosity is:

G
K = ð4:132Þ
ω

G00
K0 = ð4:133Þ
ω
G0
K 00 = ð4:134Þ
ω

η = K 0 2 þ K 00 2 ð4:135Þ

For viscoelastic materials, often the inverse size of the shear modulus, the
so-called compliance J is used.

1
J= ð4:136Þ
G
1
J = ð4:137Þ
G
J 
G = 1 ð4:138Þ

G0
J0 = ð4:139Þ
G þ G00
02 2

G00
J 00 = ð4:140Þ
G0 þ G00
2 2
212 4 Rheological Properties

Symbol Name SI unit

γ Shear angle
γ_ Shear rate s-1
γ0 Shear angle amplitude
ω Angular frequency s-1
f Frequency s-1
t Time s
τ Shear stress Pa
τ0 Shear stress amplitude Pa
Kelastic, Kviscous Constants
η Viscosity Pa
s
η Complex viscosity Pa
s
η′ Real part of the complex viscosity Pa
s
η″ Imaginary part of complex viscosity Pa
s
G Shear modulus Pa
G Complex shear modulus Pa
G′ Storage shear modulus Pa
G″ Loss shear modulus Pa
δ Phase shift
J Compliance Pa-1
J Complex compliance Pa-1
J′ Storage compliance Pa-1
J″ Loss compliance Pa-1

For a body with purely elastic properties (Hooke body, refer Sect. 4.2), the loss
module is G″ = 0, i.e. the imaginary part of the complex shear modulus is zero. In
this case, a distinction between real part and imaginary part of the shear modulus is
not necessary, because of G = G′ + 0 = G we have the simple, i.e. non-complex,
elastic case. Experimentally, this case can be recognized by the fact that δ = 0. If a
body have purely viscous properties (Newton body, refer Sect. 4.2), then is
00
tan δ = GG0 = 1, that means δ = 90°. In this case, the storage modulus G′ is zero
and the loss modulus G″ rules the complex shear module: G = 0 + G″.
Many food and biological materials have both elastic and viscous properties.
They thus show a behavior between Hook and Newtonian bodies that are experi-
mentally recognizable by a phase shift angle of 0 < δ < 90°.

Definition
Complex physical quantities consist of a complex number and a physical unit.
Complex numbers consist of a real part and an imaginary part. The real part is
p
a real number, the imaginary part is a multiple of i. i = - 1. The advantage
of complex quantities is that we can specify two properties with one size.
4.7 Rheological Measurement 213

Fig. 4.38 Shear modulus and J' tan G G'

compliance of a viscoelastic
material, schematic J '' G ''

G ',G '' J ', J ''

1 10 100 1000 10000

f / Hz

Bottom Line
In the case of the shear modulus, these are:

Complex shear modulus G = storage modulus G0 þ loss modulus G00 :

In the case of viscosity:

Complex viscosity η = real viscosity η0

þ imaginary viscosity ðelasticityÞ η00 :

If we examine a material in a shear oscillation test by varying the oscillation

frequency f (frequency sweep), we obtain elastic and viscous sizes as frequency-
dependent quantities. Figure 4.38 shows as an example of how the shear modulus,
compliance, and the phase shift δ can be related. Here, the material properties were
recorded in an oscillation rheometer with increasing frequency f.
In summary, the oscillation test allows the simultaneous determination of viscous
and elastic properties of a material. When using complex quantities, the rheological
properties of the materials can be characterized very easily by the phase shift δ. An
overview of applications in the food sector can be found in [34].
In addition to oscillating rotational rheometers, there are so-called DMA devices
(dynamic mechanical analysis), which allow the oscillating examination of samples
in the uniaxial direction.

4.7.3 Further Measurement Systems

In this section, measuring systems are addressed which play a role in addition to the
rotational measuring systems. Capillary and ball viscometers are first presented as
absolute measuring systems.
214 4 Rheological Properties

Capillary Viscosimeter
Capillary viscometers consist of a capillary with a defined diameter, through which
the fluid to be examined flows. The measured quantity is the throughput time for a
certain volume of liquid. On the basis of the Hagen–Poiseuille law for ideally
viscous fluids, the volume flow rate and the shear rate can be calculated in laminar
flow by measuring throughput time. The associated shear stress results from the
pressure difference between the ends of the capillary. For a flow in a cylindrical
capillary the force resulting from a pressure is

F P = Δp
πr 2 ð4:141Þ

the frictional force is

dv
FR = η
A
ð4:142Þ
dr
so in balance of forces

FP þ FR = 0
dv
Δp
π
r 2 = - η
2π
r
l
ð4:143Þ
dr
i.e.,

2η
l
r
dr = -
dv ð4:144Þ
Δp
integration via the pipe radius provides
r=R v=0
2η
l
r
dr = -
dv ð4:145Þ
Δp
r=0 v=v

so

1 2 2 2η
l
R -r = þ
vð r Þ ð4:146Þ
2 Δp
i.e. a parabolic velocity profile

Δp
vð r Þ =
R2- r 2 ð4:147Þ
4η
l

From the continuity equation

dV_ = vðr Þ
dA ð4:148Þ

4.7 Rheological Measurement 215

in its form

dV_ = vðr Þ
2πr
dr ð4:149Þ

the volume flow through the capillary is

r=R
Δp
V_ =
R2- r 2
2π
r
dr ð4:150Þ
4η
l
r=0

by integration we get Hagen–Poiseuille’s law

π
Δp 4
V_ =
R ð4:151Þ
8η
l
The Hagen–Poiseuille law for laminar flows is the basis of capillary viscometry.
In general, capillary viscometers work with a fixed pressure difference, i.e. with a
constant shear stress and therefore only provide the viscosity of the fluid at this given
shear load. For this reason, capillary viscometers are suitable for Newtonian fluids.
Capillary viscometers are often made of glass and work on the basis of gravity,
i.e. the pressure difference is caused by the hydrostatic pressure of the fluid to be
examined. Frequently used designs are the capillary viscometers of Ubbelohde,
Cannon–Fenske, and Ostwald. For very viscous materials, there are metal versions
in which the fluid is pressed through the capillary with the help of an external
pressure. The extrusion test carried out with texture testers (see Chap. 5), in which a
force-loaded piston pushes the mass to be examined through an opening, can be
counted as capillary viscosity. In most cases, however, this test does not provide
absolute viscosity but a relative value, i.e. a viscosity-dependent characteristic value
that allows a comparison of samples with each other.

Ball Viscosimeter
By determining the sinking rate of a sphere, the viscosity can also be determined. For
laminar flow around a sphere, the frictional force according to Stoke´s law is

F R = 6π
r
η
v ð4:152Þ

the acceleration force is

F = ð ρK - ρF Þ
V K
g ð4:153Þ

in the balance of forces is

6π
r
η
v = ðρK - ρF Þ
V K
g ð4:154Þ
216 4 Rheological Properties

i.e. with a stationary rate of descent terminal velocity is

ðρK - ρF Þ
V K
g
v= ð4:155Þ
6π
r
η
With the ball volume

4
V= π
r3 ð4:156Þ
3
is the rate of descent is

2
g
r 2 ðρK - ρF Þ
v= ð4:157Þ
9
η
Thus, the viscosity is

2
g
r 2 ð ρK - ρF Þ
η= ð4:158Þ
9
v
Ball drop measurements are among the oldest rheological test methods. A
standardized device is the ball drop viscometer according to Höppler. In order to
have laminar conditions the rate of descent must not be too high. For this reason, the
density of the drop bodies must not be too high. Because only a single shear load is
generally measured, spherical drop viscometry provides only one point of the flow
curve. It is therefore preferably suitable for Newtonian fluids. Related measurement
methods work with bodies that are pulled by an electromagnetic force or pressed in
with weights. Modified tests are fall measurements with rods or cylinders or the
bubble viscometer, in which the rise time of gas bubbles in a fluid is measured.
Very simple measurement methods for the characterization of the flow properties
are flow-out methods. Similar to capillary viscometry, the fluid is allowed to leak out
of a cup with a defined opening under the action of gravity and the outflow time of a
given volume is determined. A number of different flow cups are in use, e.g. flow
cups after Engler, Shell, Ford, Zahn, Redwood, standard flow cups after BS, DIN,
ASTM. Based on these types of cups, there are old units such as °Engler, Redwood
seconds, Saybolt seconds, etc. Most outlet cups are available with different
diameters of the outlet nozzle. The flow cup according to DIN EN ISO 2431 has a
design with a 20 mm long outlet nozzle, which allows a calculation of the shear rate
according to Hagen–Poiseuille law [10].
Further simple tests use the propagation speed or the propagation diameter of a
point-shaped given sample, the flow time on an inclined plane, or the flow path in a
given time. With the Bostwick consistometer, the sample flows along a
predetermined path after removing a barrier. The distance traveled is measured
and given as a measure of the flow behavior of the existing material. A compilation
of viscosity tests for various industries can be found at Mezger [10].
4.7 Rheological Measurement 217

Further Reading

Cookie: Influence of solid-fat-content on fracture force [35]

Bread making: Quality related rheological parameters [36]
Wheat flour dough: creep relaxation, oscillation rheometry [37, 38]
Candy gum: Burgers model for viscoelasticity [39]
Apple: elasticity and Poisson’s ratio during storage [40]
Gelatin and alginate: Investigation of gelation by oscillation rheometry [41]
Mustard: yield flow, complex viscosity, thixotropy [42]
Ketchup: influence of hydrocolloids on flow behavior [43–46]
Exopolysaccharides: rheological characterization [47]
Mayonnaise: rheological characterization [48–50]
Chocolate: flow behavior and melting behavior [51]
Powder flow properties: tests, stickiness [52, 53]
Suspensions: in-line rheometry by pulsed ultrasound velocimetry combined [54]
with pressure difference
Cosmetics rheological characterization [55]
Maxwell model for bakery product (quince sponge cake) [56]
Cheese: rheological properties [57–60]
Stickiness of foods: test methods [52]
Apple and apple puree: viscoelasticity, Poisson’s ratio, E-modulus [61–64]
Wheat flour dough: rheological properties [36, 65,
66]
Food biopolymers: rheological properties [67–70]
Pectin gel rheological properties [71]
Xanthan gum: rheological properties of concentrated solutions [72]
Mayonnaise, NMR rheology [73]
Whey protein gel: rheological and physicochemical properties [74]
3D-printing of food: rheological description [75]
Texture design by rheological data [76]

Standards:

DIN 13316 Mechanik ideal elastischer Körper

DIN 1342 Viskosität - newtonsche Flüssigkeiten
DIN 13343 Linear viskoelastische Stoffe
DIN 53014 Kapillarviskosimeter
DIN 51562 Messung der kinematischen Viskosität mit dem Ubbelohde-
Viskosimeter
DIN 53015 Messung der Viskosität mit dem Kugelfallviskosimeter nach Höppler
DIN 53019 Messung von Viskositäten und Fließkurven mit Rotationsviskosimetern
DIN 53513 Bestimmung der viskoelastischen Eigenschaften von Polymeren
DIN EN ISO Bestimmung der Auslaufzeit mit Auslaufbechern
2431

(continued)
218 4 Rheological Properties

DIN EN ISO Viskosität mittels Rotationsviskosimeter bei definiertem

3219 Geschwindigkeitsgefälle
EN ISO 5530 Weizenmehl; Physikalische Eigenschaften von Teigen; rheologische
Eigenschaften mittels Farinograph, Extensograph, Valorigraph
DIN 10331 Bestimmung der Härte von Butter

Books and book chapters on rheology:

Creep–Recovery and Oscillatory Rheology of Flour-Based Systems, Creep and [37, 77]
Relaxation of Nonlinear Viscoelastic Materials
Networks: From Rubbers to Food [78]
Mathematical Modeling of Food Processing, Food Process Modeling [79, 80]
Food texture and structure [81]
Confectionery and Chocolate Engineering [82]
Food Properties Handbook [83]
Introduction: Food Rheology and Structure [84]
Food powders: Physical properties, processing and functionally [85]
Physical Chemistry of Foods [30]
Handbook of Food Engineering Practice [86]
Rheological methods in food process engineering [9]
Fundamentals of food process engineering [87]
Viskoelastic properties of polymers [88]
Rheology and fracture mechanics of foods [89]
Rheology of fluid and semisolid Foods: Principles and Applications [11]
Yogurt: effects of Ingredients, processing and handling on rheological [90]
properties

Summary
The flow behavior of food plays an important role in quality assurance and
process control. In this chapter, the basic concepts of rheology are systemati-
cally explained and illustrated with examples. Newtonian behavior and model
laws for non-Newtonian behavior are presented and compared with each other.
The function of the rotation rheometer is demonstrated mathematically step by
step and the principle of oscillation rheometry and numerous simple measur-
ing methods are clearly explained. At the end of the chapter, application
examples are listed that can be used for further studies and as suggestions
for your own work
References 219

References
1. Moskowitz HR (1987) Food texture. Marcel Dekker, New York
2. Szabo I (1977) Geschichte der mechanischen Prinzipien. Birkhäuser, Berlin
3. Haupt P (2002) Continuum mechanics and theory of materials. Springer, Berlin. https://doi.org/
10.1007/978-3-662-04775-0
4. Tintchev F, Bindrich U, Toepfl S, Strijowski U, Heinz V, Knorr D (2013) High hydrostatic
pressure/temperature modeling of frankfurter batters. Meat Sci 94(3):376. https://doi.org/10.
1016/j.meatsci.2013.02.012
5. Skolaut W (2018) Maschinenbau. Springer, Berlin. https://doi.org/10.1007/978-3-662-55882-9
6. Williams SH, Wright BW, Truong VD, Daubert CR, Vinyard CJ (2005) Mechanical properties
of foods used in experimental studies of primate masticatory function. Am J Primatol 67:329
7. Meschede D (2006) Gerthsen Physik. Springer, Berlin. https://doi.org/10.1007/978-3-662-
45977-5
8. Tscheuschner HD (2016) Grundzüge der Lebensmitteltechnik. Behrs, Hamburg
9. Steffe J (1996) Rheological methods in food process engineering. Freeman Press, East
Lansing, MI
10. Mezger TG (2016) Das Rheologie-Handbuch. Vincentz-Verlag, Hannover
11. Rao MA (2007) Rheology of fluid and semisolid foods: principles and applications. Springer,
New York. https://doi.org/10.1007/978-0-387-70930-7
12. Senge B, Blochwitz R, Bentlin S (2004) Rheologische Stoffkennwerte richtig bestimmen.
Deutsche Milchwirtschaft 55:256
13. Lewis MJ (1996) Physical properties of foods and food processing systems. Woodhead
Publication, Cambridge, UK
14. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, Munich
15. Kohlrausch F (1996) Praktische Physik, vol 3. Teubner, Stuttgart. https://doi.org/10.1002/piuz.
19870180109
16. Taylor RA, Chung CY, Morrison K, Hawkes ER (2014) Analysis and testing of a portable
thermal battery. J Therm Sci Eng Appl 6(3). https://doi.org/10.1115/1.4026092
17. De Graef V, Depypere F, Minnaert M, Dewettinck K (2011) Chocolate yield stress as measured
by oscillatory rheology. Food Res Int. https://doi.org/10.1016/j.foodres.2011.05.009
18. Eischen J-C, Windhab EJ (2002) Viscosity of cocoa and chocolate products. Appl Rheol 12:32.
https://doi.org/10.1515/arh-2002-0020
19. Vogel DH (1921) Das Temperaturabhängigkeitsgesetz der Viskosität von Flüssigkeiten. Phys Z
22:645
20. Kulicke W-M (1986) Fließverhalten von Stoffen und Stoffgemischen. Hüthig & Wepf Verlag,
Basel
21. DETHERM - thermophysikalische Stoffdaten DECHEMA Gesellschaft für Chemische Technik
und Biotechnologie e.V. https://dechema.de/Medien/Datenbanken/Detherm-p-24.html
22. Schneider S (2002) Viskositäten unterkühlter Metallschmelzen. Doctoral, RWTH, Aachen
23. Gröber E, Grigull U (1963) Die Grundgesetze der Wärmeübertragung. Springer, Berlin. https://
doi.org/10.1007/978-3-662-29015-6
24. Williams ML, Landel RF, Ferry JD (1955) The temperature dependence of relaxation
mechanisms in amorphous polymers and other glass-forming liquids. J Am Chem Soc
77(14):3701. https://doi.org/10.1021/ja01619a008
25. Fennema OR, Damodaran S, Parkin KL (2017) Introduction to food chemistry. In: Fennema’s
food chemistry. CRC Press, Boca Raton, FL. https://doi.org/10.1201/9781315372914
26. Bhandari BR, Howes T (1999) Implication of glass transition for the drying and stability of
dried foods. J Food Eng 40(1–2):71. https://doi.org/10.1016/S0260-8774(99)00039-4
27. Peleg M, Engel R, Gonzalez-Martinez C, Corradini MG (2002) Non-Arrhenius and non-WLF
kinetics in food systems. J Sci Food Agric 82(12):1346. https://doi.org/10.1002/jsfa.1175
220 4 Rheological Properties

28. Yildiz ME, Kokini JL (2001) Determination of Williams–Landel–Ferry constants for a food
polymer system: effect of water activity and moisture content. J Rheol 45(4):903. https://doi.
org/10.1122/1.1380425
29. Burkus Z, Temelli F (2003) Determination of the molecular weight of barley beta-glucan using
intrinsic viscosity measurements. Carbohydr Polym 54(1):51. https://doi.org/10.1016/S0144-
8617(03)00139-5
30. Walstra P (2003) Physical chemistry of foods. Marcel Dekker, New York. https://doi.org/10.
1201/9780203910436
31. Deutsches Institut für Normung (2019) Taschenbuch 398 - Rheologie. Beuth, Berlin
32. DIN 53019 Viskosimetrie - Messung von Viskositäten und Fließkurven mit
Rotationsviskosimetern (2008) Beuth, Berlin
33. DIN 53019 Viscometry - Measurement of viscosities and flow curves by means of rotational
viscometers - Part 1: Principles and measuring geometry (2008) Beuth, Berlin. https://doi.org/
10.31030/1439277
34. Gunasekaran S, Ak MM (2000) Dynamic oscillatory shear testing of foods—selected
applications. Trends Food Sci Technol 11(3):115. https://doi.org/10.1016/S0924-2244(00)
00058-3
35. Moriya Y, Hasome Y, Kawai K (2020) Effect of solid fat content on the viscoelasticity of
margarine and impact on the rheological properties of cookie dough and fracture property of
cookie at various temperature and water activity conditions. J Food Meas Charac 14(6):2939.
https://doi.org/10.1007/s11694-020-00538-6
36. Dobraszczyk BJ, Morgenstern MP (2003) Rheology and the breadmaking process. J Cereal Sci
38(3):229. https://doi.org/10.1016/S0733-5210(03)00059-6
37. Sanz T, Salvador A, Hernández MJ (2017) Creep–recovery and oscillatory rheology of flour-
based systems. In: Ahmed J, Ptaszek P, Basu S (eds) Advances in food rheology and its
applications. Woodhead Publishing, Duxford, pp 277–295. https://doi.org/10.1016/B978-0-
08-100431-9.00011-5
38. Salvador A, Sanz T, Fiszman SM (2006) Dynamic rheological characteristics of wheat flour–
water doughs. Effect of adding NaCl, sucrose and yeast. Food Hydrocoll. 20(6):780. https://doi.
org/10.1016/j.foodhyd.2005.07.009
39. Csima G, Denes LD, Vozary E (2014) A possible rheological model of gum candies. Acta
Aliment 43:36. https://doi.org/10.1556/AAlim.43.2014.Suppl.6
40. Grotte M, Duprat F, Piétri E, Loonis D (2002) Young’s modulus, Poisson’s ratio and Lame’s
coefficients of golden delicious apple. Int J Food Prop 5(2):333. https://doi.org/10.1081/
JFP-120005789
41. Panouillé M, Larreta-Garde V (2009) Gelation behaviour of gelatin and alginate mixtures. Food
Hydrocoll 23(4):1074. https://doi.org/10.1016/j.foodhyd.2008.06.011
42. Juszczak L, Witczak M, Fortuna T, Banys A (2004) Rheological properties of commercial
mustards. J Food Eng 63(2):209. https://doi.org/10.1016/j.jfoodeng.2003.07.002
43. Mansouripour S, Mizani M, Rasouli S, Gerami A, Sharifan A (2017) Effect of inulin and
galactooligosaccharides on particle size distribution and rheological properties of prebiotic
ketchup. Int J Food Prop 20(1):157. https://doi.org/10.1080/10942912.2016.1152478
44. Juszczak L, Oczadly Z, Galkowska D (2013) Effect of modified starches on rheological
properties of ketchup. Food Bioprocess Technol 6(5):1251. https://doi.org/10.1007/s11947-
012-0813-x
45. Koocheki A, Ghandi A, Razavi SMA, Mortazavi SA, Vasiljevic T (2009) The rheological
properties of ketchup as a function of different hydrocolloids and temperature. Int J Food Sci
Technol 44(3):596. https://doi.org/10.1111/j.1365-2621.2008.01868.x
46. Bayod E, Willers EP, Tornberg E (2008) Rheological and structural characterization of tomato
paste and its influence on the quality of ketchup. LWT Food Sci Technol 41(7):1289. https://
doi.org/10.1016/j.lwt.2007.08.011
47. Hundschell CS, Wagemans AM (2019) Rheology of common uncharged exopolysaccharides
for food applications. Curr Opin Food Sci 27:1. https://doi.org/10.1016/j.cofs.2019.02.011
References 221

48. Ma L, Barbosa-Cánovas GV (1995) Rheological characterization of mayonnaise. Part II: flow

and viscoelastic properties at different oil and xanthan gum concentrations. J Food Eng 25(3):
409. https://doi.org/10.1016/0260-8774(94)00010-7
49. Ma L, Barbosa-Cánovas GV (1995) Rheological characterization of mayonnaise. Part I: slip-
page at different oil and xanthan gum concentrations. J Food Eng 25(3):397. https://doi.org/10.
1016/0260-8774(94)00011-W
50. Barbosa-Cánovas GV, Kokini JL, Ma L, Ibarz A (1996) The rheology of semiliquid foods. In:
Taylor SL (ed) Advances in food and nutrition research, vol 39. Academic Press, New York, pp
1–69. https://doi.org/10.1016/S1043-4526(08)60073-X
51. Oba S, Toker OS, Palabiyik İ, Konar N, Goktas H, Cukur Y, Artik N, Sagdic O (2017)
Rheological and melting properties of sucrose-free dark chocolate. Int J Food Prop 20:2096.
https://doi.org/10.1080/10942912.2017.1362652
52. Adhikari B, Howes T, Bhandari BR, Truong V (2001) Stickiness in foods: a review of
mechanisms and test methods. Int J Food Prop. https://doi.org/10.1081/JFP-100002186
53. Ermis E, Yasasin M, Bozkurt F, Yilmaz MT (2018) Powder flow tester (PFT): a new tool to
evaluate wheat flour dough behavior by measuring unconfined failure strength under variable
consolidation stresses. J Food Meas Charac 12(3):1678. https://doi.org/10.1007/s11694-018-
9783-7
54. Wiklund J, Birkhofer BH, Jeelani S, Stading M, Windhab EJ (2012) In-line rheometry of
particulate suspensions by pulsed ultrasound velocimetry combined with pressure difference
method. Appl Rheol 22:422
55. Brummer R (2006) Rheology essentials of cosmetic and food emulsions. Springer, New York.
https://doi.org/10.1007/3-540-29087-7
56. Salehi F, Kashaninejad M (2018) Texture profile analysis and stress relaxation characteristics of
quince sponge cake. J Food Meas Charac 12(2):1203. https://doi.org/10.1007/s11694-018-
9734-3
57. Fox PF, Guinee TP, Cogan TM, McSweeney PLH (2017) Fundamentals of cheese science.
Springer, New York. https://doi.org/10.1017/CBO9781107415324.004
58. Gunasekaran S, Mehmet Ak M (2003) Cheese rheology and texture. CRC Press, Boca
Raton, FL
59. Tariq S, Giacomin AJ, Gunasekaran S (1998) Nonlinear viscoelasticity of cheese. Biorheology
35(3):171. https://doi.org/10.1016/S0006-355X(99)80006-7
60. Ray CA, Gholamhosseinpour A, Ipsen R, Hougaard AB (2016) The effect of age on Cheddar
cheese melting, rheology and structure, and on the stability of feed for cheese powder manu-
facture. Int Dairy J 55:38. https://doi.org/10.1016/j.idairyj.2015.11.009
61. Grotte M, Duprat F, Pietri E, Loonis D (2002) Young’s modulus, Poisson’s ratio, and Lame’s
coefficients of golden delicious apple. Int J Food Prop 5(2):333. https://doi.org/10.1081/
Jfp-120005789
62. Morales-Contreras BE, Wicker L, Rosas-Flores W, Contreras-Esquivel JC, Gallegos-Infante
JA, Reyes-Jaquez D, Morales-Castro J (2020) Apple pomace from variety “Blanca de Asturias”
as sustainable source of pectin: composition, rheological, and thermal properties. LWT 117:
108641. https://doi.org/10.1016/j.lwt.2019.108641
63. Salinas DE, Garvin A, Ibarz R, Ibarz A (2019) Effect of apple fibre addition and temperature on
the rheological properties of apple juice and compensation study. LWT 116:108456. https://doi.
org/10.1016/j.lwt.2019.108456
64. Espinosa-Muñoz L, Renard CMGC, Symoneaux R, Biau N, Cuvelier G (2013) Structural
parameters that determine the rheological properties of apple puree. J Food Eng 119(3):619.
https://doi.org/10.1016/j.jfoodeng.2013.06.014
65. Amjid MR, Shehzad A, Hussain S, Shabbir MA, Khan MR, Shoaib M (2013) A comprehensive
review on wheat flour dough rheology. Pak J Food Sci 23:105
66. Anderssen RS (2007) Wheat-flour dough rheology. In: The chemical physics of food. Wiley,
New York. https://doi.org/10.1002/9780470995792.ch7
67. Stokes JR (2012) Food biopolymer gels, microgel and nanogel structures, formation and
rheology. In: Food materials science and engineering. Wiley-Blackwell, New York, pp
151–176. https://doi.org/10.1002/9781118373903.ch6
222 4 Rheological Properties

68. Foegeding EA (2007) Rheology and sensory texture of biopolymer gels. Curr Opin Colloid
Interface Sci 12(4–5):242. https://doi.org/10.1016/j.cocis.2007.07.001
69. Morris ER (2000) Rheology of biopolymer co-gels. Elsevier Science, Amsterdam, pp 135–146.
https://doi.org/10.1016/b978-044450178-3/50078-0
70. Russ N (2016) Rheology and thermodynamics of starch-based hydrogel-mixtures. Johann-
Gutenberg-University, Mainz
71. Fraeye I, Doungla E, Duvetter T, Moldenaers P, Van Loey A, Hendrickx M (2009) Influence of
intrinsic and extrinsic factors on rheology of pectin-calcium gels. Food Hydrocoll. https://doi.
org/10.1016/j.foodhyd.2009.03.022
72. Song KW, Kim YS, Chang GS (2006) Rheology of concentrated xanthan gum solutions: steady
shear flow behavior. Fibers Polymers 7(2):129. https://doi.org/10.1007/Bf02908257
73. Kort D, Daan W, Nikolaeva T, Dijksman JA (2018) Rheo-NMR: applications to food. In: Webb
G (ed) Modern magnetic resonance. Springer, Cham
74. Zhang XF, Sun XM, Gao F, Wang JQ, Wang CN (2019) Systematical characterization of
physiochemical and rheological properties of thermal-induced polymerized whey protein. J Sci
Food Agric 99(2):923
75. Zhu S, Stieger MA, van der Goot AJ, Schutyser MAI (2019) Extrusion-based 3D printing of
food pastes: correlating rheological properties with printing behaviour. Innov Food Sci Emerg
Technol 58:102214. https://doi.org/10.1016/j.ifset.2019.102214
76. Matignon A, Michon C, Reichl P, Barey P, Mauduit S, Sieffermann JM (2016) Texture design
based on chemical-physics knowledge of dairy neutral desserts: instrumental and sensory
characterizations. Food Hydrocoll 52:289. https://doi.org/10.1016/j.foodhyd.2015.05.029
77. Findley WN, Davis FA (2013) Creep and relaxation of nonlinear viscoelastic materials. Dover
Publications, New York
78. Zielbauer BI, Schönmehl N, Chatti N, Vilgis TA (2016) Networks: from rubbers to food. In:
Stöckelhuber KS, Das A, Klüppel M (eds) Designing of elastomer nanocomposites: from theory
to applications. Advances in polymer science, vol 275. Springer, Cham
79. Farid MM (2010) Mathematical modeling of food processing. CRC Press, Boca Raton,
FL. https://doi.org/10.1201/9781420053548
80. Trujillo FJ (2016) Food process modeling. Elsevier, New York. https://doi.org/10.1016/b978-0-
08-100596-5.03419-3
81. Chen J, Rosenthal A (2015) Food texture and structure. In: Modifying food texture. Woodhead
Publishing, Duxford, pp 3–24. https://doi.org/10.1016/b978-1-78242-333-1.00001-2
82. Mohos FÁ (2010) Confectionery and chocolate engineering: principles and applications.
Wiley-Blackwell, Hoboken, NJ, pp 52–96. https://doi.org/10.1002/9781444320527
83. Rahman S (2009) Food properties handbook. CRC Press, Boca Raton, FL. https://doi.org/10.
1201/9781420003093
84. Rao MA (2007) Introduction: food rheology and structure. In: Rheology of fluid and semisolid
foods. Springer, New York, pp 1–26
85. Barbosa-Canovas GV, Ortega-Rivas E, Juliano P, Yan H (2005) Food powders: physical
properties, processing and functionally. Springer, New York. https://doi.org/10.1007/0-387-
27613-0
86. Taoukis PS, Labuza TP, Saguy IS (1997) The handbook of food engineering practice. Springer,
New York. https://doi.org/10.1201/9781420049077
87. Toledo R, Singh R, Kong F (1991) Fundamentals of food process engineering. Springer,
New York
88. Ferry JD (1980) Viscoelastic properties of polymers. Wiley-VCH, Weinheim
89. Vliet T (2013) Rheology and fracture mechanics of foods. CRC Press, Boca Raton, FL. https://
doi.org/10.1201/b15681
90. Clark S, Michael M, Schmidt KA (2019) Rheological properties of yogurt: effects of
ingredients, processing and handling. In: Joyner HS (ed) Rheology of semisolid foods.
Springer, Cham, pp 203–229. https://doi.org/10.1007/978-3-030-27134-3_7
Texture
5

Before we start explaining the texture study of foods, let us clarify the meaning of the
term texture. The Latin root word (lat. textura, woven material) goes back to the
characterization of woven fabrics. According to the Oxford English Dictionary
(1989), texture is0 the description of “constitution, structure or substance of anything
with regard to its constituents, formative elements.” According to ISO 5492, food
texture is “all of the mechanical, geometrical, surface and body attributes of a
product perceptible by means of kinaesthesis and somesthesis receptors and
(where appropriate) visual and auditory receptors from the first bite to final
swallowing” [1].
Accordingly, the texture of a food is a category of sensory characterization of
food and includes sensorially perceptible, mechanical, and geometric properties of a
food [2, 3]. Sensorially perceptible phenomena include haptic, auditory, and visual
phenomena. Haptic (felt) impressions are divided into tactile and kinesthetic
impressions: While tactile impressions are created when touching the food with
fingertips and tongue, kinesthetic impressions occur when breaking, bending,
chewing, or swallowing with hands, teeth, or the oral cavity. Auditory phenomena
are audible sounds that arise when breaking, biting, or chewing and often complete
the haptic impression. Visual phenomena are impressions that are perceived with the
naked eye.
After Escher [4] the overall sensory quality of a food is composed of color, flavor,
and texture. If one understands by flavor everything that can be grasped with the
sense of smell and taste, then for the concept of texture, what can be grasped are the
following:
Texture is:

– Everything that the tongue and palate grasp except taste and smell
– Everything that the eye captures except color
– Everything that the sense of touch captures
– Everything that captures hearing.

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 223
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_5
224 5 Texture

The advantage of this definition of texture is that flavor (taste and smell) and color
are clearly treated separately, i.e., must be clearly determined outside the texture
examination. Here are some examples of terms that describe texture.

Examples for terms used to describe the texture of food:

firm, crisp, soft, juicy, smooth, rough, creamy, icy, gritty, crunchy, mealy,
crusty.

If we look at the terms of sensory texture description from a physical point of

view, we can see relationships to the rheology of solids and liquids. Many texture
features can be traced back to mechanical properties such as flow behavior,
viscoelasticity, elasticity, and fracture behavior. Table 5.1 shows how texture
properties can be attributed to their physical reasons.
With the help of instrumental texture analysis [2, 5–9] attempts are made to
objectively quantify the sensory description of texture features, i.e., to make them
independent of the subjective impressions of the test persons. In the following, some
principles and measurement methods of instrumental texture analysis will be
presented.

5.1 Measurement

Individual measurement methods differ greatly depending on the physical question

at hand. For example, investigation of fracture behavior requires different
experiments from those used for determination of viscoelasticity or creaminess.
Measuring methods can first be sorted according to which type of stress is placed
on the sample or what degree of responding strain is measured. Table 5.2 shows a list
of possible measuring principles.
A distinction is made between static and dynamic tests. In static tests, the
specified load is constant over time. In dynamic tests, the applied load is varied
over time. For most dynamic tests, the load is linearly ramped up, stepped up, or
modulated in a sinusoidal form. The rate of change of the load is also of great
importance and can carried out in stages. Then, it is necessary to determine the

Table 5.1 Texture properties and their physical reasons

Texture properties Physical reasons
Hard, soft, firm Rigidity, elasticity
Elastic, rubbery Elasticity
Pasty, doughy, plastic, spreadable, tough Flow behavior, viscosity, plasticity, viscoelasticity
Short, crumbly, splitty, crispy Fracture behavior
Greasy, creamy, gel-like, spreadable Yield point, flow behavior
Granular, icy, mealy Particle shape, homogeneity
Pulpy, fibrous Particle shape
5.1 Measurement 225

Table 5.2 Options of Load Measurement of e.g.

combination of stress and
Stress σ Strain ε or strain rate ε_
measured variable
Strain ε Stress σ
Shear stress τ Shear angle γ or shear rate γ_
Shear γ Shear stress τ
Force F Distance s
Distance s Force F
Strain rate s_ Force F
Shear rate γ_ Shear stress τ

Table 5.3 Static, step-like, or dynamic load

Static load Stepwise load Dynamic load

B B B

t t t

Table 5.4 Terms in static and dynamic testing

Term Explanation
Static test Load and measured variables are constant over time
Dynamic test Load is variable over time (e.g., programmed)
Steady state Measured variables are constant over time
Transient state Measured variables are functions of time

waiting time between the individual stages. Table 5.3 illustrates the differences
between these basic principles. If the applied load in a dynamic test is periodic, it
is referred to as an oscillation test.
A further classification of measurement methods is possible based on the question
of whether equilibrium states or non-equilibrium states are recorded. In static
measurement methods, it is assumed that equilibrium states are the goal of the
measurement. To ensure this, it is advisable to install appropriate waiting times
before reading the measured variable so that thermal or mechanical balancing
processes come to an end. In dynamic testing methods, if equilibrium states are
the goal of the measurement, the rate of change of the stress must also be sufficiently
slow or the waiting time stages must be sufficiently long.
If dynamic tests have the goal to study transient states or kinetic processes like
relaxation, we have to plan conversely. Here, the reading of the measured variable
must be sufficiently fast or sufficiently frequent to record the response of the material
to dynamic stress. Table 5.4 summarizes the cases.
226 5 Texture

4 4

3 3

5 5

2 2

1 1
Fig. 5.1 Compression test between parallel plates (left). Penetration test or cutting test (right).
1 table, 2 sample, 3 force measurement, 4 lift, 5 tool

Table 5.5 Ideal material behavior under uniaxial load and shear in comparison
Elastic deformation Viscous deformation—flow
Uniaxial strain σ=E
ε σ = ηE
ε_
Shear τ=G
γ τ = η
γ_

After dealing with the characterization of viscous and elastic materials in detail in
Chap. 4, we want to start texture analysis with these material properties, too. As a
load, we apply a stress to the material and measure the reaction of the material. In
contrast to shear, we want to apply the stress uniaxially, i.e., as compressive stress or
tensile stress, at which the force acts perpendicular to the loaded surface. The left
image of Fig. 5.1 shows schematically how this can look like. A rectangular sample
is loaded on the front side, e.g., with a tensile stress or compressive stress. If the
material is elastic, the sample reacts to this load with a reversible elongation or
compression. If the material is viscous, the sample reacts to this load with an
irreversible continuous deformation, which we call flowing. In the case of uniaxial
flow, instead of viscosity, we speak of extensional viscosity. Table 5.5 compares
these terms.

σ stress in Pa
τ shear stress in Pa
E modulus of elasticity in Pa
G shear modulus in Pa
ε strain in m.m-1
γ shear angle in rad
ε_ strain rate in s-1
γ_ shear rate in s-1
η viscosity in Pa.s
ηE extensional viscosity in Pa.s
5.1 Measurement 227

Fig. 5.2 Texture testing

tools, examples

Definition
Uniaxial means “in one spatial direction.” Let us imagine a Cartesian coordi-
nate system with x-, y-, and z-axes. A movement or load along the x-axis
(or the y- or z-axis) is called uniaxial.

Experiments with uniaxial loading of a sample are comparatively easy to carry

out. Figure 5.1 schematically shows the principle of a simple tensile/compression
testing apparatus in which the sample can be compressed, stretched, penetrated, cut,
broken, etc. by a low-speed tool. Instruments used for this purpose are equipped with
load cells that measure and record load, as well as a means for measuring and
recording distance as the test proceeds. A force–distance curve is recorded and
evaluated. With known sample geometry, the force–distance curve can be used to
generate stress–strain curves.
These instruments can also be fitted with various jigs and fixtures that provide for
carrying out tensile tests, shear tests, peeling tests, and much more. With needles,
cutting edges, and conical or spherical tools (example in Fig. 5.1 right) penetration
tests can be carried out in which the penetration depth is measured as a function of
the load. Figure 5.2 shows some common tools for texture testing. The dimensions
of the tools, their ball radius, or conical angle are selected according to the sample.
Figure 5.3 shows examples of simple stress–strain diagrams obtained with solid
materials. The black dot at the end of a graph symbolizes the end of the measurement
due to material failure. Failure strain and stress can be read on the diagram axes and
used to characterize the material (Table 5.6).

Table 5.6 Simple charac- No Type E modulus Failure stress

terization of solid materials,
1 Hard strong Large High
ref. Fig. 5.3
2 Hard weak Large Low
3 Soft strong Small High
4 Soft weak Small Low
228 5 Texture

Fig. 5.3 Simple stress–strain

curves of solid materials: 1
1 hard strong, 2 hard weak, 3
3, soft strong, 4 soft weak
V
2
4

Fig. 5.4 Stress–strain curves 1

of solid bodies: 1 ideal elastic, 2
2 nonlinear elastic, V0
3 nonlinear elastic plastic, 4
4 above the yield point
σ 0 plastic flowing
V
3

In Fig. 5.4 stress–strain curves of materials are shown which, in addition to elastic
behavior, also have the ability to flow (curves 3 and 4).
Looking at Fig. 5.4, it is easy to determine the modulus of elasticity of a sample.
Table 5.6 shows the relative magnitude of the modulus of elasticity E in solid
material behaviors. The slope of the linear part of the stress–strain curve modulus
of elasticity E. Elasticity describes the degree to which the strain recovers. If the
stress-strain return path differs from the loading path, we have deviations from pure
elasticity and the area bounded by the two paths represents absorbed energy
(Fig. 5.5). However, the sample geometry and the speed of the loading tool must
be determined in advance. Let us imagine a simple uniaxial compression test on a
cubic sample compressed onto a downward plate producing a recorded stress–strain
curve.
The rate of deformation is called the strain rate ε_ ; negative strain rates are also
referred to as extension rate ε_ . It reads:

1 dL
ε_ =
ð5:1Þ
Lðt Þ dt

with
5.1 Measurement 229

Fig. 5.5 Stress–strain

diagram of a viscoelastic
material. E elastic energy,
V dissipated energy

dL
Lðt Þ = L0 -
t ð5:2Þ
dt
it is

1 dL
ε_ =
ð5:3Þ
L0 - dL
dt
t dt

so

1
ε_ = ð5:4Þ
L0
dL
dt
-t

L length in m
L0 initial length in m
ε strain in m.m-1
ε_ strain rate in s-1
A area in m2
A0 initial area in m2

It turns out that the strain rate ε_ acting on the sample is not constant but changes
over time due to the change in length of the sample as testing proceeds. This effect
becomes greater as the value of L0 decreases and the value of dL dt increases. For this
reason, small rates of change dt are recommended to keep the strain rate ε_ constant
dL

over time. With

L0
≫ time period of measurement
ðdL
dt Þ

it is

1 dL
ε_ ≈
= const:
L0 dt
230 5 Texture

Example
In order to measure the modulus of elasticity, an initially 50 mm long
rectangular sample is compressed uniaxial by 2 mm. A measurement typically
takes 1–2 minutes. With

L0
≫ time period of measurement
ðdL
dt Þ

the recommended strain rate can be calculated by

dL L0 50mm
≪ = = 25mm
min - 1
dt time period of measurement 2 min

For this to apply, the strain rate should be 2.5 mm/min or less.

Also, the area of a sample under stress is not constant during the test. In the case
of incompressible materials, it is

L0
Aðt Þ = A0

L0 - ΔL
i.e., the area increases slightly during a compression. To permit the enlarging sample
area to run smoothly, the use of suitable lubricants between the sample and the plate
may be advantageous. In addition, care must be taken when preparing samples to be
sure that the sample surface is precisely parallel to the plate on the testing instrument.
For example, if a sample with a height of L0 = 1 cm is examined and the elastic area
ends with a strain of ε = 5% then the experimental change in length is only
ΔL = ε
L0 = 0.5 mm. This means that the sample surface must not deviate from
remaining parallel by more than approximately 0.1 mm.
When examining materials by strain or shear, one has the choice between
recording the stress (or force) when specifying a deformation or deformation
rate—the so-called stress test—and recording the deformation (or path) according
to a given stress—the so-called creep test. There are different experimental variants
for both variants:

5.2 Test Options

When we plan a texture examination, the first thing to do is to determine which basic
measurement mode you want to use:
5.2 Test Options 231

1. Load on the sample by a force: measurement of the change in length

2. Load on the sample by a change in length: measurement of the force

If we work with a well-defined sample geometry, e.g., cube or cylinder with

defined dimensions, stress can be used instead of force and strain can be used instead
of change in length. Then the distinction is

1. Load on the sample by a stress: Measurement of strain

2. Load on the sample by a strain: Measurement of stress

Depending on the direction in which the force is applied, it can be a tensile test
when the sample is being stretched (increasing length or positive strain). Or it can be
a compression test when the sample is being compressed (decrease in length or
negative strain).
For an ideally elastic body, Hooke´s law applies
F = D
Δl ð5:5Þ

with the size of loaded area, A, and original sample length, l:

F D
l Δl
=
ð5:6Þ
A A l
with the abbreviations

F
σ= : ð5:7Þ
A
D
l
E= : ð5:8Þ
A
Δl
ε= : ð5:9Þ
l
Hooke’s law is
σ = E
ε: ð5:10Þ

For the ideally viscous body, Newton‘s law applies

dl
F=k
ð5:11Þ
dt
with the size of the loaded area A and the initial length l of the sample

F k
l 1 dl
=

ð5:12Þ
A A l dt
with the abbreviations
232 5 Texture

F
σ= : ð5:13Þ
A
k
l
ηE = : ð5:14Þ
A
1 dl
ε_ =
: ð5:15Þ
l dt
Newton‘s law is

σ = ηE
ε_ ð5:16Þ

F force in N.m-1
A area in m2
l length in m
Δl change in length in m
D elastic constant in N.m-1
σ stress in Pa
ε strain in m.m-1
E modulus of elasticity in Pa
ε_ strain rate in s-1
γ_ shear rate in s-1
k viscous constant in N.s.m-1
ηE extensional viscosity in Pa.s

Note that time-dependent deformation occurs in Newtonian material, while none

occurs in Hookean material. Therefore, when planning to conduct a texture test, we
must decide whether measurements should be applied over time or not. Table 5.7
helps with the decision.
Table 5.7 shows that the application of the load or strain over time t provides new
information, especially for materials that have not only elastic properties but also
viscous properties. For viscoelastic materials, therefore, the stress relaxation test or
the creep test is often chosen.
An overview like in Table 5.7 can be created analogously for shear loading. Then
the strain ε is replaced by the shear γ and the strain rate ε_ is replaced by the shear rate
γ_ . In addition to the test options mentioned (specification of a strain or specification
of a stress), there are other possibilities for viscous materials, e.g., specification of a
strain rate or a stress rate and measurement of the stress. You get flow curves, e.g.,
σ = fkt:ðε_ Þ or τ = fkt:ðγ_ Þ (refer Figs. 5.6 and 5.7). The overshoot in the flow curves is
a consequence of the elastic properties of the material.
5.2 Test Options 233

Table 5.7 Overview: planning of texture tests, uniaxial load

Load on the sample by a strain: Measurement of stress

Elastic behavior Viscoelastic behaviour Viscous behaviour (note: The

(stress relaxation) overshoot only occurs with elastic
component)
Load on the sample by a stress: Measurement of strain

Elastic behavior Viscoelastic behaviour Viscous behaviour. The slope is the

(creep) strain rate

Fig. 5.6 Flow test on a

viscoelastic material by 3
applying different shear rates
γ_ 2

W
1

Fig. 5.7 Flow test on a

viscoelastic material by 3
applying shear strain rates ε_
2

V ,F
1

t
234 5 Texture

5.3 Stress Relaxation

Figure 5.8 shows the time course of the stress after applying a sudden sustained
strain on a sample. A viscoelastic sample shows an exponential drop in stress over
time, which is caused by the fact that the sample is not only elastically deformed but
also slowly flows.
To understand viscoelasticity on a molecular basis, we recall that a material
consists of molecular sites that interact with each other via bonds. Strong bonds
cannot be easily broken, but they can be deformed while deformation energy is
stored, which is released again during reversible back-deformation. Weaker bonds
(i.e., non-covalent bonds) can be opened and re-connected with neighboring mole-
cule sites. This induces flow and permanent deformation of the material. Energy
used for flow is not stored in deformation states but is lost as heat; this is called
energy dissipation. The relaxation test shows how much a material tends to flow
under a given deformation (Fig. 5.8). A viscoelastic material is deformed abruptly
and the stress is measured. As the material begins to flow, the stress drops over time,
displaying stress relaxation. Kinetics of relaxation can be described by an exponen-
tial decay function with the characteristic relaxation time trel. Relaxation time is the
time required for the relaxing stress to reach the level 1e = 0:37 = 37% of its initial
value.
A schematic mechanical model illustrating viscoelastic and stress relaxation
behavior is the Maxwell element from mechanics; see Fig. 5.9. It consists of a spring
and dashpot (or damper) connected in series. The spring represents the elastic
component of the viscoelastic material behavior, while the dashpot represents the
viscous component. When the element is strained abruptly, the spring first absorbs
the deformation, while a gradual yielding of the damper leads to a relaxation of the
spring tension. The time course of the relaxation curve, characterized by the relaxa-
tion time (Fig. 5.8), provides information about the ratio of the elastic and viscous
parts of the material.

Fig. 5.8 Behavior of a

viscoelastic sample after
sudden application (at t = 0)
of a constant deformation. The
stress σ drops over time. The
exponential decay function is
characterized by the relaxation
time trel
5.3 Stress Relaxation 235

Fig. 5.9 Maxwell element as

a model of a viscoelastic
material

A closer look at the Maxwell element shows that the applied stress for the hook
element and the Newton element is the same, i.e.,

σ = σ Hooke = σ Newton ð5:17Þ

which means the deformation of the Maxwell element results additively.

ε = εHooke þ εNewton : ð5:18Þ

The time derivative of the deformation is

dε d
ε_ = = ðεHooke þ εNewton Þ ð5:19Þ
dt dt
with

E = const
d 1 1 dσ σ
ε_ =
σ þ εNewton =
þ ð5:20Þ
dt E E dt η
or

η dσ
σþ
= η
ε_ ð5:21Þ
E dt
with

η
trel = ð5:22Þ
E
so

dσ
σ þ t rel
= η
ε_ ð5:23Þ
dt
with
236 5 Texture

ε_ = 0 because of ε = const.
which results in

dσ
σ þ λrel
= 0: ð5:24Þ
dt
σ 1
=- dt: ð5:25Þ
dσ t rel
σ0 0
σ 1
=- dt: ð5:26Þ
dσ t rel
σ t

σ t
ln =- : ð5:27Þ
σ0 t rel
i.e.

-tt
σ = σ0
e rel : ð5:28Þ

Example
A Camembert cheese product at 10 °C has a relaxation time of

t rel = 120 min

i.e. in a relaxation test, the stress after the first minute is

σ - t
= e trel = e - 120 = 0:99 = 99%:
1

σ0
The stress relaxation after one minute is therefore only about 1%. To have
an observation time appropriate for the sample, it has to be much longer than
1 min.

5.4 Creep Test

Figure 5.10 shows the time course of deformation after application of a sudden
sustained stress to a viscoelastic sample. The deformation or strain on the sample
asymptotically approaches a maximum value, i.e., the final deformation that is
reached after a time delay due to the viscous behavior of the sample. This behavior
is known as strain retardation and is called creep.
5.4 Creep Test 237

Fig. 5.10 Time-dependent

deformation when a
viscoelastic material is loaded
with a constant stress (creep
test). To describe the
exponential kinetics, the
retardation time tret is used

Fig. 5.11 Kelvin element as

a model of a viscoelastic
material

The mechanical model used to describe creep behavior is the Kelvin element, in
which the spring and dashpot are connected in parallel (Fig. 5.11). In this case, the
dashpot initially absorbs the sudden applied stress, which gradually transfers to the
spring slowly, relaxing the stress on the dashpot.
Because the strain is the same for the Hook element and the Newton element

ε = εHooke = εNewton ð5:29Þ

the stress of the Kelvin element results additively:

σ = σ Hooke þ σ Newton ð5:30Þ

which means

σ = E
ε þ ηE
ε_ ð5:31Þ

or

1 dσ η d_ε

= ε_ þ E
ð5:32Þ
E dt E dt

ηE
with t red = ð5:33Þ
E
238 5 Texture

1 dσ d_ε

= ε_ þ t red
ð5:34Þ
E dt dt

dt = 0.
and with σ = const:, i:e: dσ
we get by integration

1 ε
- σ = : ð5:35Þ
E 0 e - tredt - 1
σ0 - t
ε=
1- e tred : ð5:36Þ
E
The deformation of the Kelvin element thus proceeds in the form of a (1 - e-time)-
function. Therefore, it is suitable as a model for describing the creep behavior of a
viscoelastic material. The kinetics of reaching the maximum value can be
characterized by an exponential saturation function and the time constant tred.
If a material shows a more complicated course of the creep curve, the Burgers
element can be used, consisting of a parallel connection of several Kelvin elements
(refer section viscoelasticity in Chap. 4).

Definition
Strain is the response of a material to a load, e.g., uniaxial stress or shear stress.
So strain has the meaning of a deformation, e.g., change in length or change of
shear angle. Depending on the load, strain can be positive or negative.

In addition, there are distinctions in the definitions of strain, e.g., Cauchy strain
(engineering strain)

ΔL
ε= ð5:37Þ
L0
and Hencky strain (true strain)

L
dL Lðt Þ
ε= = ln : ð5:38Þ
L ðt Þ L0
L0

Peleg [10] recommends the use of the Hencky definition for large stretches. For
small deformations, Cauchy and Hencky strain are approximately the same.
5.5 Deborah’s Number 239

5.5 Deborah’s Number

Is a Camembert cheese at 10 °C a solid or a fluid? What about at 25 °C? To

investigate this question, we first want to look at the lane grooves on roads. The
formation of lane grooves in asphalt roads is described with the viscoplastoelastic
properties of the material. The road at normal ambient temperature is undoubtedly a
solid. We leave the temperature constant, e.g., at 20 °C to keep the example simple.
However, an observer who examines the road surface from two photos taken ten
years apart will discover ruts and grooves in the surface that developed over the ten
years. The observer must conclude that the road is made of a material that can flow
slowly. The rut example shows that the observation time plays a role in our
assessment. In order to characterize the relationship between observation time and
flow time, Reiner [11] introduced Deborah´s number De. It is the ratio of the
characteristic time of the material and the observation time.
characteristic time of the material
De = : ð5:39Þ
observation time
As characteristic time of the material, the relaxation time can be used; then it is

t rel
De = : ð5:40Þ
t
Table 5.8 shows an overview: A large Deborah’s number indicates predominance
of solid-state behavior (elasticity). Small Deborah’s numbers indicate a preponder-
ance of fluid behavior (viscosity). Deborah’s number around 1 means that elastic and
viscous properties are both strongly represented (viscoelasticity).

Example
A Camembert cheese has a relaxation time of 120 min. With an observation
duration of 1 min,

120 min
De = = 120; therefore, the material is to be classified as solid:
1 min
At an observation duration of 8 h,

2h
De = = 0:25; the material is to be classified as flowing:
8h

Looking at ideal bodies, the concept becomes clear. The Hookean body (solid-
state properties only) has a Deborah’s number De = 1 because of its infinite
relaxation time. Pascal’s body (fluid properties only) has a relaxation time of zero;
therefore, Deborah’s number is De = 0.
Accordingly, the larger the De number, the more solid-state properties a body has.
If the observation time of a solid body is extended by a few years or perhaps by a few
240 5 Texture

Table 5.8 Deborah’s Case De Material is

number indicating material
t > trel De < 1 Viscous
properties
t < trel De > 1 Elastic
t ≈ trel De ≈ 1 Viscoelastic

thousand years, the same body can look like a fluid (low De number). This means
that if we consider large (historical) periods in addition to laboratory tests (small
periods), solids such as window glass, glaciers, and even mountains can be under-
stood not only as solids but also as flowing. Heraclitus’ saying pantha rhei (“every-
thing flows”) seems to already mean this.

Bottom Line
Solid materials have a large De number; liquid materials have a small De
number. Surprisingly, the duration of observation plays a role.

Remark
Deborah’s number is named after the clairvoyant, Deborah, mentioned in
Judges 5,5 (The Bible) where she said “The mountains flow before the Lord.”

5.6 Oscillating Load

Just as with shear, we can put an oscillating load on a body. Let us apply an
oscillating strain on a material in uniaxial direction and measure the resulting stress.
With the strain

ε=E
σ ð5:41Þ

which is oscillating, e.g., in the form of a sinus function

ε = ε0
sin ωt ð5:42Þ

the strain rate is

dε d
ε_ = = ðε0
sin ωt Þ ð5:43Þ
dt dt
i.e.
5.6 Oscillating Load 241

ε_ = ε0
ω
cos ωt ð5:44Þ

the stress is

σ = kelastic
ε þ kviscous
ε_ ð5:45Þ

or

σ = E
ε þ ηE
ε_ ð5:46Þ

or

σ = E0
ε þ E 00
ε_ ð5:47Þ

with

σ
E0 = cos δ ð5:48Þ
ε
and

σ
E 00 = sin δ ð5:49Þ
ε
we can write

σ
E = = E0 2 þ E00 2 ð5:50Þ
ε

This introduces the complex elasticity module E, which consists of the real part
E and the imaginary part E″.
′

ε Strain
ω Angular frequency in s-1
t Time in s
σ Stress in Pa
σ0 Stress amplitude in Pa
kelastic, kviscous Constants
E Modulus of elasticity in Pa
E Complex modulus of elasticity in Pa
E′ Storage modulus of elasticity in Pa
E″ Loss modulus of elasticity in Pa
δ Phase shift

So-called DMA instruments (dynamic mechanical analysis) close the gap

between measuring instruments that bring uniaxial compression stresses to the
sample and devices that exert oscillating shear stresses. Instruments used for
242 5 Texture

conducting DMA put uniaxial oscillating load on a sample, and frequency and
amplitude can be varied. These DMA measurements are fast and only need small
material samples [12, 13].

5.7 Fracture Tests

Higher stress or strain on solid and semi-solid materials often leads to fracture
(rupture) of the material. Before breakage, the behavior of the materials is usually
far from the behavior of an ideal elastic material. First there are deviations from the
elastic behavior leading to plastic deformation (flow) and finally to the irreversible
separation of bonds between the molecules (or atoms) of the solid material. This
applies analogously to uniaxial compression and shear. In addition, the rate of the
loading plays a role in the investigation of fracture behavior. There are a number of
methods for characterizing fracture behavior [9]; most simple are uniaxial compres-
sive stress or uniaxial tensile stress.
Again, we have to choose between:

• Loading by stress (tensile or compressive) up to breakage, recording of strain

or

• Loading by strain (positive or negative) up to breakage, recording of stress

The measured fracture strain (rupture strain), fracture stress, or fracture force is
specified. In addition, the fracture work Wf and the volume-related fracture work can
be determined, i.e., the mathematical product of stress and strain (shaded area in
Fig. 5.12).

F f sf Wf
σ f
εf =
= ð5:51Þ
A L0 V0

F f
sf = W f ð5:52Þ

σf Fracture stress in Pa
εf Fracture strain in s
Ff Fracture force in N
L0 Initial length in m
Wf Fracture work in J
sf Fracture deformation in m
A Area in m2
V0 Initial volume in m3
5.7 Fracture Tests 243

Fig. 5.12 Compression

fracture test, schematic. Top:
Stress–strain diagram, Vf
bottom: Force–displacement
diagram. The shaded area
symbolizes the fracture work
V

Hf
H

Ff

sf
s

The choice of testing tools is of particular importance here. For example, the
sharpness of a blade, the angle of a cone, and the diameter of a ball, which are used
for the fracture test, influence the measurement result. In addition, the attachment of
the sample plays a role. If we want to measure the fracture strain by an uniaxial
tensile test, different sample fastenings for, e.g., a foil, a noodle, or a sausage are
required. The sample attachments must be selected in such a way that they do not
affect the measurement result as far as possible, e.g., by damaging the sample. If the
sample breaks first at the clamping point, this is an indication of damage to the
sample due to the attachment. In the bending fracture test, the sample is loaded at
3 points (Fig. 5.13). Here, not only the shape of the test tool but also the geometry of
the entire arrangement must be specified when reporting the measurement result. By
the shape of the support points (round, sharp, etc.), the test can be adapted to the
question of investigation.
The 3-point bending test is particularly suitable for samples that are long com-
pared to their thickness. In the bending stress, the lower part of the sample is strained
while a compression takes place in the upper area (see Fig. 5.13).
244 5 Texture

Fig. 5.13 Three-point

bending test. N neutral axis, C
compressed axis, E strained
axis

C N
E

5.8 Texture Profile Analysis

In instrumental texture profile analysis (TPA), loads are repeatedly applied to a

sample and the sample behavior is recorded. These can be all combinations of load
and measurements presented in the sections above. Often these tests attempt to
imitate the real mechanical stress of a food, e.g., during chewing or cutting. To
imitate the scenario in the oral cavity, a fixed number of chewing movements are put
to the sample and the measurement signal is recorded over time. The evaluation can
be used to describe the time-dependent behavior of the food between the tongue and
the palate or between the teeth from the first to the last bite, and to determine cutting
force, breaking force, or stickiness of the product.

Further Reading

Theme References
Cake: texture profile analysis and stress relaxation characteristics of quince [14]
sponge cake
Cheese: rheological properties [15–18]
Stickiness in foods—test methods [19]
Food texture affected by ohmic heating [20]
Gels: relation between mechanical properties and texture [21]
Meat texture: MRI-aided texture analysis [22]
Banana: texture kinetics during ripening [23]

Summary
This chapter deals with texture as a subfield of solid-state rheology. With the
help of axial loads, typical solid-state properties such as elasticity and fracture
can be measured, but also relaxation and retardation phenomena, which are

(continued)
References 245

used to describe the viscoelasticity of materials. At the end of the chapter,

application examples are listed which can be used for further studies and as
suggestions for ongoing investigations.

References
1. ISO 5492 sensory analysis - vocabulary (2009) Beuth, Berlin
2. Bourne MC (2002) Food texture and viscosity - concept and measurement. Academic Press,
San Diego
3. Joyner HS (2018) Explaining food texture through rheology. Curr Opin Food Sci 21:7–14.
https://doi.org/10.1016/j.cofs.2018.04.003
4. Jowitt REF, Hallstrom B, Meffert HFT, Spiess WEL, Vos G (eds) (1983) Physical properties of
foods. Elsevier Applied Science, London
5. Bourne MC (2002) Physics and texture. In: Bourne MC (ed) Food texture and viscosity -
concept and measurement. Academic Press, San Diego. https://doi.org/10.1016/B978-
012119062-0/50003-6
6. Chen J, Rosenthal A (2015) Food texture and structure. Woodhead Publishing, Sawston, pp
3–24. https://doi.org/10.1016/b978-1-78242-333-1.00001-2
7. Moshenin NN (1986) Physical properties of plant and animal material. Gordon & Breach
Science Publishers, New York
8. Lillford PJ (2016) Oral perception of food texture. Elsevier, New York. https://doi.org/10.1016/
b978-0-08-100596-5.03431-4
9. Day L, Golding M (2016) Food structure, rheology, and texture. Elsevier, New York. https://
doi.org/10.1016/B978-0-08-100596-5.03412-0
10. Steffe J (1996) Rheological methods in food process engineering. Freeman Press, East
Lansing, MI
11. Reiner M (1964) The Deborah number. Phys Today 17(1):62
12. Zhong Q, Daubert CR (2013) Food rheology. In: Kutz M (ed) Handbook of farm, dairy and
food machinery engineering. Academic Press, San Diego, pp 403–426. https://doi.org/10.1016/
B978-0-12-385881-8.00015-X
13. Meyvis TKL, Stubbe BG, Van Steenbergen MJ, Hennink WE, De Smedt SC, Demeester J
(2002) A comparison between the use of dynamic mechanical analysis and oscillatory shear
rheometry for the characterisation of hydrogels. Int J Pharm 244(1):163–168. https://doi.org/10.
1016/S0378-5173(02)00328-9
14. Salehi F, Kashaninejad M (2018) Texture profile analysis and stress relaxation characteristics of
quince sponge cake. J Food Meas Charac 12(2):1203–1210. https://doi.org/10.1007/s11694-
018-9734-3
15. Fox PF, Guinee TP, Cogan TM, McSweeney PLH (2017) Fundamentals of cheese. Science.
https://doi.org/10.1017/CBO9781107415324.004
16. Gunasekaran S, Ak MM (2003) Cheese rheology and texture. CRC Press, Boca Raton, FL
17. Tariq S, Giacomin AJ, Gunasekaran S (1998) Nonlinear viscoelasticity of cheese. Biorheology
35(3):171–191. https://doi.org/10.1016/S0006-355X(99)80006-7
18. Ray CA, Gholamhosseinpour A, Ipsen R, Hougaard AB (2016) The effect of age on Cheddar
cheese melting, rheology and structure, and on the stability of feed for cheese powder manu-
facture. Int Dairy J 55:38–43. https://doi.org/10.1016/j.idairyj.2015.11.009
19. Adhikari B, Howes T, Bhandari BR, Truong V (2001) Stickiness in foods: a review of
mechanisms and test methods. Int J Food Prop. https://doi.org/10.1081/JFP-100002186
246 5 Texture

20. Gavahian M, Tiwari BK, Chu Y-H, Ting Y, Farahnaky A (2019) Food texture as affected by
ohmic heating: mechanisms involved, recent findings, benefits, and limitations. Trends Food
Sci Technol 86:328–339. https://doi.org/10.1016/j.tifs.2019.02.022
21. Santagiuliana M, Piqueras-Fiszman B, van der Linden E, Stieger M, Scholten E (2018)
Mechanical properties affect detectability of perceived texture contrast in heterogeneous food
gels. Food Hydrocoll 80:254–263. https://doi.org/10.1016/j.foodhyd.2018.02.022
22. Bajd F, Skrlep M, Candek-Potokar M, Sersa I (2017) MRI-aided texture analyses of compressed
meat products. J Food Eng 207:108–118. https://doi.org/10.1016/j.jfoodeng.2017.03.026
23. Nannyonga S, Bakalis S, Andrews J, Mugampoza E, Gkatzionis K (2016) Mathematical
modelling of color, texture kinetics and sensory attributes characterisation of ripening bananas
for waste critical point determination. J Food Eng 190:205–210. https://doi.org/10.1016/j.
jfoodeng.2016.06.006
Interfaces
6

The contact surface between two phases is called an interface. Phase is understood
here to be a homogeneous state form of a substance, which can be distinguished from
another phase by this recognizable separation surface. A prerequisite for the forma-
tion of an interface is the incomplete miscibility of the phases involved. If two
substances are completely miscible into each other (such as two gases or a pinch of
salt in water) no permanent interfaces are formed. Assuming three states of aggrega-
tion (solid, liquid, and gaseous), the following interfaces are possible (Table 6.1):
Interfaces at solid materials are also referred to as solid interfaces due to their lack
of mobility. In Chap. 1, the sorption at solid interfaces is treated. Fluid–fluid
interfaces are referred to as fluid interfaces, sometimes also referred to as liquid
interfaces, because of their mobility. Those interfaces in which one of the phases
involved is gaseous are called surfaces. Then, we speak of surface tension instead of
interfacial tension.

Definition
An interface is the area that is located at the boundary of two phases. It
separates these phases (solid, liquid, and gaseous) from each other. Surfaces
belong to the interfaces.

Since the physical properties of the interfaces differ from the physical properties
within the substance of the same material, the surface layer of a phase is sometimes
referred to as an independent phase, the so-called interfacial phase [1]. This is
especially true in the case of molecular layers that are adsorbed at an interface.
Examples of this are water molecules on solid surfaces (see Chap. 1) as well as
adsorbed surfactants or emulsifiers at liquid interfaces. While a molecule inside a
substance is surrounded by the same molecules, this is not the case at interfaces. At
the interface to another phase, interactions with non-identical molecules occur. As a
result, the molecules at the interface differ in their energetic state and their physical

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 247
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_6
248 6 Interfaces

Table 6.1 Classification of interfaces

Fluid–fluid interfaces Solid interfaces
Liquid–gaseous e.g., beer–CO2 Solid–gaseous e.g., NaCl–air
Liquid–liquid e.g., water–oil Solid–liquid e.g., sugar–oil
Solid–solid e.g., glass–plastic

behavior from molecules within the substance. The difference in energetic state is the
cause of interfacial phenomena such interfacial tension. For phases that are not
composed of molecules but of individual atoms, e.g., noble gases or metals, this
applies analogously. In this chapter, we will first look at two-phase systems and look
at three-phase systems.

6.1 Interfacial Tension

A molecule M in the substance of a liquid is surrounded on all sides by molecules of

the same substance. The intermolecular interaction forces on the molecule M are the
same in all spatial directions, so that the resulting force on M is zero (Fig. 6.1). At the
interface, on the other hand, the interaction forces do not compensate each other in
all spatial directions, so that a resulting force is created. If the intermolecular
interaction forces in the liquid phase are greater than the intermolecular interaction
forces between the molecules of the liquid phase and the adjacent gas phase (in the
case of water-air, as in Fig. 6.1), then a force F results for a liquid molecule M, which
is directed into the liquid phase.
If we want to increase the interface area, we have to bring liquid molecules from
the volume to the interface against this force F. Enlargement or creation of an
interface is therefore associated with a work or energy expenditure
(work = force
distance), while the reduction of the interfacial area is associated
with an energy gain.

Bottom Line
Molecules in interfaces have a higher energy than molecules within the same
substance. The surface-specific energy of interfaces is given in J
m-2 and also
referred to as interfacial tension.

Fig. 6.1 Intermolecular forces within the substance and at the interface. A molecule M experiences
a net force F at the interface, which is directed into the substance
6.1 Interfacial Tension 249

Fig. 6.2 Creation an element db

of interface dA

s dA F

Since interfaces are high in energy, free liquid droplets tend to take on a spherical
shape. The spherical surface is a minimal surface. At a given volume, all other three-
dimensional shapes have a larger interface, i.e., a greater energy than the sphere
(refer to Platon’s bodies in Chap. 3). However, most liquid droplets do not have an
ideal spherical shape, because of other influences such as gravity, adhesive forces, or
frictional forces in a flow (e.g., air friction).
Let us look at an interface represented by the thin wire of length, s, shown in
Fig. 6.2. The area of this interface can be enlarged by moving the wire a distance, db,
represented by the cross-hatched area in Fig. 6.2. We can see that the creation of this
additional interface size, dA = s
db, is related to an energy expenditure of dE = F
db.
This area-related energy is called specific interfacial energy σ which can also be
called area-specific interfacial energy or specific interfacial energy.
ΔE dE
σ = lim = ð6:1Þ
ΔA → 0 ΔA dA
F
db F
σ= = ð6:2Þ
s
db s
F force in N
A area in m2
σ interfacial tension in N∙m-1
E specific interfacial energy in J∙m-2
b length in m
s width in m

The surface-specific interfacial energy is identical to the interfacial tension. The

units of surface-related energy and interfacial tension are identical:

J Nm N
1 =1 2 =1
m2 m m
.
The energy for the generation of an interface dA or ΔA can thus be calculated
according to
dE = σ
dA ð6:3Þ
250 6 Interfaces

or

ΔE = σ
ΔA ð6:4Þ

The energy used to generate an interface ΔA is thus directly proportional to the

value of the interfacial tension between the two phases. In processes where interface
creation occurs, knowledge of interfacial tension is helpful, as well as knowledge
how to reduce it. A lower interfacial tension means that the energy for the interfacial
formation process is lower.
Figure 6.1 shows the forces on a molecule in a flat interface. The forces tangential to
the interface cancel each other out to zero. The forces perpendicular to the interface do
not compensate for each other, so there is a resulting net force directed into the liquid
volume. In curved interfaces, the tangential forces similarly do not fully compensate for
each other. This is the reason why curved interfaces have additional effects.

6.1.1 Curved (Convex/Concave) Interfaces

When a beaker is overfilled with water, a curved interface between the water and air
is formed at the top surface, and the water reaches a slightly higher level at the center
than at the rim of the beaker. It looks as though an invisible thin skin would retain the
water at the top of the beaker. This invisible “thin skin” is caused by what we call
“surface tension” and is a result of the interfacial energy discussed previously. This
phenomenon is also the reason why the molecules in the inner volume are exposed to
a slight pressure which would not be there without that “skin.” This pressure is called
capillary pressure.
The curvature effect is caused by the fact that at a curved interface, the tangential
forces acting on a molecule in the interface do not compensate to zero. This produces
a concave interface resulting from an increase in the forces that are directed inward
into the volume of liquid (see Fig. 6.3).

Fig. 6.3 Forces on a

molecule in a liquid interface.
With a concave curved
interface, the resulting force
directed into the liquid
volume is greater than with a
flat interface
6.1 Interfacial Tension 251

Fig. 6.4 Cross-section of a dr

spherical droplet of fluid
whose radius is increased
by dr
r

Virtual displacement: When we want to increase the volume of a liquid drop by

adding a given amount of liquid to it (see Fig. 6.4), we have to do work against the
capillary pressure pσ . This form of work is called volume work dWP

dW P = - pσ
dV ð6:5Þ

WP volume work in N∙m

Eσ specific interfacial energy in N∙m
pσ pressure in N∙m-2
V volume in m3

For a sphere with V = 43 π
r 3 the dependence of the volume on the radius is

dV d 4
=
π
r 3 = 4π
r 2 ð6:6Þ
dr dr 3
used in Eq. (6.5) gives

dW P = - pσ
4π
r 2
dr: ð6:7Þ

The droplet, whose radius was increased by dr, gets a larger interface. The
interfacial energy required for this is

dE σ = σ
dA: ð6:3Þ

dr for a sphere with A = 4π
r we get

2
Writing the differential dA

dA d
=
4π
r 2 = 8π
r ð6:8Þ
dr dr
so

dA = 8π
r
dr ð6:9Þ
252 6 Interfaces

That means

dEσ = σ
8π
r
dr ð6:10Þ

According to energy conservation, the volume work performed is equal to the

interfacial energy Eσ , i.e.,

dEσ þ dW P = 0 ð6:11Þ

or

dE σ = - dW P ð6:11Þ

used in Eq. (6.10) gives

σ
8π
r
dr = - - pσ
4π
r 2
dr ð6:12Þ

so

2σ
pσ = : ð6:13Þ
r
This is called Laplace’s equation. The pressure in an interfacial sphere is called
capillary pressure according to Laplace. The radius r in Laplace’s equation is the
radius of curvature of the observed interface. Convex curved interfaces have positive
curvature radii, while concave curved interfaces have negative radii of curvature.

Example
Mist consists of fine water droplets distributed in air. In a spherical drop of
water with a radius of 70 nm and

σ = 70 mN
m - 1

the pressure according to Laplace is

2σ 2
70mN
m - 1 2
10 - 3 N
m - 1
pσ = = = = 20
105 Pa ≈ 20bar:
r 70
10 - 9 m 10 - 9 m

The capillary pressure according to Laplace is the overpressure of the phase with
the concave-shaped interface compared to the adjacent phase. If phases 1 and 2 have
absolute pressures p1 and p2, then the difference in absolute pressures provides the
overpressure in the concave phase. Concave-shaped particles are typically droplets
or gas bubbles.
6.1 Interfacial Tension 253

2σ 12
p1 - p2 = = pσ ð6:14Þ
r1
σ 12 interfacial tension between phase 1 and 2 in N m-1
r1 radius of curvature on phase 1 in m
p1 absolute pressure in phase 1 (droplet) in Pa
p2 absolute pressure in phase 2 (surrounding) in Pa
pσ capillary pressure (overpressure of phase 1) in Pa

Bottom Line
The pressure inside a drop of liquid or in a gas bubble is increased compared to
the environment. The smaller the radius of curvature, the greater the pressure
difference to the environment.

Laplace’s equation is also applicable to concave-shaped interfaces. Because

negative values for the radius of curvature are used, the capillary pressure is
negative; i.e., there is a condition of underpressure compared to the other phase.
At a flat interface, the radius of curvature is infinitely large and the Laplace’s
equation returns the value zero; i.e., the pressure is the same on both sides of the
interface. Table 6.2 summarizes these cases.

Attention
It is easy to get confused with the terms concave and convex: A concave-
shaped interface looks convex to a viewer localized behind it. Think of a cereal
bowl turned upside down on a table. If you were inside the bowl, the curved
surface above would appear concave, but outside the bowl, it will appear
convex. To avoid this confusion, we must be clear “on which side we are,” i.e.,
for which of the two phases we perform the calculation.

Table 6.2 Capillary pressure of differently curved interfaces

Interface Curved Curved Not curved
Appearance of the Phase 2 Phase 2 Phase 2
interface, schematic
Phase 1 Phase 1 phase 1
Interface of phase 1 is... concave convex flat
Radius of curvature r1 r1 < 0 r1 > 0 r1 = 1
p1 - p2 = 2σ 12
r1
<0 >0 0
In phase 1 there is Underpressure Overpressure Same pressure
compared to phase 2
Example: Gas bubble (phase 2) in Droplet (phase 1) in Flat water
water (phase 1) air (phase 2) surface
254 6 Interfaces

Example
We have an air bubble with a radius of 1.4 mm in water. We imagine to sit
inside the air bubble: We are in phase 1 and see an concave interface with
radius r1 of curvature. Laplace’s equation p1 - p2 = 2σ
r1 = pσ gives a positive
-3
value: pσ = 2σ
r1 = 1:4
10 - 3 Pa = 100 Pa, i.e., the pressure in phase 1 is greater
2
70 10

than in phase 2, i.e., we are sitting in overpressure. Now we change sides and
enter the water phase. We are now in phase 2 and see the air bubble interface in
front of our eyes in the form of a convex interface, i.e., r2 = - 1.4 mm.
Laplace’s equation p1 - p2 = 2σ r2 = pσ gives

2σ 2
70 10 - 3
pσ = = Pa = - 100 Pa:
r2 - 1:4
10 - 3
So we are sitting in underpressure. The pressure in phase 2 is lower than in
phase 1.

If a drop does does not have an ideal spherical shape, its shape can be described,
for approximately, e.g., as an ellipsoid with 2 main radii of curvature r+ and r-. Then
Laplace’s equation is

1 1
pσ = σ
þ : ð6:15Þ
rþ r -

For spherical drops r+ = r- = r resulting again in Eq. (6.13): pσ = 2σr .

Capillary pressure causes the vapor pressure of particles to be higher than the
vapor pressure of the flat phase [2]. This effect is higher the smaller the particle
diameter is. Kelvin’s equation used in Chap. 1 allows the calculation of the vapor
pressure of a liquid drop as a function of its radius.

Example
Vapor pressure of a spherical water droplet with a diameter of 20 nm at 20 °C
according to Kelvin:

p 1 2
σ 1 2
72:25
10 - 3 N
m - 1
ln =
=
:
p0 r p ρ
R s
T rp 999 kg
m - 3
461:9 J
K
kg - 1
293:15 K

(continued)
6.1 Interfacial Tension 255

p 1
ln =
1:0682 nm = 0:10682
p0 10 nm
p
= e0:10682 = 1:1127
p0

The vapor pressure of a water droplet with a diameter of 20 nm is 11.1%

higher than the vapor pressure of water, which is free with a flat surface.

Because of this increased vapor pressure, small droplets evaporate more quickly
in a gas phase than larger droplets. Small droplets in a liquid, e.g., in an emulsion,
have a higher solubility than in the flat phase because of this increased vapor
pressure. Since this effect is independent of the physical state and is based exclu-
sively on the curvature of the interface, it also occurs with solid substances. This
means that small solid particles have a higher vapor pressure and thus a higher
solubility or a higher sublimation pressure than the same substance in a flat phase.

Example
Two drops of liquid of different sizes are brought to the touch, so that a
pressure equalization of their capillary pressures can take place. What will
happen?

(a) the larger drop becomes smaller and the smaller one larger
(b) the larger drop becomes larger; the smaller one becomes smaller and
disappears completely
(c) both drops take on the same size.

Answer: The smaller drop has a higher pressure. Therefore, the smaller
drop fills the larger one until the small one is completely gone. Answer b is
correct.

A disperse system with different particle diameters such as a spray—is not

energetically stable. The different pressures of the particles with different diameters
lead to a balancing process in which small particles disappear in favor of larger
particles. This effect, known as Ostwald ripening, occurs in liquid, solid, and
gaseous particles. Therefore, small particles often have a short life time because
the energy of the particles depends on their radii. The disappearance of the small
particles, i.e., the coarsening of the disperse system, leads to a reduction in the total
interfacial energy of the system [3].
256 6 Interfaces

6.1.2 Temperature Dependence of Interfacial Tension

The interfacial tension between two phases decreases with increasing temperature.
The surface tension of liquids goes to zero when the critical temperature is reached.
Above the critical point, there is no longer an interface between the liquid and
gaseous phases. If we assume a liquid approximately composed of cube-shaped
molecules of edge length l, the volume of one mole of these molecules, the so-called
molar volume, is

V m = N A
l3 ð6:16Þ

a molecule in the surface requires the area

2
Vm 3
l2 = ð6:17Þ
NA

One mole of this substance therefore requires the area Am

1 2
Am = N A
l2 = N A 3
V m 3 ð6:18Þ

The surface energy of a mole of these molecules is

1 2
σ m = σ
Am = σ
N A 3
V m 3 ð6:19Þ

with the molar volume

M
Vm = ð6:20Þ
ρ

The surface energy of a mole is

2
1 M 3
σm = σ
N A 3
ð6:21Þ
ρ

According to Eötvös, the temperature dependence of the molar interfacial

energy is
σ m = kE
ðT C- T θ - T Þ ð6:22Þ

Equalizing (6.21) and (6.22) provides the equation for the temperature depen-
dence of the interfacial tension according to Ramsay and Shields [4]

σ
N A 3
M 3
σ - 3 = k E
ðT C- T θ - T Þ
1 2 2
ð6:23Þ
6.1 Interfacial Tension 257

Fig. 6.5 Temperature

dependence of interfacial
tension

i.e.,

σ = N A - 3
M - 3
ρþ3
kE
ðT C- T θ - T Þ
1 2 2
ð6:24Þ

For substances with a strong association of liquid molecules (as is the case with
water), the correction factor χ is introduced:

σ = N A - 3
ðM
χ Þ - 3
ρþ3
kE
ðT C- T θ - T Þ
1 2 2
ð6:25Þ

kE Eötvös coefficient
T temperature in K
TC critical temperature in K
Tθ material constant in K
NA Avogadro’s number
Vm molar volume in m3
mol-1
ρ density in kg
m-3
σ interfacial tension in N
m-1
M molar mass in kg
mol-1
χ correction factor

Figure 6.5 schematically represents the temperature dependence of the interfacial

tension of a pure substance.

Example
Calculation of the surface tension of water at 20 °C.
with data for water

kE Eötvös coefficient for H2O: 7.5 J
K-1 mol-1

TC critical temperature for H2O: 647.1 K
Tθ material constant for H2O: 6

(continued)
258 6 Interfaces

NA Avogadro’s constant 6.023∙1023 mol-1

ρ density of H2O: 998.2 kg∙m-3
M molar mass of H2O: 18.015∙10-3 kg∙mol-1
χ correction factor for molecular association of water at 20 °C: 0.47)

we get
2
ð998:2 kg
m - 3 Þ3
7:5 N
m
K - 1
mol - 1
σ= 1 2
6:023
1023
mol - 1 3

0:47
18:015
10 - 3 kg
mol - 1 3

ð647:15 K - 6 K - T=KÞ

this numerical equation can be summarized to

mN
σ = 0:2135
ð368:0 - ϑ= ° CÞ
m
for 20 °C we get

σ = 0:2135mN
m - 1
ð368:0 - 20:0Þ

σ = 74:3mN
m - 1

This numerical equation is based on fundamental quantities, such as Avogadro’s

number and molecular weight, and easily provides values for the surface tension of
water at different temperatures. However, not taking into account the temperature
dependence of the density and the association factor diminishes the utility of that
equation. A better equation can be produced by fitting the experimental data to a
mathematical function without attaching importance to any physical meaning of the
mathematical parameters (“curve fitting”). Table 6.4 shows which improvements of
fitting quality can be achieved. Three numerical value equations were used as
examples (refer Table 6.3) and the deviations from the experimentally determined
values of the surface tension listed.
A comparison of the calculated values with the experimentally obtained values
shows that C provides the smallest deviations. Also, the simple linear regression
according to B provides a better fit than A. While B and C are purely empirical
adaptations, A is derived from fundamental variables. However, it still contains the
empirically determined correction factor χ, which depends on the temperature. The
quality of A can be improved if we take into account the temperature dependence of
the correction factor χ and the density. However, by this, the equation becomes more
complicated.
Why this comparison of the models?
For process engineering calculations, material data are always required. The
interfacial tension here serves as an example of such a material property. If we
6.1 Interfacial Tension 259

Table 6.3 Equations for calculating the surface tension of water (examples)
A Ramsay, Shields [4] σ = 0:2135 mN
m
ð368:0 - ϑ= ° CÞ
-1
B Kohlrausch [5] σ/mN
m = 76.056 - 0.1675
ϑ/ ° C
μ
C International steam tables [6] σ=mN
m - 1 = B
TC - T

1þb
TC - T
TC TC
with
B = 235.8
10-3 Nm-1
b = -0.625
μ = 1.256
TC = 647.096 K

Table 6.4 Comparison of functions for calculating the surface tension of water
Experimental
value [5] Calculated with A Calculated with B Calculated with C
σ cal/ σ cal/ σ cal/
ϑ/° mN∙m- mN∙m- mN∙m-
C σ exp/mN∙m-1 1
Deviation 1
Deviation 1
Deviation
0.01 75.65 78.57 3.9% 76.05 0.5% 75.65 -0.001%
10 74.22 76.43 3.0% 74.38 0.2% 74.22 0.006%
20 72.74 74.30 2.1% 72.71 0.0% 72.74 0.000%
30 71.20 72.16 1.3% 71.03 -0.2% 71.20 -0.003%
40 69.60 70.03 0.6% 69.36 -0.4% 69.60 0.001%
50 67.95 67.89 -0.1% 67.68 -0.4% 67.95 -0.002%
60 66.24 65.75 -0.7% 66.01 -0.4% 66.24 0.005%
70 64.49 63.62 -1.3% 64.33 -0.2% 64.49 -0.006%
80 62.68 61.49 -1.9% 62.66 0.0% 62.68 -0.003%
90 60.82 59.35 -2.4% 60.98 0.3% 60.82 0.003%

find what we are looking for in food material data collections, we often have to be
satisfied with the value of a material of similar composition, and we have to accept
the value as an approximation. Another possibility is measurement of the required
property in compliance with the rules of food physics measurement technology. This
results in a true value for the material actually present. However, further
measurements will be necessary as soon as the temperature changes or, e.g., the
water content is to be changed slightly. An alternative is the use of mathematical
functions which take into account the influence of, e.g., temperature and composi-
tion. Such functions are referred to as mathematical models and are used for
automatic calculation of the desired material data and ensure that the process
calculation is not based on incorrect data. If there are several functions from which
to choose, a decision is necessary as to which model should be used. Deviation
between the calculated and experimentally determined value is often used as a
decision criterion for the quality of the function to be chosen. Such a comparison
is shown in Table 6.4. In addition to a comprehensible decision for one of the
models, a quantitative measure about the quality of the adaptation can be obtained by
giving the average or maximum deviation of the data used. An indication of the
260 6 Interfaces

reliability of the process calculation then can be made by means of error propagation
calculation. In process calculations in which dozens of parameters and material data
are included, it is highly recommended to quantify the uncertainty of the calculation
in this way.

Bottom Line
When physical substance data of foods and their components are missing,
approximate values must be used for process calculations. In order to be able
to quantitatively specify the resulting uncertainties, it is recommended that
substance data of the individually available material be experimentally
recorded and expressed in the form of mathematical functions of their
parameters, such as temperature, pressure, or composition. The quality of the
function used can then be expressed by comparing the calculated and
measured values, thus providing a measure of the quality of the
approximation.

The values of the surface tensions of some foods are listed in the appendix.

6.1.3 Concentration Dependence of Interfacial Tension

Amphiphilic molecules—surfactants—emulsifiers
Substances that arrange themselves at the interface between two phases and thus
cause a change in interfacial tension are referred to as interfacial active substances.
This effect is particularly evident in amphiphilic molecules. Due to their molecular
structure, even small amounts of these dissolved substances lead to a significant
reduction in interfacial tension. Such substances are, therefore, also called
surfactants. Amphiphilic molecules have both lyophilic (“solvent-friendly”) and
lyophobic (“solvent-hostile”) groups. If the solvent is water, we speak of hydrophilic
(“water-friendly”) and hydrophobic (“water-hostile”) groups. While the lyophilic
molecular parts enter into strong interactions with the solvent molecules, the ener-
getically most favorable state for the lyophobic molecular parts is the avoidance of
interactions with the molecules of the solvent. For this reason, the amphiphilic
molecules accumulate in the boundary layer, known as boundary layer adsorption
[7]. In the boundary layer, the molecules orient themselves in such a way that the
lyophilic molecular parts interact strongly with the solvent and the lyophobic
molecular parts avoid interaction with the solvent. An interface occupied by inter-
face-active molecules in that way is in an energetically favorable situation; therefore,
its specific interfacial energy (see Sect. 6.1) is lower than in the absence of these
molecules. With increasing concentration of the surface-active substance, the occu-
pancy of the interface increases and the interfacial tension continues to decrease (see
Fig. 6.6). When complete monomolecular occupation of the interface is achieved,
there is no more space in the interface for further molecules; the maximum interface
occupancy is reached. On further increase in the surfactant concentration, the
6.1 Interfacial Tension 261

Fig. 6.6 Concentration

dependence of interfacial
tension by surface-active
substances

cmc
c

surfactant molecules therefore start to form associates in the volume of the solvent,
so-called micelles. The concentration where micelle formation begins is called
critical micelle formation concentration, cmc. Micelles can be spherical, rod shape,
tubular, or lamellar. Micelles can interact with each other and can form structures. In
addition, there are phase transitions in which the structure of the micellar aggregates
changes [8].

Definition
Micelles are colloidal aggregates formed by association of amphiphilic
molecules in solutions.

Since the addition of surface-active substance beyond the cmc no longer changes the
occupancy at the interface, the interfacial tension remains unchanged above this
concentration. The interfacial tension achieved at the cmc is therefore a characteristic
quantity for the interfacial active substance used. In Fig. 6.7, the achievement of the
maximum interface occupancy is shown in a glass beaker, schematically.
The concentration dependence of the interfacial tension for surface-active
substances can be described mathematically with the Szyszkowski equation [3]:

c
σ 0 - σ = a
ln 1 þ ð6:26Þ
b
σ interfacial tension in N
m-1
σ0 interfacial tension bei c = 0
c concentration of the surface-active substance in kg
m-3
b material constant in m3
kg-1
a material constant in N
m-1

Analogous to the cmc, the temperature at which micelles form is called the critical
micelle temperature (cmt). If the temperature of a surfactant solution is lowered,
micelles form when passing through the cmt. Conversely, on increasing the
262 6 Interfaces

Fig. 6.7 Once the maximum

interface occupation is
reached, the interfacial tension
remains constant

Fig. 6.8 Surfactants in an

aqueous solution: Monomers
in solution (I). Micelles in
solution (II). Liquid crystal
suspension (III). Krafft point
(K ). after [9]

temperature, cmt can be recognized by the beginning clearness of a cloudy, colloidal

micelle dispersion.
The intersection of the cmt curve and the cmc curve in the surfactant phase
diagram is called Krafft point; the critical micelle formation temperature is also
called Krafft temperature (Fig. 6.8) [10].
The structure of the micelles is ruled by the most energetically favorable state of
the amphiphilic molecules in the solvent at given concentration and temperature.
Since during micelle formation both the entropy and the energy of the system
change, the free enthalpy (Gibbs energy; see Chap. 8) indicates the stability of the
6.1 Interfacial Tension 263

system. The minimum of free enthalpy determines the stability of the system
[7]. Association and orientation of the molecules take place in such a way that
lyophilic interactions are in favor before interactions of lyophobic molecular parts
and solvent molecules. For example, hydrophilic groups are interacting with the
water molecules, while hydrophobic groups are facing each other. If lamellar
bilayers form aggregates with a cavity, this is referred to as vesicles.

Attention
Vesicles and micelles are different forms of association aggregates. Vesicles
consist of concentrically arranged lamellar molecular bilayers and have a
solvent-filled cavity. Vesicles are also called liposomes.

Micelles and vesicles are used as capsules and vehicle systems for poorly soluble
active ingredients. For example, casein micelles can be used for microencapsulation
[11] or water-insoluble substances are introduced into body cells with the help of
vesicles [8].
There are substances which, in contrast to surfactants, increase the interfacial
tension of a liquid when dissolved. In the case of aqueous systems, these are mainly
strong electrolytes (inorganic salts) and hydroxyl-rich compounds such as
carbohydrates. The reason for this is the strong interaction of these substances
with water (hydration) and thus an accumulation of the substances in the liquid
volume, i.e., a depletion of the substances at the interface.

6.1.4 Emulsions

An emulsion is a disperse system of two immiscible liquid phases. Disperse means

that one phase is distributed in the form of small particles - here liquid droplets—in
the other liquid phase. A prerequisite for the formation of an emulsion is the
non-miscibility of the two phases. Otherwise, a solution would be formed and not
a disperse system. If the disperse phase consists of oil and the continuous phase of
water, it is called an O/W emulsion (oil-in-water emulsion). During forming of an
emulsion, interface is created, as the following example shows.

Example
In a beaker, 100 ml of oil and 100 ml of water are layered on top of each other.
Assume the beaker has diameter of 10 cm. The size of the circular oil–water
interface is

A0 = π
r 2 = π
ð5 cmÞ2 = 7:85
10 - 3 m2

(continued)
264 6 Interfaces

An O/W emulsion is prepared from this. The resulting oil droplets are
spherical with a uniform diameter of 20 μm. What is the area of oil–water
interface in the emulsion?
Solution:
First we calculate the volume of a single oil droplet:

4 4 3 4
Vd = π
r 3d = π
10 - 5 m = π
10 - 15 m3
3 3 3
Then the surface of this oil droplet:

2
Ad = 4π
r 2 = 4π
10 - 5 m = 4π
10 - 10 m2

Now we calculate the number of oil droplets by dividing the existing

volume of oil by the volume of a droplet:

V oil 100 ml 100
10 - 6 m3 3

N= = =4 - 15 3
=
1011 = 2:387
1010
Vd Vd
3 π
10 m 4π

The area of the interface in the emulsion results from the sum of all oil
droplet interfaces:

Atotal = N
Ad = 2:387
1010
4π
10 - 10 m2 = 30 m2

If we compare this size of the interface with the initial interface in the
beaker, we get

Atotal 30 m2
= = 3821
A0 7:85
10 - 3 m2
i.e., during the production of the emulsion, the area of the oil–water interface is
increased by some thousands.

The interfacial energy of the emulsion is calculated according to

ΔE = σ
ΔA ð6:4Þ

By lowering the interfacial tension σ, the interfacial energy of the emulsion can be
reduced. For this purpose, suitable surface-active substances (see Sect. 6.1.3) are
used. Analogous to the O/W emulsions, such as milk, cream, or salad dressing, there
are water/oil (W/O) emulsions, such as butter, in which water droplets are dispersed
in an oil phase.
By adding of suitable co-surfactants, we get optically transparent emulsions. Such
clear or translucent preparations are called microemulsions. That is because it was
assumed that the emulsified particles are smaller than the wavelength of visible light,
6.1 Interfacial Tension 265

i.e., below 400 nm. In contrast to emulsions, microemulsions are considered ther-
modynamically stable. Since no droplets can be detected even in the electron
microscope, it is questionable whether it is an emulsion at all or rather a single-
phase, ternary colloidal system of water, oil, and surfactant [12].
Cosmetic products with encapsulated oil in capsules with diameters in the range
of 1–100 nm are sometimes referred to as nanoemulsions [13, 14]. The capsules
consist, e.g., of a phosphatidylcholine single membrane that encloses a liquid lipid
core. In order to distinguish between capsules and conventional emulsion droplets,
preparations with nanoscale capsules should be called nano-dispersions.
Pickering emulsions are emulsions in which solid particles are used as surface-
active substances. With suitable solids the amount of classic emulsifiers can be
reduced [15].
In addition, it is possible to disperse an emulsion in another phase. In this way,
e.g., O/W/O emulsions or W/O/W emulsions, so-called multiple emulsions can be
formed [16–18].

Definition
Multiple emulsions are emulsions in which the dispersed phase itself is an
emulsion.
There are numerous options to make such emulsions usable. One option is
to have the inner phase carry an active ingredient and the outer phases used to
estabish the desired flow behavior of the emulsion. By use of co-surfactants,
viscosity-increasing thickeners, and gelling agents, the number of technologi-
cal options can be further increased. While three or more different substances
are used in such systems, we still have a two-phase system (liquid–liquid). In
the following section, we want to learn about interfacial phenomena in which
three phases are involved.

6.1.5 Liquid–Liquid–Gas Interface

The shape of a liquid drop floating on the surface of another liquid is determined by
the interfacial tensions between all three phases. The phases involved are gas phase,
liquid A, and liquid B. In Fig. 6.9, such a drop shape is shown schematically. In the

Fig. 6.9 Interfacial tensions

at the contact point of three V 13
phases: 1 gas phase. 2 Liquid
A. 3 Liquid B V 12 M1 1
M3 3
M2
2
V 23
266 6 Interfaces

Fig. 6.10 Vectors of

interfacial tensions, shifted in V13 J V23
parallel to apply the cosine
theorem D E
E V12

equilibrium of forces, the tension σ 12 is just as great as the result of the vector
addition of forces σ 13 and σ 23. The contact angle φ3 takes the value at which this
equilibrium is realized.
With the help of the Cosine theorem for a isosceles triangle

c2 = a2 þ b2 - 2a
b
cos γ ð6:27Þ

can be formulated (Fig. 6.10):

σ 12 2 = σ 13 2 þ σ 23 2 - 2σ 13
σ 23
cos γ ð6:28Þ

because of

α þ β = φ3 and γ þ φ3 = 180∘ i:e: γ = 180∘ - φ3 ð6:29Þ

is
cos γ = - cos φ3 ð6:30Þ

therefore

σ 12 2 = σ 13 2 þ σ 23 2 þ 2σ 13
σ 23
cos φ3 ð6:31Þ

this results in the contact angle

σ 12 2 - ðσ 13 2 þ σ 23 2 Þ
cos φ3 = ð6:32Þ
2
σ 13
σ 23
Three cases can be distinguished in which the contact angle φ3 is above 90°,
below 90°, or near 0°. A contact angle above 90° leads to ellipsoid drops of phase
3 floating on phase 2. In an extreme case, the contact angle is 180°; then the floating
drops have spherical shape. In the other extreme case, the contact angle is 0°; then
phase 3 forms a thin film on phase 2. This is called spreading. By the use of
spreading, very thin films can be created down to the thickness of a single mono-
layer. In the other cases with contact angles between 0 and 90°, phase 3 forms
“lenses” that float on phase 3. Whether these lenses are more flat or more curved is
determined by the contact angle φ3, i.e., from the vectorial addition of the interfacial
tensions involved. Table 6.5 compiles these cases with some examples.
6.1 Interfacial Tension 267

Table 6.5 Cases with different contact angles

Case Cos φ3 φ3 Phase 3 forms... Example
I 1 0 Film on phase 2 Gasoline on water
Benzene on water
II Positive 0°–90° Lens on phase 2 Olive oil on water
Decane on water
III Negative 90°–180° Ellipsoid/sphere on phase 2 Water on silicone oil

Definition
Wetting is the formation of a small contact angle between a liquid and a solid
interface. Complete wetting is given at a contact angle of 0°. Then the liquid
forms a film on the solid surface. Complete non-wetting is given at a contact
angle of 180°.

The concept of wetting can also be understood from an energetic point of view.
The creation of an interface is associated with an energy expenditure. With three
phases involved, there are three types of interfaces that can be formed: solid–liquid,
solid–gaseous, and liquid–gaseous. Depending on the molecular structure of the
phases involved, the energy expenditure for each type is different. By varying the
areas of the three interfaces, the total interfacial energy of the system can be reduced
to a minimum value. This state of equilibrium can be characterized by the contact
angle; refer Fig. 6.13.

Definition
Complete wetting, i.e., a contact angle of 0° leads to the spreading of a liquid
on the subphase. The spreading pressure is the difference between the adhe-
sion energy and the cohesion energy. Substances with positive spreading
pressure are called film formers.

Liquids whose adhesion on another phase is associated with a greater energy gain
than the cohesion of the liquid will spread and form a film on the other phase. The
tendency toward spreading is characterized by the spreading pressure. With φ3 = 0
the vector σ 12 acts for adhesion and the vector sum (σ 13 + σ 23) for cohesion. The
difference between the two is the spreading pressure π:
π = σ 12 - ðσ 13 þ σ 23 Þ ð6:33Þ

If the spreading pressure π is positive, the liquid spreads on the subphase.

268 6 Interfaces

Example
A drop of a liquid hydrocarbon shall be placed on a water surface. Decane and
benzene are available. What will happen? Will the drop form a film on the
surface of the water or swim like a lens on the water?
Solution: The material data for the interfacial tensions at room temperature
are used to calculate the spreading pressure:

π = σ 12 - ðσ 13 þ σ 23 Þ

σ 12 air–water air–water
σ 12 = 72.4 mN
m-1 σ 12 = 72.4 mN
m-1
σ 13 Air–benzene Air–decane
σ 13 = 28.7 mN
m-1 σ 13 = 23.9 mN
m-1
σ 23 Water–benzene Water–decane
σ 23 = 33.7 mN
m-1 σ 23 = 51.2 mN
m-1

Spreading pressure of decane on water:

π=mN
m - 1 = 72:4 - ð23:9 þ 51:2Þ = - 2:7

Spreading pressure of benzene on water:

π=mN
m - 1 = 72:4 - ð28:7 þ 33:7Þ = 10:0

It turns out that the spreading pressure of benzene on water is positive; i.e.,
the benzene droplet will spread on the water surface. The drop decane, on the
other hand, will form a lens-like drop on the water surface.

The spreading pressure π is also called surface pressure. By analogy to the

equation of state for ideal gases

p
V =n
R
T ð6:34Þ

can be derived for surface pressure π

π
A=n
R
T ð6:35Þ

With a constant amount of substance n and a constant temperature T, there are

analogues to the Boyle–Mariotte law

1
p ð6:36Þ
V
6.1 Interfacial Tension 269

i.e.,

1
π ð6:37Þ
A
This means that the surface pressure of a film increases due to compression of the
film (reduction of the area A) and vice versa. With a so-called Langmuir trough, the
π - A isotherm of a film can be recorded with the help of a Langmuir balance. In this
way, aggregate states and polymorphic states of films can be characterized [19].
If we dissolve benzene from the above example in a high-molecular-weight
substance and give a drop on a water surface, then this substance can be distributed
with the help of the spreading of benzene to a monomolecular layer on the water
surface. After evaporation of the spreading liquid, a monolayer of this substance
remains on the water surface. This monolayer can be transferred to a solid substrate,
e.g., to a polymer or a glass. Such a film on a solid surface is called Langmuir–
Blodgett film. Film formation techniques like this make it possible to modify the
wetting properties or the biofunctionality of materials [20, 21].

6.1.6 Solid–Liquid–Gas Interface

We now replace the underlying liquid phase with a solid material. When a small
amount of a liquid is applied to this solid surface, the shape of the liquid is again
determined by the interfacial stresses involved (refer Fig. 6.11).
In equilibrium, Young‘s equation applies:

σ 12 = σ 23 þ σ 13
cos φ ð6:38Þ

i.e.,
σ 12 - σ 23 = σ 13
cos φ ð6:39Þ

This is called wetting tension σ B.

σ B = σ 12 - σ 23 = σ 13
cos φ ð6:40Þ

The wetting tension has a positive value if the adhesion of the liquid on the solid
surface is energetically more advantageous than the cohesion of the liquid. With a
negative wetting tension, it is reversed. In this case, the liquid lying on the solid tends

Fig. 6.11 Interfacial tensions

at the point of contact of three
phases. 1 gas, 2 solid, 3 liquid
270 6 Interfaces

to contract and forms drops. The cosine of the contact angle is positive in this case;
i.e., the contact angle is between 0° and 90°.
σ 12 - σ 23
cos φ = ð6:41Þ
σ 13
If we look again on Fig. 6.11 and take the arrows as vectors, we see that the
→
wetting tension is the horizontal component of the vector σ 13 , i.e.,

σ B = σ 13
cos φ: ð6:42Þ

In Table 6.6 are compared three cases, in which the contact angle φ is above 90°,
below 90°, or at 0°.
The wetting of a solid surface according to the listed cases I, II, and III is shown in
Fig. 6.12 schematically. The difference between a well-wetting liquid and a low
wetting liquid is illustrated in Fig. 6.13.

Table 6.6 Different contact angles on a solid surface

Case cos φ φ Consequence
I σ 12 - cos φ ≥ 1 =0 Complete wetting, creep of the
σ 23 ≥ σ 13 liquid, film forming
II 0 > σ 12 - 0 > cos 90° > φ > 180° Low or no wetting, liquid forms
σ 23 < σ 13 φ > -1 drops
III 0 < σ 12 - 0 < cos 90° > φ > 0° Good or partial wetting, liquid
σ 23 < σ 13 φ<1 forms flat drops

I II III

Fig. 6.12 Wetting of a solid surface: I film formation, II partial wetting, III roll off of a liquid

Fig. 6.13 Contact angle φ of a well wettable (I) and a poor wettable (II) solid surface. The contact
angle is below 90°. With a non-wettable solid surface, the contact angle is above 90° (III)
6.2 Kinetic Phenomena at Interfaces 271

6.2 Kinetic Phenomena at Interfaces

The extent to which an interface is occupied with surface-active substances can

depend on the speed of the processes involved. Examples of such processes are

• Formation or degradation processes of surface-active substances

• Molecular orientation of surface-active substances
• Formation or degradation of associates such as micelles or vesicles
• Adsorption or desorption at interfaces

In these cases, the interfacial occupancy depends on the time. Interfacial tension,
contact angle, and spreading pressure are functions of time.
To illustrate we can imagine the adsorption of proteins at an interface, e.g., the
adsorption of a milk protein at the interface to a fat droplet. For the speed of
adsorption, first the speed of mass transport to the interface must be determined.
This means that the kinetics of the interface stabilization depend on the diffusion rate
of the molecule. This in turn depends on molecular weight, viscosity of the solution,
and, thus, the temperature. In addition, the molecular alignment, i.e., the orientation
of the surface-active molecule, can play a role in adsorption. If the protein forms
micelles, its availability for interfacial adsorption depends on the balance between
speed of micelle formation and degradation. This equilibrium, as well as the speed of
equilibrium adjustment, can depend on factors such as pH or electrolyte content. The
interaction of all those processes leads to the kinetics of adsorption and thus to
kinetics of interface stabilization. As soon as there is an equilibrium between the
processes, the interface occupancy is constant. Such a state of equilibrium exists at a
given temperature for each concentration of the adsorbent. These equilibrium states
can be represented with adsorption isotherms, analogous to the isotherms presented
in Chap. 1. If a strong bond, such as chemisorption, is found between adsorbent and
adsorbate, these interfacial adsorption isotherms have the form of the Langmuir type
(see Chap. 1 on Water activity) [22].
Kinetic phenomena in interfacial stabilization play a role in the production of
emulsions, in which newly formed interfaces must be stabilized in such a short time
that reversing processes such as droplet coalescence are minimized. Here, it is
important to avoid delays in the kinetics of the interface occupation.
The reverse effect, the depletion of surface-active substances, is also called
negative adsorption. Here, an increase in interfacial tension is observed when the
concentration of surface-active substance is increased.

Definition
Dynamic interfacial tension: In contrast to an interfacial tension which is
constant over time (static interfacial tension), a time-varying interfacial tension
is referred to as dynamic interfacial tension.
272 6 Interfaces

6.3 Measurement

Measurement of Interfacial Tension

One of the methods used for measurement of interfacial (surface) tension is known
as the bow wire method. Figure 6.14 shows a schematic of how this method is
applied. A piece of wire of length l is lifted for a distance dh out of a liquid interface.
This causes the interfacial surface area to increase by dA = 2
l
dh.The force
F necessary for lifting the wire is measured and used to calculate the surface tension.
When using a flat plate instead of the wire, one speaks of the plate measuring method
according to Wilhelmy (Fig. 6.15).
In the ring method according to Du Noüy, a circular wire ring is used instead of a
wire bracket. When the ring is pulled out of the interface, the interface is increased
by the amount dA = 2
l
dh, in the case of the ring-shaped wire it is l = π
d. By
determining the necessary force F, the interfacial tension can be determined by
dE F
dh F
σ= = = : ð6:43Þ
dA cos φ
2
l
dh cos φ
2
π
d
The contact angle φ indicates the wetting of the ring by the measuring liquid.
With complete wetting of the ring, φ = 0, it is

dE F
dh F
σ= = = ð6:44Þ
dA 2
l
dh 2
π
d

Fig. 6.14 Schematic for 1

measuring surface tension F
with the bow wire.
1 Forcemeter, 2 bow wire, 2
3 horizontal piece of wire, 3
l
4 liquid sample 4
F h

Fig. 6.15 Wilhelmy plate for

measuring interfacial tension,
schematic I
z

M
6.3 Measurement 273

σ interfacial tension in N
m-1

E interface energy in Nm
m-2
A area in m2
F force in N
φ wetting angle
h height in m
l length in m

The value thus obtained is to be multiplied by a tabulated correction factor that

eliminates the effect of the volume of liquid hanging on the ring. The correction
factor according to Harkins and Jordan in turn depends on the density of the liquid
and the ring diameter [23].
The ring method is often used to determine the surface tension of aqueous liquids.
For a ring made of platinum-iridium, it is assumed that the liquid ideally wets the
wire or the plate. That is, we use the approximation that the contact angle between
liquid and wetted solid is φ = 0. For most aqueous solutions, this approximation is
well fulfilled. If we want to measure the interfacial tension between two liquids with
the ring method, we must not neglect buoyancy effects and wetting angles of the
fluids.
With the plate method the maximum force F is measured with a wetted plate that
is used to “raise” the underlying liquid. The interfacial tension is
dE F
dh
σ= = : ð6:45Þ
dA cosφ
2
ðz þ lÞ
dh

With φ = 0 it is

dE F
σ= = : ð6:46Þ
dA 2
ðz þ lÞ

Conversely, if we want to determine the contact angle between liquid and plate, a
liquid of known surface tension is used. In this way, the contact angle φ of a material
or a surface coating can be determined.

F
φ = arc cos ð6:47Þ
σ
2
ðz þ lÞ

Definition
The contact angle is the angle between liquid interface and solid surface. It is
also known as the wetting angle. When a solid material is completely wetted
by the liquid, the contact angle is 0°.
274 6 Interfaces

Fig. 6.16 Capillary method

rK
for determining the surface
tension
M

h
FG

In the capillary method, the height to which a liquid rises within a capillary tube is
measured and used to calculate the surface tension. The wetting of a capillary creates
a curved liquid surface (Fig. 6.16) with a pressure difference between the two sides
of that interface (see Table 6.2). The liquid column in the capillary rises until the
hydrostatic pressure of the column corresponds to the pressure above the curved
surface. The surface tension can be calculated from the pressure equilibrium.
The capillary pressure above the concave fluid meniscus is according to Laplace
(refer Sect. 6.1.1).

2
σ
cos φ
pK = : ð6:48Þ
rK
Here, φ is the contact angle, i.e., the angle between the liquid surface and the wall
of the capillary (see Fig. 6.16) and is compensated by the hydrostatic pressure of the
fluid column below the meniscus
p = Δρ
h
g, ð6:49Þ

where Δρ is the difference in density of liquid phase and gas phase.

Equalizing the pressures provides

2
σ
cos φ
= Δρ
h
g ð6:50Þ
rK
and thus for interfacial tension

Δρ
h
g
r K
σ= : ð6:51Þ
2
cos φ
With the approximations Δρ ≈ ρFl and φ = 0 which is well fulfilled in aqueous
systems in glass capillaries, the surface tension therefore is

ρFl
h
g
r K
σ= ð6:52Þ
2
6.3 Measurement 275

σ interfacial tension in N
m-1

ρFl density of the liquid in kg
m-3
g gravitational acceleration in m
s-2
rK capillary radius in m
φ wetting angle
h height in m
l length in m

Example
In a 1 mm capillary, a liquid rising height of 28 mm is determined for an
aqueous solution at room temperature. The surface tension of the liquid thus is

998kg
m - 3
28
10 - 3 m
9:81m
s - 2
0:5
10 - 3 m

σ=
2
σ = 68:5mN
m - 1

Example
Estimation: How high will an aqueous solution with a surface tension of
70 mN∙m-1.
rise in a capillary with D = 0.75 mm inner diameter at room temperature?
With

2
σ
h=
rK
ρ
g
it is

4
70
10 - 3 N
m - 1
h= = 38:1 mm
0:75
10 - 3 m
998 kg
m - 3
9:81 m
s - 2

For estimations like this we can use:

4
70
10 - 3 N
m - 1 280
10 - 4 m 28
10 - 3 m
h≈ = =
D=m
10 kg
m - 3
10 m
s - 2
3 D=m D=10 - 3 mm

(continued)
276 6 Interfaces

i.e.,

28 mm
h≈
D=mm

i.e., for further capillary diameters:

D/mm h/mm
0.5 56
1 28
2 14
5 5.6
10 2.8

Drop Shape Analysis

The measurement of the shape of droplets offers further possibilities for the deter-
mination of interfacial tension and contact angles.

Pendant Drop Method

Under the influence of gravity, a hanging drop does not have a spherical shape but an
ellipsoidal shape. The height-dependent hydrostatic pressure in the drop causes the
deviation from the spherical shape. With a microscopic drop contour analysis, the
surface tension of a liquid can be calculated in this way at known droplet volume and
known densities of the liquid and gaseous phase.
In balance of forces, the Laplace pressure in the hanging drop is as high as the
hydrostatic pressure of the droplet liquid

1 1
pσ = σ
þ =ρ
g
h ð6:53Þ
r1 r2

i.e.,
ρ
g
h
σ= ð6:54Þ
1
r1 þ r12

Sessile Drop Method

If the drop is located on a solid material, microscopic techniques can be used to
determine the contact angle between the liquid and the material. From the contact
angle, conclusions can be drawn about the interfacial tensions of the substances
involved (cf. Table 6.6).
6.3 Measurement 277

Fig. 6.17 Spinning drop 1 2 3

method, schematic.
1 temperature control jacket,2
2rz
rotating drop, 3 liquid in
rotating capillary Y

Fig. 6.18 Droplet formation

at the end of a capillary

Spinning Drop Method

If a liquid-filled capillary, in which a drop of a liquid that cannot be miscible with the
ambient phase, is brought into rotation, the shape of the drop changes depending on
the rotation frequency (see Fig. 6.17). With known density difference, the interfacial
tension between the two liquids can be determined from the video drop contour
analysis.

Drop Volume Techniques

The volume of liquid droplets escaping from a capillary depends on the interfacial
tension of the liquid (Fig. 6.18).
A drop leaves a capillary orifice when the weight force of the drop can overcome
the tension force [24], i.e.,

m
g = 2π
r
σ ð6:55Þ

At a known density of the drop liquid, the interfacial tension can be determined
from the drop radius:
2π
r
σ = m
g ð6:56Þ

2π
r
σ = V
ρ
g ð6:57Þ

V
ρ
g
σ= ð6:58Þ
2π
r
In order to improve the accuracy, a predetermined correction factor C is used. It is
to be determined by calibration with a substance of known interfacial tension.

V
ρ
g
σ= ð6:59Þ
2π
r
C
278 6 Interfaces

Fig. 6.19 Stalagmometer for

the determination of the
surface tension

Specially shaped glass capillaries for falling droplets, so-called stalagmometers

(Fig. 6.19), are used for the rapid determination of surface tension. A fixed volume of
the liquid to be examined is slowly dropped through the stalagmometer. By counting
the drops, the volume of the individual drop can be determined.

σ interfacial tension in N
m-1

ρ density of the droplet-forming liquid in kg
m-3
g gravitational acceleration in m
s-2
r drop radius in m
V drop volume in m3
V_ flow in m3
s-1
ΔN number of drops
Δt duration in s
C correction factor

Drop Volume Tensiometer

In a continuously operating drop volume tensiometer, a precision pump is used to
deliver a defined volume flow rate which leads to a continuous generation of falling
drops.
ΔV ΔN
V
V_ = = : ð6:60Þ
Δt Δt
By counting the drops per unit of time Δt, e.g., by means of a photodiode, the
volume V of a drop can be determined:
6.3 Measurement 279

V_
Δt
V= ð6:61Þ
ΔN
then we can calculate

V_
Δt
ρ
g
σ= ð6:62Þ
N
2π
r
C
Instead of droplets falling into air (measuring surface tension), droplets falling
into liquid (e.g., water droplets in oil) or rising drops (e.g., oil droplets in water) can
also be examined.
With a continuously operating drop volumeter, the so-called dynamic interfacial
tension can be measured. Increasing the flow of the pump leads to faster drop
generation. The time span between two escaping drops is the time that a drop has
spent hanging on the capillary. This is called life time of the interface or age of the
interface. By increasing the flow of the pump, it is possible to create droplets with
very short life time up to the time the interfacial tension is measured. In this way, it is
possible to determine whether the interfacial tension depends on the life time of the
droplets. If this is the case, the continuous droplet volume tensiometer can be used to
study the interfacial kinetics of the present system.

Bubble Point Tensiometer

With a bubble point tensiometer, the air pressure is measured that is necessary to
produce gas bubbles of defined size in a liquid. According to the Laplace’s equation,
the surface tension between liquid and gas can be calculated from the measured
internal bubble pressure.
Since the interfacial tension is temperature-dependent (Table 6.4) such
measurements must always be carried out under temperature-controlled conditions.

Dynamic Measurement of Interfacial Tension

The continuous droplet volume tensiometer for determining the time-resolved inter-
facial tension on droplets has been described in the previous section. Another way to
examine surfaces dynamically is an overflowing cylinder technique [25, 26]. With
the help of a pump, the liquid to be examined is allowed to climb up a cylinder until it
overflows. In the overflowing surface, the surface tension is measured, e.g., using
Wilhelmy technique. By increasing the flow of the pump, the overflow speed and
thus the surface pressure increase and the surface lifetime is reduced.
A Langmuir–Pockels trough (commonly: Langmuir trough) consists of a shallow
tub that is filled with the liquid to be examined. With the help of a movable barrier, a
film of surface-active substances can be compressed or expanded. During compres-
sion, the interfacial tension is continuously measured, which gives information about
occupancy densities and film states. The speed of the moving barrier can be changed,
which allows the study of kinetic effects [27–29].
Contact angle measurements can also be carried out dynamically. For this
purpose, the contact angle is measured during movement of a drop at an interface.
280 6 Interfaces

The direction of movement can be wetting or dewetting. Often a hysteresis is

observed between wetting and dewetting which can be used to characterize a solid
interface, e.g., its roughness.
Dynamic investigations of interfacial effects are also possible by the drop contour
analysis of lying or hanging droplets whose volume is not constant but pulsating.
With the pulsating drop method the interface experiences a continuously alternating
dilation and contraction. By varying the pulsation frequency, time-dependent effects
of surface-active substance on can be investigated.
The dynamic behavior of interfacial films can be described with variables such as
interfacial elasticity and viscosity. As shown in Chap. 4 both quantities can be
described together by using the complex interfacial elasticity. The complex modulus
of interfacial elasticity consists of a real part for interfacial elasticity and an imagi-
nary part for interfacial viscosity. In interfacial rheology, often the interfacial
viscoelasticity as a function of the pulsation frequency is investigated [30, 31].

Further Reading

Theme ref.
Espresso coffee foam: interfacial rheology [32]
Alginate capsules: interfacial elasticity [33]
Curcuma encapsulation by coacervation [34]
Double emulsion in gel droplets from Ca-alginate [35]
Egg white: dynamic interface measurement by pendant drop technique [36]
Emulsions with droplets <1 μm: influence of disperse phase [37]
Icecream: foam stability and interfacial properties [38, 39]
Casein micelles: nano vehicle for nutraceuticals [11]
Sucrose solutions in industry: viscosity and interfacial tension [40]
Emulsions: Interfacial versus bulk rheology [41]
Beer foam: surface rheology [42]
Whey proteins, W/W emulsions: interfacial and gel rheology [43]
Surfactant glycyrrhizin: surface tension and rheology [44]
Plasma chemistry: modification of interfacial properties [45]
Pickering emulsions: preparation and application [46–49]
W/O/W emulsion containing vitamin B [50]
Nanobubbles in food processing [51]

Summary
At interfaces materials have different properties than those within their sub-
stance. Specific interfacial energy depends on the adjacent materials in
two-phase and three-phase systems and on the curvature of the interface. In
this chapter, basic relationships are explained step by step and calculated in

(continued)
References 281

numerous examples. The relation to disperse systems such as emulsions and

foams is established. Concentration- and time-dependent effects, such as those
found in micelles and surface films, are treated. Basic measurement techniques
for interfacial tension are presented and explained. At the end of the chapter.
Application examples are listed that can be used for further studies and as
suggestions for own investigations

References
1. Lauth JG, Kowalczyk J (2016) Einführung in die Physik und Chemie der Grenzflächen und
Kolloide. Springer Spektrum, Heidelberg
2. Adamson AW, Gast AP (1997) Physical chemistry of surfaces. Wiley-VCH, Weinheim
3. Martin AN, Swarbrick J, Cammarata A (1980) Physikalische Pharmazie. Wissenschaftliche
Verlagsgesellschaft, Stuttgart
4. Ramsay W, Shields J (1916) Sci Rep Tohoku Imp Univ 4(I):373
5. Kohlrausch F (1996) Praktische Physik, vol 3. Teubner, Stuttgart. https://doi.org/10.1002/piuz.
19870180109
6. Kretzschmar H-J, Wagner W (2019) International steam tables - properties of water and steam
based on the industrial formulation IAPWS-IF97. Springer, Berlin. https://doi.org/10.1007/978-
3-662-53219-5
7. Maeda M, Takahara A, Kitano H, Yamaoka T, Miura Y (2019) Molecular soft-interface.
Science. https://doi.org/10.1007/978-4-431-56877-3
8. Hildebrandt A (2002) Physikochemische Charakterisierung von Gallensalz-Mischmizellen als
Grundlage für innovative Arzneistoffträgersysteme. Martin-Luther-University, Halle
9. Carey MC, Small DM (1970) The characteristics of mixed micellar solutions with particular
reference to bile. Am J Med 49(5):590. https://doi.org/10.1016/S0002-9343(70)80127-9
10. DIN EN 13955 Trübungsmessung (2003) Beuth, Berlin
11. Semo E, Kesselman E, Danino D, Livney YD (2007) Casein micelle as a natural nano-capsular
vehicle for nutraceuticals. Food Hydrocoll 21:936. https://doi.org/10.1016/j.foodhyd.2006.
09.006
12. Burauer S, Belkoura L, Stubenrauch C, Strey R (2003) Bicontinuous microemulsions revisited:
a new approach to freeze fracture electron microscopy (FFEM). Colloids Surf A Physicochem
Eng Asp 228(1):159. https://doi.org/10.1016/S0927-7757(03)00357-1
13. Gutiérrez JM, González C, Maestro A, Solè I, Pey CM, Nolla J (2008) Nano-emulsions: new
applications and optimization of their preparation. Curr Opin Colloid Interface Sci 13(4):245.
https://doi.org/10.1016/j.cocis.2008.01.005
14. McClements DJ (2012) Nanoemulsions versus microemulsions: terminology, differences, and
similarities. Soft Matter. https://doi.org/10.1039/c2sm06903b
15. Berton-Carabin CC, Schroen K (2015) Pickering emulsions for food applications: background,
trends, and challenges. Annu Rev Food Sci Technol 6:263. https://doi.org/10.1146/annurev-
food-081114-110822
16. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, Munich
17. Schuchmann HP, Schuchmann H (2005) Lebensmittelverfahrenstechnik. Wiley-VCH,
Weinheim. https://doi.org/10.1002/9783527623549
18. Schuchmann HP, Köhler K (eds) (2012) Emulgiertechnik. Behr’s, Hamburg
19. Dauphas S, Beaumal V, Riaublanc A, Anton M (2006) Hen egg yolk low-density lipoproteins
film spreading at the air–water and oil–water interfaces. J Agric Food Chem 54(10):3733.
https://doi.org/10.1021/jf053174e
282 6 Interfaces

20. Przykaza K, Woźniak K, Jurak M, Wiącek AE, Mroczka R (2019) Properties of the Langmuir
and Langmuir–Blodgett monolayers of cholesterol-cyclosporine A on water and polymer
support. Adsorption 25(4):923. https://doi.org/10.1007/s10450-019-00117-2
21. Furtado FAS, Caseli L (2020) Enzyme activity preservation for galactose oxidase immobilized
in stearic acid Langmuir-Blodgett films. Thin Solid Films 709:138253. https://doi.org/10.1016/
j.tsf.2020.138253
22. Azizian S, Eris S, Wilson LD (2018) Re-evaluation of the century-old Langmuir isotherm for
modeling adsorption phenomena in solution. Chem Phys 513:99. https://doi.org/10.1016/j.
chemphys.2018.06.022
23. Harkins WD, Jordan HF (1930) A method for the determination of surface and interfacial
tension from the maximum pull on a ring. J Am Chem Soc 52(5):1751. https://doi.org/10.1021/
ja01368a004
24. Tate T (1864) On the magnitude of a drop of liquid formed under different circumstances.
Philos Mag 27:176. https://doi.org/10.1080/14786446408643645
25. Bergink-Martens DJM (1993) Interface dilation: the overflowing cylinder technique. Dr Phil,
WUR, Wageningen
26. Bain CD (2008) The overflowing cylinder sixty years on. Adv Colloid Interf Sci 144(1):4.
https://doi.org/10.1016/j.cis.2008.08.006
27. Murray BS, Nelson PV (1996) A novel Langmuir trough for equilibrium and dynamic
measurements on air-water and oil-water monolayers. Langmuir 12(25):5973. https://doi.
org/10.1021/la960748o
28. Shen Y, Powell RL, Longo ML (2008) Interfacial and stability study of microbubbles coated
with a monostearin/monopalmitin-rich food emulsifier and PEG40 stearate. J Colloid Interface
Sci 321(1):186. https://doi.org/10.1016/j.jcis.2008.01.044
29. Murray BS, Durga K, Yusoff A, Stoyanov SD (2011) Stabilization of foams and emulsions by
mixtures of surface active food-grade particles and proteins. Food Hydrocoll 25(4):627. https://
doi.org/10.1016/j.foodhyd.2010.07.025
30. Ivanov IB, Danov KD, Ananthapadmanabhan KP, Lips A (2005) Interfacial rheology of
adsorbed layers with surface reaction: on the origin of the dilatational surface viscosity. Adv
Colloid Interf Sci 114-115:61. https://doi.org/10.1016/j.cis.2004.11.001
31. Wulff-Pérez M, Torcello-Gómez A, Martín-Rodríguez A, Gálvez-Ruiz MJ, de Vicente J (2011)
Bulk and interfacial viscoelasticity in concentrated emulsions: the role of the surfactant. Food
Hydrocoll 25(4):677. https://doi.org/10.1016/j.foodhyd.2010.08.012
32. Piazza L, Gigli J, Bulbarello A (2008) Interfacial rheology study of espresso coffee foam
structure and properties. J Food Eng 84(3):420. https://doi.org/10.1016/j.jfoodeng.2007.06.001
33. Martins E, Poncelet D, Marquis M, Davy J, Renard D (2017) Monodisperse core-shell alginate
(micro)-capsules with oil core generated from droplets millifluidic. Food Hydrocoll 63:447.
https://doi.org/10.1016/j.foodhyd.2016.09.018
34. Mohammadian M, Salami M, Alavi F, Momen S, Emam-Djomeh Z, Moosavi-Movahedi AA
(2019) Fabrication and characterization of curcumin-loaded complex coacervates made of gum
Arabic and whey protein nanofibrils. Food Biophys 14(4):425. https://doi.org/10.1007/s11483-
019-09591-1
35. Lee B-B, Bhandari BR, Ching SH, Howes T (2019) Improving hydrophilic barriers of
encapsulated compounds in ca-alginate microgel particles through a new ionotropic gelation
method for double emulsion droplets. Food Biophys 14(4):365. https://doi.org/10.1007/
s11483-019-09586-y
36. Dabestani M, Yeganehzad S, Krzan M, Miller R (2019) Characterisation of egg white adsorp-
tion layers under equilibrium and dynamic conditions. Colloids Surf A Physicochem Eng Asp
568:29. https://doi.org/10.1016/j.colsurfa.2019.01.066
37. Gao S, McClements DJ (2019) Influence of disperse phase transfer on properties of
nanoemulsions containing oil droplets with different compositions and physical states. Food
Biophys 14(4):355. https://doi.org/10.1007/s11483-019-09585-z
38. Goff HD, Hartel RW (2013) Ice cream. Springer, Berlin
References 283

39. Goff HD, Vega C (2007) In: McClements DJ (ed) Structure-engineering of ice-cream and foam-
based foods. Woodhead Publishing, Sawston, pp 557–574. https://doi.org/10.1533/
9781845693671.4.557
40. Schmidt T (2000) Viskositäts- und Oberflächenspannungsverhalten von reinen und technischen
Saccharoselösungen. Technical University, Berlin
41. Kamkar M, Bazazi P, Kannan A, Suja VC, Hejazi SH, Fuller GG, Sundararaj U (2020)
Polymeric-nanofluids stabilized emulsions: interfacial versus bulk rheology. J Colloid Interface
Sci 576:252. https://doi.org/10.1016/j.jcis.2020.04.105
42. Lu Y, Choi B, Nylander T, Bergenståhl B, Nilsson L (2020) Surface rheology and morphology
of beer protein and iso-humulone at air-liquid surface. Food Hydrocoll 108:105897. https://doi.
org/10.1016/j.foodhyd.2020.105897
43. Madadlou A, Famelart M-H, Pezennec S, Rousseau F, Floury J, Dupont D (2020) Interfacial
and (emulsion) gel rheology of hydrophobised whey proteins. Int Dairy J 100:104556. https://
doi.org/10.1016/j.idairyj.2019.104556
44. Ralla T, Salminen H, Braun K, Edelmann M, Dawid C, Hofmann T, Weiss J (2020)
Investigations into the structure-function relationship of plant-based surfactant glycyrrhizin:
interfacial behavior & emulsion formation. LWT 120:108910. https://doi.org/10.1016/j.lwt.
2019.108910
45. Jiang Y-H, Cheng J-H, Sun D-W (2020) Effects of plasma chemistry on the interfacial
performance of protein and polysaccharide in emulsion. Trends Food Sci Technol 98:129.
https://doi.org/10.1016/j.tifs.2020.02.009
46. Murray BS (2019) Pickering emulsions for food and drinks. Curr Opin Food Sci 27:57. https://
doi.org/10.1016/j.cofs.2019.05.004
47. Jiang H, Sheng Y, Ngai T (2020) Pickering emulsions: versatility of colloidal particles and
recent applications. Curr Opin Colloid Interface Sci 49:1. https://doi.org/10.1016/j.cocis.2020.
04.010
48. Dickinson E (2020) Advances in food emulsions and foams: reflections on research in the
neo-Pickering era. Curr Opin Food Sci 33:52. https://doi.org/10.1016/j.cofs.2019.12.009
49. Zafeiri I, Wolf B (2019) Sustainable Pickering emulsions. Food. Sci Technol 33(3):20. https://
doi.org/10.1002/fsat.3303_6.x
50. Zaghian N, Goli M (2020) Optimization of the production conditions of primary (W1/O) and
double (W1/O/W2) nano-emulsions containing vitamin B12 in skim milk using ultrasound
wave by response surface methodology. J Food Meas Charac 14(6):3216. https://doi.org/10.
1007/s11694-020-00567-1
51. Thi Phan KK, Truong T, Wang Y, Bhandari B (2020) Nanobubbles: fundamental characteristics
and applications in food processing. Trends Food Sci Technol 95:118. https://doi.org/10.1016/j.
tifs.2019.11.019
Permeability
7

Transport of substance, mass, heat, electric charge, or other forms of mass or energy
across a porous membrane is known as permeability. To answer questions about the
permeation of substances through packaging materials, for heat transport through
solid bodies or fluids, and for the transport of electric current in solid substances, the
so-called transport variables are required. A distinction is made between transport
variables that already take into account the geometry of the given system and
geometry-independent transport variables. The geometry-independent transport
variables are material properties.
For the transport of mass particles or molecules in stationary fields, analogous
laws apply as for the transport of heat, phonons (quantized waves) or electric charge
carriers, e.g., ions or electrons. We can treat mass transport by diffusion with
analogous equations such as heat conduction or the conduction of electric current.
By applying electro-analog rules such as the addition of resistors (see Sect. 7.3), the
calculation of steady state transport processes becomes very simple. In this way,
estimates of mass transport through multi-layer packaging films can be carried out
with little computational effort.

7.1 Steady State Diffusion in Solids

Mass transport will first be considered using an example: A packaging film is

intended to prevent the loss of a gaseous flavoring substance from a product. The
film thus separates the package with a high concentration of the flavoring substance
(potential φ is high) from the environment in which the concentration of the
flavoring substance is low (potential φ is low). How much of the flavoring substance
diffuses through the film (see Fig. 7.1) depends on:

• how permeable the film is to this flavoring substance (material property)

• how great is the potential difference (gradient) in concentrations on opposite sides
of the membrane (the “driving force”) Δφ

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 285
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_7
286 7 Permeability

• how thick the film is (geometric size)

• how much film area is available (geometric size)

The quotient dφ
dx in Fig. 7.1 is called potential gradient.
If this gradient is constant over time, a constant flow (quantity of diffusing
substance per unit of time) occurs as a result of stationary diffusion through the
film. The general transport equation for steady state is:
quantity flow = - coefficient
gradient

so
quantity potential difference
= - coefficient

area
time length

shorter
1 dM dφ

= -K
ð7:1Þ
A dt dx

A area
M quantity
t time
φ potential
x length
K coefficient

The term “quantity” is a generic term for types of quantities such as mass, amount
of substance, volume or even amount of heat or electric charge or number of
particles. Depending on the quantity type used, different terms are used for the
flow, see Tables 7.1 and 7.2. Depending on the application, the potential difference
has the meaning of concentration difference, temperature difference, or electrical
voltage. Equation (7.1) shows the capability of the general transport equation.
The negative sign in the general transport equation ensures that a flux in direction
of x has a positive sign. Such a flux with a positive sign occurs when the potential
decreases in direction of x, i.e. when the gradient has a negative sign (cf. Fig. 7.1).

Fig. 7.1 Curve of the

potential φ across a thin film
with thickness d. As a result of
the potential gradient mass
flow M occurs
7.1 Steady State Diffusion in Solids 287

Table 7.1 Terms in trans- Quantity Transport flux is ...

port processes
Moles of material molar flow rate in mol
m2
s-1
Mass mass flow rate in kg
m-2
s-1
Volume volume flow rate in m3
m-2
s-1
Heat heat flow rate in J
m-2
s-1
Electric charge electric current in C
m-2
s-1

Table 7.2 Forms of the general transport equation with their associated coefficients and units
Flow Caused by gradient of ... Coefficient is called
Mass flow concentration in kg
m-3 diffusion coefficient in m2
s-1
Volume flow concentration in m3
m-3 diffusion coefficient in m2
s-1
Substance flow concentration in mol
m-3 diffusion coefficient in m2
s-1
Mass flow partial pressure in Pa permeability coefficient in kg
s-1
m-1
Pa-1
Volume flow partial pressure in Pa permeability coefficient in m2
s-1
Pa-1
(standard volume)
Substance flow partial pressure in Pa permeability coefficient in mol
s-1
m-1
Pa-1

In many cases, only the absolute amount of the flux is of interest only, since the
direction is clearly known. Then, the transport equation in amount notation is used.
Using the first quantity given in Table 7.1, we can write the transport of a
substance, e.g., a flavoring substance through a film as a molar flow rate density
that occurs as a result of a concentration gradient. The transport equation thus is
amount of substance
- = coefficient
concentration gradient
area
time
so
1 dn dc
-
=D
ð7:2Þ
A dt dx
This form of the transport equation is also called first Fick’s law for steady state
diffusion. The coefficient D is called diffusion coefficient.

A area in m2
n amount of substance in mol
t time in s
c concentration in mol
m-3
x length in m
D diffusion coefficient in m2
s-1

Depending on the choice of concentration units (mol
m-3, l-1, g∙cm-3, ppm o.ä.),
there are corresponding other units for D, which can be converted into each other.
288 7 Permeability

If the quantity is not the amount of substance but the mass (second line in
Table 7.1), the transport equation provides the mass flow rate through the film. If
the partial pressure is used instead of the concentration, which is useful for gaseous
substances, the transport equation has the form
mass
- = K
partial pressure difference
area
time
so

1 dm dp
-
=P
ð7:3Þ
A dt dx

A area in m2
n amount of substance in mol
t time in s
p partial pressure in Pa
x length in m
P coefficient of permeability in kg∙s-1∙m-1∙Pa-1

The coefficient P is called coefficient of permeability and is used as an indicator

of the permeability of a material. The unit of P depends on which unit is used for the
partial pressure (Pa, bar, atm, psi, etc.).
If the volume of the diffusing flavoring substance is chosen instead of the mass
(third line Table 7.1), the transport equation takes the form

volume
- = K
difference of partial pressures
area
time
i.e.,

1 dV dp
-
=P
ð7:4Þ
A dt dx

A area in m2
V volume in m3
t time in s
p partial pressure in Pa
x length in m
P permeability coefficient in m2
s-1∙Pa-1

We see that the permeability coefficient now includes a volume in m3. Since the
volume of a gas depends on external conditions of pressure and temperature, it is
advantageous to use a so-called standard volume measured under specified standard
conditions of atmospheric pressure and temperature. After DIN 1343 [1] the standard
volume of a gas is the volume of the gas at 0 °C and 1013 hPa (Table 7.6).
7.2 Conductivity, Conductance, Resistance 289

Definition
Permeation is the transport of a substance through a solid material. The
permeability of a material is quantified by the permeability coefficient for a
particular permeate, e.g., for gaseous oxygen. While in permeation through a
solid material the diffusing substance comes from the environment, in migra-
tion the diffusing substance comes from the material itself.

7.2 Conductivity, Conductance, Resistance

The coefficient K of the general transport equation has the function of a specific
conductivity. “Specific” in this sense means the conductivity is specific for the
material in which the transport is observed. The coefficient does not depend on the
geometry of the system (A and d). The quantity which integrates the geometric sizes
A and d and conductivity is called conductance G. The reciprocal of G is the
resistance R. Writing Eq. (7.1) with the material thickness d we get
1 _ Δφ
- M=K
ð7:5Þ
A d
with

A
G=K
ð7:6Þ
d
the flow rate is simply
_ = G
Δφ
-M ð7:7Þ

in words
- flow = conductance
potential difference

or because of G = 1
R

_ = 1
Δφ
-M ð7:8Þ
R

M_ flow rate
K coefficient
A area
d thickness
φ potential
G conductance
R resistance
290 7 Permeability

The conductance G and its reciprocal value, the resistance R, include both the
material properties and the geometry of the present system.
The coefficient K as specific conductivity, on the other hand, contains no geome-
try information, so it is a pure material property. The reciprocal size of K is also
called resistivity and also contains no geometry information. Depending on the
application, these quantities are further specified. Table 7.7 summarizes those
designations.

7.3 Steady State Transport Through Solid Multi-layers

Consideration of resistance or conductance values simplifies the calculation of

transport processes for more complicated systems. Transport processes through
solids consisting of several layers can be treated in this way as transport processes
through several resistors. A distinction must be made here as to whether the resistors
must be overcome one after the other (series connection) or simultaneously (parallel
connection). Figure 7.2 illustrates the difference between a flow-through serial
resistors and parallel resistors. For the resistors we can use symbols from electrical

1
1 2 3
2

Mengenstrom Mengenstrom

Three materials in series Three materials in parallel

analogue to electric circuits:

serial connection parallel connection
R1

R1 R2 R3
R2

R3

Fig. 7.2 Flow-through serial resistors or parallel resistors

7.3 Steady State Transport Through Solid Multi-layers 291

engineering and design electric circuit diagrams for the present transport problem
(see Fig. 7.2 bottom).
Regardless of whether it is material flow, mass flow, volume flow, heat flow, or
electrical current, total resistance can be calculated in the same way: In the case of
series connection, the individual resistances are simply added together and the total
resistance is obtained. In the case of parallel connection, the reciprocal total resis-
tance is obtained (this is the total conductance) by adding the individual conductance
values R1 .
n
at serial connection : Rtotal = Ri
i=1

n
1 1
at parallel connection : =
Rtotal i=1
Ri

Packaging films are often produced from several layers to obtain the desired
barrier properties. Calculation of a permeation current through such a composite film
of different thicknesses can be carried out in a simple way with the concept of
additive individual resistors (see examples). It is important to recognize whether the
resistors are connected in series or parallel. More complicated systems can also
consist of resistors, some of which are connected in series and some in parallel. Then
the advantage of the electro-analog model becomes clear, because the system
remains manageable.
Resistances and conductances contain the geometric dimensions A and d of the
present system. Therefore, a simple addition of resistance and conductance values is
possible even if A and d are not the same size for every resistor.
However, if A and/or d are the same size for all individual resistors, the calcula-
tion can be further simplified. Then it is possible to calculate the total resistance by
adding only the coefficients or the reciprocal coefficients (Table 7.3):
Below we will see two calculation examples in which we come back to this
simplification.

Table 7.3 Calculation of Serial connection Parallel connection

the total resistance at simple n n
geometry Rtotal = Ri 1
Rtotal = 1
Ri
i=1 i=1
because of
R = Ad K1 1
R = Ad K
Rtotal = di
Ai
1
Ki
1
Rtotal = Ai
di Ki
i i
with A1 = A2 = A3...
Rtotal = 1
A
di
Ki
1
Rtotal =A Ki
di
i i
with also d1 = d2 = d3...
Rtotal = Ad K1i 1
Rtotal = A
d Ki
i i
292 7 Permeability

7.4 Food Packaging

One of the original purposes of food technology is to achieve a certain shelf life of
food. This refers to ensuring public health safety by preventing the possibility of any
spoilage-related food poisoning, even after a period of storage. Once safety is
assured, a second purpose is to maintain quality until the time of consumption.
The purpose of food packaging is therefore to support the stabilization of the food
against loss of quality and spoilage. In addition, there are other purposes of packag-
ing such as mechanical transport protection, the formation of prepackages, carriers of
legal information or manufacturer’s instructions, etc.

Definition
Shelf life is usually the period of time between the production and consump-
tion of a food during which no unacceptable quality losses occur [2]. The
specified shelf life applies only to foods maintained under proper storage and
handling conditions. For example, proper storage and handling for perishable
foods may require keeping them in cold storage or under refrigeration at all
times until purchased by the consumer. Many foods have a significantly lower
shelf life without suitable packaging, think of liver sausage, beer, or olive oil.

In order to minimize the loss of quality of food, packaging can be used to control
the following influences:

1. Incidence of light
2. Invasion of foreign microorganisms or parasites
3. Absorption of gaseous substances from the environment (water vapor, oxygen,
foreign odors)
4. Loss of gaseous substances (flavorings, water vapor, oxygen, nitrogen, carbon
dioxide)

The last two points can be achieved by selecting packaging materials that have the
desired permeation properties. It is therefore a matter of choosing the right materials
and thickness in order to achieve the desired shelf life durability (e.g., days, weeks,
months, or years) [3, 4]. A package can consist of several packaging materials
[5, 6]. Depending on the food, packaging materials with specified permeability are
required, e.g., for O2, CO2, H2O. Packaging materials are often combined in the form
of multi-layer plastic films or composite materials. In this way, for example, a low
permeability for O2 is obtained by a thin special material and mechanical stability or
shadowing of light by the other material. In Fig. 7.3 the flow of a permeate through a
film is shown schematically.
In steady state ( p1, p2 = const.) the permeability coefficient P can be expressed
via the mass flux of the permeate:
7.4 Food Packaging 293

Fig. 7.3 Permeation through

a film, schematic A p1
d ∆p
p2
dm
dt

1 dm Δp

=P
ð7:9Þ
A dt d

m_
d
P= ð7:10Þ
A
Δp

A area in m2
m mass in kg
t time in s
P permeability coefficient in kg
s-1
m-1
Pa-1
Δp partial pressure difference in Pa
d thickness in m

or via the volume flux density:

1 dV Δp

=P
ð7:11Þ
A dt d

V_
d
P= ð7:12Þ
A
Δp

V volume in m3
P permeability coefficient in m2
s-1
Pa-1

or via the molar flux density:

1 dn Δp

=P
ð7:13Þ
A dt d

n_
d
P= ð7:14Þ
A
Δp

n amount of substance in mol

P permeability coefficient in mol
s-1
m-1
Pa-1
294 7 Permeability

Example
Calculating the oxygen permeability into an initially oxygen-free tubular bag
(film surface 14 cm2, thickness 20 μm) permeates 0.2 cm3 O2 within 20 h. The
oxygen concentration of the air in the vicinity of the tubular bag is 20.9% (V/V),
temperature 25 °C, pressure 1020 hPa. The permeability P of the packaging
material shall be calculated.
Solution:
Calculating the standard volume of oxygen using the law for ideal
gases:

p1 V 1 p2 V 2 pSTP V STP
= =
T1 T2 T STP
p1 V 1
T STP T p 273:5 1020
V O2 ,STP = = V 1
STP
1 = 0:2 cm3


T 1
pSTP T 1 pSTP 298:15 1013:15
= 0:184 cm3 = 1:84
10 - 7 m3

Calculating the partial pressure difference:

pO2 ,outer = ptotal
xO2 = 1020 hPa
0:209 = 213:18 hPa

Δp = 213:18 hPa - 0 hPa = 213:18 hPa

Calculating the permeability coefficient:

1 dV Δp

=P

A dt d
_V d
P=

A Δp
1:84
10 - 7 m3
20
10 - 6 m 1:84
10 - 12 m3
m
P= -4 2
=
14
10 m
213:18 hPa
20
3600 s 1:4
m2
2:1318 Pa
3:6 s
105
1:84
10 - 17 m3
m - 18 m
m
3
= = 1:71
10
1:4
2:1318
3:6
m2
Pa
s m2
Pa
s
cm3
cm m3
m
= 1:71
10 - 13 = 1:71
10 - 9 2
cm
bar
s
2 m
bar
s
cm 3

cm
P = 1:48
10 - 4 2
cm
bar
day
7.4 Food Packaging 295

Example
Rough estimation of the permeability of the same bag
Omitting standard conditions, taking the oxygen partial pressure difference
to 0.2 bar and rounding 20 h to 1 day we get

_
jVj d 0:2 cm3
0, 002 cm -4 cm3
cm
P=
≈ = 1:4
10
A Δp 14 cm2
0:2 bar
1 day cm2
bar
day

7.4.1 Composite Films

Composite films are multi-layer films with each layer made of a different material
having different permeability in order to control several transport phenomena at the
same time, such as gas transmission, moisture transmission, light transmission, etc.
Calculating total combined permeability of such a multi-layer film is easy with the
concept of additive resistors (refer Fig. 7.2).

Example
The permeability of a composite film against CO2 is to be calculated. It
consists of:

film 1: 40 μm, P1 = 2
105cm3
μm
m-2
bar

film 2: 10 μm P2 = 5.65cm3
μm
m‐2
bar.

Composite film: 40 μm film 1 and 10 μm film 2

Solution:
We make a sketch showing that the films are permeated one after the other,
analogous to a serial connection of resistors. To get the total resistance per
square meter, we add the individual resistances over an area of 1 m2.
From

(continued)
296 7 Permeability

1 di
Rtotal =
A i
Ki

we get

2
1 d 1 di 1 d1 d2
Rtotal =
= =
þ
A P ges
A i=1
Pi A P1 P2

d 40 10 m2
bar m2
bar
= þ = 1:77
P total 2
10 5 5:65 cm 3 cm3

From this, total permeability can be calculated

dtotal ð40 þ 10Þμm

Ptotal = = = 28:2 cm3
μm
m - 2
bar - 1
1:77 m2
bar 1:77 m2
bar
If we compare the permeability of the composite film with that of film 1, we
can see that the permeability of the composite film is lower by a factor of 7000:

Ptotal 28:2 cm3
μm
m - 2
bar - 1 1

= = 14
10 - 3 =
P1 2
105 cm3
μm
m - 2
bar - 1 7092

Example
Consider a bag made of packaging material intended to prevent transmission of
water vapor into the bag. The bag consists of an aluminum-coated film (film 1:
area 400 cm2, thickness 20 μm, P = 2 cm3
μm
cm-2
bar-1), and a transparent
film (film 2: area 80 cm2, thickness 50 μm, P = 40 cm3
μm
cm-2
bar-1),
which serves as a viewing window.
Solution:
We make a sketch.

(continued)
7.4 Food Packaging 297

This shows that the films are permeated at the same time, i.e., it is a parallel
connection of resistors. To obtain the total resistance, we add the conductance
values, as such

2
1 Ai
K i Ai
Pi Ai
Pi
= = = =A
Rtotal i
di i
di i=1
di

A1
P1 A2
P2 Ai
Pi

þ =
d1 d2 di total

Ai
Pi A1
P1 A2
P2 400
2 80
40 cm3
= þ = þ
di total
d1 d2 20 80 bar
cm3 cm3
= ð20
2 þ 40Þ = 80
bar bar

Atotal cm3
Ptotal
= 80
dtotal bar
dtotal cm3
Ptotal =
80
Atotal bar
ð20μm þ 50μmÞ cm3 70 cm3
μm cm3
μm
=
80 =
80 = 11:7
ð400 cm2 þ 80 cmÞ bar 480 bar
cm2 bar
cm2

cm3
μm
Ptotal = 11:7
bar
cm2
Comparing the value of Ptotal with the permeability coefficient of film 1, it
turns out that the total permeability coefficient is about a factor of 6 higher than
the value of film 1.

Ptotal 11:7 cm3
μm
cm - 2
bar - 1

= ≈6
P1 2
cm3
μm
cm - 2
bar - 1
This increased permeability is a consequence of the viewing window used.
298 7 Permeability

7.5 Permeation as a Molecular Process

On a molecular level, permeation through a solid material is a subsequent process of

• Transport of the permeate from the gas phase to the interface of the material
• Adsorption of the permeate at the interface of the material
• Solubilization of the permeate in the material
• Diffusion of the permeate in the material in the direction according to partial
pressure gradient
• De-solubilization of the permeate from the opposite interface
• Adsorption of the permeate at the interface
• Removal (desorption) from the interface into the gas phase

For this reason, both the diffusion coefficient D of the permeate in the solid and its
solubility S in this material influence the permeability coefficient P. It is:
P=D
S ð7:15Þ

When looking for suitable barrier materials, this connection must be taken into
account. Biopolymers are often much more hydrophilic than conventional plastics.
Due to the solubility of the permeates in the biopolymer, this has consequences for
the permeation properties of the material. Often there is a lack of reference values for
the solubility of a gaseous substance in a solid. For low pressures, the solubility S can
be estimated using Henry’s law:

c
S= ð7:16Þ
p
Here concentration c is calculated as a quotient of the gas volume (standard
volume) and the volume of the solid phase.

P permeability coefficient in m2
s-1
Pa-1

D diffusion coefficient in m2
s-1
3
S solubility in mm3
Pa
3
c concentration of permeate in solid material in m
m3
p partial pressure of permeate in Pa

To determine the permeability coefficient of water vapor, the use of mass flow is
advantageous. This is especially advantageous if the permeability is determined by a
gravimetric measurement method (refer 7.7). For permanent gases, on the other
hand, the use of volume flow is common. Here, for better comparability the
determined gas volumes are given as standard gas volumes [1].
7.6 Temperature Dependency 299

→ Attention
There are several conventions for the standard state of gases that differ
slightly. It is therefore advisable to specify the standard used:

Standard Pressure Temperature

DIN 1343 1013.25 hPa 273.15 K = 0 °C
DIN ISO 2533 1013.25 hPa 288.15 K = 5 °C
DIN EN ISO 13443 1013.25 hPa 288.15 K = 15 °C
ISO 6358 and ISO 8778 1.0 bar 293.15 K = 20 °C

7.6 Temperature Dependency

Both solubilization and diffusion are temperature-dependent processes. In the sim-

plest case, their temperature dependence can be described with the Arrhenius
approach. Since permeation is based on the diffusion and solubilization of the
permeate, the temperature dependence of permeability can also be described by [2]:
EP
P = P0
e - RT ð7:17Þ
ED
D = D0
e - RT ð7:18Þ

Δsol H
S = S0
e - RT ð7:19Þ

with

EP = E D þ Δsol H ð7:20Þ

Ep activation energy of permeation in J
mol-1

ED activation energy of diffusion in J
mol-1
ΔsolH solution enthalpy in J
mol-1

If we know the permeability at, e.g., 20 °C and require the permeability coeffi-
cient at, e.g., 6 °C, the Arrhenius approach provides the following calculation option:

PðT 2 Þ E 1 1
ln =- P - ð7:21Þ
PðT 1 Þ R T2 T1

If the value of Ep is not available, we can estimate it from results of experimental

measurements taken at different temperatures. Experimentally determined perme-
ability values P are plotted as the logarithm of P against the reciprocal absolute
temperature T (the so-called Arrhenius plot, Fig. 7.4). If there is a linear relationship,
300 7 Permeability

Fig. 7.4 Temperature

dependence of permeability—
Arrhenius plot

lnP

1
T

one can assume the validity of the Arrhenius approach and use the slope m of the
graph to calculate the value of Ep [7].
EP
m= - ð7:22Þ
R

m
EP = - ð7:23Þ
R

7.7 Measurement of Permeability

A distinction must be made between steady state and transient measuring methods.
In case of steady state the pressure gradient is constant over time, i.e., the difference
between the partial pressures on the outside and the inside of the material to be
examined is constant, but in the transient case, the pressure gradient changes over
time. In both cases, the increase or decrease in the amount of permeating gas is
measured.
Solid films can easily be characterized with a so-called permeation cell. The film
is clamped into the cell in such a way that the film divides the cell into two air-tight
chambers, one on either side of the film. One chamber is filled with the test gas (this
is the permeating gas), while the other chamber contains the gaseous atmosphere into
which the test gas will permeate. The test gas will begin to permeate into the opposite
chamber and increase over time, while being measured and recorded. Depending on
the device and permeate, pressure sensors, gas sensors, gas chromatographs, or
sorbents are used [8].
When measuring permeability using manometric technique [9, 10] one of the two
gas rooms is evacuated and the differential pressure between the two gas rooms is
recorded. The permeability coefficient of the film material can be determined from
the temporal drop in the differential pressure due to permeation of the test gas.
A simple method of determining water vapor permeability is performed with the
help of a water vapor-absorbing substance in the permeation cell. Such a sorbent
(refer Chap. 1) is placed at the bottom of the permeation cell and the cell is sealed
with the examining film. The sealed permeation cup is placed in a chamber with
7.7 Measurement of Permeability 301

Fig. 7.5 Gravimetric

determination of the water 5 1
vapor permeability of a film,
schematic. 1. Climatic
chamber, 2. cup, 3. balance,
4. desiccant, 5. film 2
4
3

defined humidity (climatic chamber). Figure 7.5 shows the experimental setup
schematically.
The film now separates the water vapor-containing outdoor atmosphere and the
water vapor-free air in the cup. As a result of the partial pressure gradient, water
vapor permeates through the film. The water molecules entering the cup are absorbed
by the sorbent. The temporal increase in the mass of the sorbate is a measure of the
water vapor permeability of the film. The mass increase is tracked and recorded by
weighing. The water vapor permeability can be given in the unit g
m-2
day-1
[11]. To ensure steady state conditions, the water vapor partial pressure and temper-
ature on both sides of the film must be kept constant during the test. A constant water
vapor partial pressure pH 2O
D can be achieved by creating an atmosphere with constant
relative humidity, e.g., with the help of a climate chamber. Circular samples of the
film are used, e.g., circular foil punching pieces with a diameter of 90 mm. As
sorbents CaCl2 or silica gel (grain size preferably 1.5–2 mm) can be used.
In order to increase the comparability of measurements from different
laboratories, standard climates are recommended, e.g., temperature 23 °C and
humidity 50% r.h. [12]. Defined humidity also prevails in air, which is in equilibrium
with a saturated salt solution (ref. Chap. 1). If there is no automatic climate chamber,
the permeation cell can be placed in a vessel together with a saturated saline solution.
If, for example, saturated KCl solution is used at 20 °C, the humidity after setting the
equilibrium is 85% r.h., with a saturated NaCl solution it is 76% r.h. [13].
This gravimetric permeation method can also be applied to entire packages. Then
we do without the permeation cup and enclose the sorbent in the packaging, e.g., in a
bag. Then we place the bag in the desired climate and record its mass increase
(Table 7.4).
The m-t curve must be recorded over a sufficiently long period of time. Only
when at least three measuring points are on a straight line, one can assume steady
state conditions. From the slope of the m-t curve, the mass increase in g
day-1 is
determined. The quotient of the mass flow determined in this way and the film
surface is the water vapor permeability W with the unit g
m-2
day-1.
1 Δm
W= ð7:24Þ
A Δt
302 7 Permeability

Table 7.4 Climate ϑ (°C) φ (% r.h.) Code Ref.

standards
50 23 23/50 [14]
20 65 20/65 [14]
27 65 27/65 [14]
25 75 C [15]
20 85 E [15]
23 85 D [15]
25 90 A [15]
38 90 B [15]

Example
Gravimetric determination of the water vapor permeability of a household film
(thickness 10 μm). The circular foil sample has a diameter of 1 in. The
measurement is carried out at 25 °C and a relative humidity of 66% r.h.
Solution:

A = π
r 2 = π
ð12:7 mmÞ2 = 5:1
10 - 4 m2

d = 10μm

ϑ = 25 ° C

φinside = 0

φoutside = 66%r:h:

Measured data (see Fig. 7.6):

Time stamp t (h) m (g)

05.10.22 15:56 0 78.5908
06.10.22 08:10 16.233 78.6063
06.10.22 15:20 23.400 78.6155
09.10.22 08:45 88.817 78.6841
09.10.22 16:25 96.483 78.6928
10.10.22 08:25 112.483 78.7092
11.10.00 11:15 139.317 78.7370

mass flow from the slope, refer Fig. 7.6:

Δm
m_ = = 3
10 - 10 kg
s - 1
Δt

1 Δm 1
W= =
3
10 - 7 g
s - 1 = 50:8 g
day - 1
m - 2
A Δt 5:1
10 - 4 m2
7.7 Measurement of Permeability 303

Fig. 7.6 Mass increase due

78,74
to water vapor permeation
78,72
78,70
78,68 ∆m
m/g 78,66
78,64 ∆t
78,62
78,60

0,00 50,00 100,00 150,00

t/h

Example
Calculating the water vapor permeability of the same film
Solution:
First, we calculate the water vapor partial pressures:

ps ð25 ° CÞ = 31:666 hPa

p = φ
ps

poutside = 0:66
31:666 hPa = 2089:6 Pa

pinside = 0
31:666 hPa = 0 Pa

difference:

Δp = poutside - pinside = 2089:6 Pa - 0 Pa = 2089:6 Pa

Calculation of the permeability coefficient:

Δm d
P=

Δt A
Δp
10
10 - 6 m
P = 3
10 - 10 kg
s - 1
2
π
12:7
10 - 3 m2
2089:6 Pa
P = 2:8
10 - 15 kg
m
m - 2
s - 1
Pa - 1

other ways to express the result are listed in Table 7.5.

Measurements with the help of permeation cells require circular samples of the
packaging materials to be examined. Permeation measurements can also be carried
out for complete packaging. Thus, the influence of closures or sealing seams on the
gas permeability can also be investigated. While the permeability of metals and glass
304 7 Permeability

Table 7.5 Different ways of indicating water vapor permeability

Formula For the example above
Based on the mass flow Δm
Δt =
3
10-10kg
s-1 = 26 mg
day-1
Quotient of mass flow and area 1 Δm
A
Δt = 50.8 g
m-2
day-1 = 5.1 mg
cm-
2

day-1
Consideration of the thickness of the 1
A
Δm
Δt
d =
50.7 mg
μm
cm-2
day-1
film
Consideration of the pressure 1
A
Δm
Δt
Δp =
d
2:8
10 - 15 kg
m
m - 2
s - 1
Pa - 1
difference or
2:4 g
μm
cm - 2
day - 1
bar - 1

is negligible, the permeability of films made of polymers or biopolymers and

composite materials is of particular interest. In addition, the influences of technical
design such as sealants, closure techniques, modification of coatings, film thickness,
sealing and adhesive connections, etc. are of interest [16].

7.8 Analogous Transport Phenomena: Heat and Electricity

After applying the general transport equation for steady state gas diffusion, let’s turn
briefly to heat conduction and conduction of electric current. Both heat in solids and
electrons in electrical conductors move according to the laws of diffusion and can
therefore be described with the general transport equation.
In the case of steady state heat conduction, the term quantity is to be replaced with
heat and the gradient with temperature gradient. So the transport equation is
heat
- = coefficient
temperature gradient
area
time
or

1 dQ dT
-
=λ
ð7:25Þ
A dt dx
In this form, the transport equation is identical to Fourier’s fist law. The coeffi-
cient is the thermal conductivity coefficient λ of the solid under consideration. It is
called thermal conductivity for short.
When considering an electric current, the term quantity is to be replaced by
electric charge. The gradient now is calculated with the help of the potential
difference, which is the electrical voltage. The transport equation for steady state is
electric charge
- = coefficient
voltage gradient
area
time
or
7.8 Analogous Transport Phenomena: Heat and Electricity 305

1 dQc dφ
-
=κ
ð7:26Þ
A dt dx
In this form, the transport equation is identical to Ohm’s law. The coefficient is
the specific electrical conductivity.

Q heat in J
A area in m2
t time in s
T temperature in K
x length in m
φ electric potential in V
Qc electric charge in C
κ specific electrical conductivity in S
m-1
λ specific heat conductivity in W
K-1
m-1

It can be seen that the general transport equation is useful for a number of
different phenomena that at first glance do not reveal any relationship, such as
permeation and heat conduction. The reason is that the transport mechanism can
be described with the model of steady state diffusion. Tables 7.6 and 7.7 compile the
names for the coefficients, resistances, and guide values for the respective cases.
The general transport equation is a partial differential equation that is suitable for
describing the transport of several components in n dimensions. In this chapter, we
have applied the transport equation only for the steady state, one-dimensional case,
for a single substance. We have seen that the concept of additive resistors known
from electricity can also be used for mass transfer. The transport equation can also be
used to describe flow, heating, and drying processes [17]. In the application of the
general transport equation for the transient case, there are similar analogies (Fick’s
second law, Fourier’s second, etc.), which again allow the application of electro-
analog models there.

Table 7.6 Cases for use of the transport equation and naming of their respective coefficients
Flow of Caused by gradient of ... Coefficient is called
Mass concentration in kg
m-3 diffusion coefficient in m2
s-1
Volume concentration in m3
m-3 diffusion coefficient in m2
s-1
Substance concentration in mol
m-3 diffusion coefficient t in m2
s-1
Mass partial pressure in Pa permeability coefficient in kg
s-1
m-1
Pa-1
Volume partial pressure in Pa permeability coefficient in m2
s-1
Pa-1
Substance partial pressure in Pa permeability coefficient in mol
s-1
m-1
Pa-1
Heat temperature specific heat conductivity in W
K-1
m-1
Electricity electric potential specific electric conductivity in S
m-1
306 7 Permeability

Table 7.7 Terms of different usages of the general transport equation

Transport Transport
phenomena equation R is called
1
R = G is called 1
R = K
Ad K is called 1
K is called
Substance n_ = R1
Δc resistance conductance of 1
R = D
Ad coefficient of diffusion
amount (Fick’s law) of diffusion diffusion in resistivity
(mol) diffusion in m3
s-1 m2
s-1 in s
m-2
in s
m-3
Mass (kg) m_ = R1
Δp resistance conductance of 1
R = P
Ad coefficient permeation
(Permeation of permeation of resistivity
of gases) permeation in permeation in
in kg
s-1
Pa-1 in kg
m
s-1
Pa
s
m
kg-1
Pa
s
kg-1 m-2
Pa-1
Charge (C) I = R1
U electric electrical 1
R = κ
Ad specific electric
(Ohm’s law) resistance conductance electric resistivity
in Ω in S conductivity in Ω
m-1
in S
m-1
Heat (J) Q_ = R1
ΔT thermal thermal 1
R =λ
dA specific heat
(Fourier’s resistance conductance of thermal resistivity
law) in K
W-1 in W
K-1 conductivity in K
m
W-1
in
W
K-1
m-1

Further Reading

Nanotechnology in food packaging films [18–24]

Nanofunctional packaging [25–31]
Antimicrobial films [32–37]
Nonporous films of high permeability [38, 39]
Shelf life prediction on permeability data [40]
Diffusion in food: compilation of literature data [41]
Nano-inspired oxygen barrier coatings for food packaging [42]
Agar-based edible films for food packaging applications [43]
Films with insect repellent agents [44]
Hydrogels with integrated color indicator [45, 46]
Temperature-sensitive smart packaging [47]
Biodegradable films for food packaging [48]
Control of adhesion at packaging films [49]
Active packaging for meat [50]
Gas permeability test of whole packaging [51]
Oxygen indicator in intelligent food packaging [52–55]
Film design for respiration in packaging [56]
DSC characterization of packaging films [7]
Pizza: heat and mass transfer properties during baking [57]
Transport phenomena in food processing [17]
Permeability of edible packaging [58]
References 307

Summary
The permeation of gaseous substances through packaging materials is based
on diffusion as a mass transport mechanism. Under the assumption of station-
ary conditions, the determination of the water vapor and oxygen permeability
on plastic films is demonstrated and illustrated with examples. Comparison
with conduction of electric current or heat flow allows the development of
electro-analog models for the calculation of permeation. In this way, values for
films made of several materials can also be easily calculated in order to design
composite films with required properties, e.g., from biopolymers. Using the
example of permeation steady state diffusion-controlled transport processes
can be understood and the multitude of specific terms be ordered. At the end of
the chapter examples from packaging technology are listed, which can be used
for further studies and as suggestions for own scientific work.

References
1. DIN 1343 (1990) Referenzzustand, Normzustand, Normvolumen; Begriffe und Werte. Beuth,
Berlin. https://doi.org/10.31030/2333155
2. Robertson GL (2012) Food packaging: principles and practice. Marcel Dekker, New York
3. Rodríguez-Rojas A, Arango Ospina A, Rodríguez-Vélez P, Arana-Florez R (2019) ¿What is the
new about food packaging material? A bibliometric review during 1996–2016. Trends Food Sci
Technol 85:252. https://doi.org/10.1016/j.tifs.2019.01.016
4. Majid I, Ahmad Nayik G, Mohammad Dar S, Nanda V (2018) Novel food packaging
technologies: innovations and future prospective. J Saudi Soc Agric Sci 17(4):454. https://
doi.org/10.1016/j.jssas.2016.11.003
5. Deutsches Institut für Normung (2015) Verpackung 1 - DIN-Taschenbuch 490/1. Beuth, Berlin
6. Deutsches Institut für Normung (2012) Verpackung 2 - DIN-Taschenbuch 490/2. Beuth, Berlin
7. Gajdoš J, Galić K, Kurtanjek Ž, Ciković N (2000) Gas permeability and DSC characteristics of
polymers used in food packaging. Polym Test 20(1):49. https://doi.org/10.1016/s0142-9418
(99)00078-1
8. Giacinti Baschetti M, Minelli M (2020) Test methods for the characterization of gas and vapor
permeability in polymers for food packaging application: a review. Polym Test 89:106606.
https://doi.org/10.1016/j.polymertesting.2020.106606
9. ISO 15105 (2007) Kunststoffe - Folien und Flächengebilde - Bestimmung der
Gasdurchlässigkeit - Teil 1: Differentialdruck-Verfahren. Beuth, Berlin
10. ASTM D1434 (2015) Standard test method for determining gas permeability characteristics of
plastic film and sheeting. ASTM International, West Conshohocken, PA. https://doi.org/10.
1520/D1434-82R15E01
11. DIN 53122 (2001) Prüfung von Kunststoff-Folien, Elastomerfolien, Papier, Pappe und anderen
Flächengebilden - Bestimmung der Wasserdampfdurchlässigkeit - Teil 1: Gravimetrisches
Verfahren. Beuth, Berlin
12. ISO/CIE 11664-6 (2014) Farbmetrik - Teil 6: CIEDE2000 Formel für die Farbdifferenz. Beuth,
Berlin
13. Winston PW, Bates DH (1960) Saturated solutions for the control of humidity in biological
research. Ecology 41(1):232. https://doi.org/10.2307/1931961
308 7 Permeability

14. Bleisch G, Langowski H-C, Majschak JP (2014) Lexikon Verpackungstechnik. Behrs,

Hamburg
15. DIN 50014 (2018) Normalklimate für Vorbehandlung und/oder Prüfung – Festlegungen. Beuth,
Berlin. https://doi.org/10.31030/2866227
16. Elsner P, Eyerer P, Hirth T (2012) Domininghaus - Kunststoffe - Eigenschaften und
Anwendungen. Springer, Heidelberg. https://doi.org/10.1007/978-3-642-16173-5
17. Welti-Chanes J, Vélez-Ruiz JF, Barbosa-Cánovas GV (2002) Transport phenomena in food
processing. CRC Press, Boca Raton. https://doi.org/10.1201/9781420006261
18. Rehman A, Jafari SM, Aadil RM, Assadpour E, Randhawa MA, Mahmood S (2020) Develop-
ment of active food packaging via incorporation of biopolymeric nanocarriers containing
essential oils. Trends Food Sci Technol 101:106. https://doi.org/10.1016/j.tifs.2020.05.001
19. Enescu D, Cerqueira MA, Fucinos P, Pastrana LM (2019) Recent advances and challenges on
applications of nanotechnology in food packaging. A literature review. Food Chem Toxicol
134:110814. https://doi.org/10.1016/j.fct.2019.110814
20. Silvestre C, Duraccio D, Cimmino S (2011) Food packaging based on polymer nanomaterials.
Prog Polym Sci 36(12):1766. https://doi.org/10.1016/j.progpolymsci.2011.02.003
21. Topuz F, Uyar T (2020) Antioxidant, antibacterial and antifungal electrospun nanofibers for
food packaging applications. Food Res Int 130:108927. https://doi.org/10.1016/j.foodres.2019.
108927
22. Jildeh NB, Matouq M (2020) Nanotechnology in packing materials for food and drug stuff
opportunities. J Environ Chem Eng 8(5):104338. https://doi.org/10.1016/j.jece.2020.104338
23. Alfei S, Marengo B, Zuccari G (2020) Nanotechnology application in food packaging: a
plethora of opportunities versus pending risks assessment and public concerns. Food Res Int
137:109664. https://doi.org/10.1016/j.foodres.2020.109664
24. Dudefoi W, Villares A, Peyron S, Moreau C, Ropers MH, Gontard N, Cathala B (2018)
Nanoscience and nanotechnologies for biobased materials, packaging and food applications:
new opportunities and concerns. Innovative Food Sci Emerg Technol 46:107. https://doi.org/
10.1016/j.ifset.2017.09.007
25. Nakazato G, Kobayashi RKT, Seabra AB, Duran N (2017) Use of nanoparticles as a potential
antimicrobial for food packaging. Academic Press, pp 413–447. https://doi.org/10.1016/b978-
0-12-804303-5.00012-2
26. Ntim SA, Noonan GO (2017) Nanotechnology in food packaging. Janua 2017:118–142. https://
doi.org/10.1039/9781782626879-00118
27. Singh P, Ikram S (2017) Nanotechnology in food packaging: an overview. J Adv Mater 1:19
28. Khalaj MJ, Ahmadi H, Lesankhosh R, Khalaj G (2016) Study of physical and mechanical
properties of polypropylene nanocomposites for food packaging application: nano-clay
modified with iron nanoparticles. Trends Food Sci Technol 51:41. https://doi.org/10.1016/j.
tifs.2016.03.007
29. Carbone M, Donia DT, Sabbatella G, Antiochia R (2016) Silver nanoparticles in polymeric
matrices for fresh food packaging. J King Saud Univ Sci 28(4):273. https://doi.org/10.1016/j.
jksus.2016.05.004
30. Mihindukulasuriya SDF, Lim LT (2014) Nanotechnology development in food packaging: a
review. Trends Food Sci Technol. https://doi.org/10.1016/j.tifs.2014.09.009
31. Youssef AM, El-Sayed SM (2018) Bionanocomposites materials for food packaging
applications: concepts and future outlook. Carbohydr Polym 193:19. https://doi.org/10.1016/j.
carbpol.2018.03.088
32. Mousavi Khaneghah A, Hashemi SMB, Limbo S (2018) Antimicrobial agents and packaging
systems in antimicrobial active food packaging: an overview of approaches and interactions.
Food Bioprod Process 111:1. https://doi.org/10.1016/j.fbp.2018.05.001
33. Santos JCP, Sousa RCS, Otoni CG, Moraes ARF, Souza VGL, Medeiros EAA, Espitia PJP,
Pires ACS, Coimbra JSR, Soares NFF (2018) Nisin and other antimicrobial peptides: produc-
tion, mechanisms of action, and application in active food packaging. Innovative Food Sci
Emerg Technol 48:179. https://doi.org/10.1016/j.ifset.2018.06.008
References 309

34. Otoni CG, Espitia PJP, Avena-Bustillos RJ, McHugh TH (2016) Trends in antimicrobial food
packaging systems: emitting sachets and absorbent pads. Food Res Int 83:60. https://doi.org/10.
1016/j.foodres.2016.02.018
35. Zanetti M, Carniel TK, Dalcanton F, dos Anjos RS, Gracher Riella H, de Araújo PHH, de
Oliveira D, Antônio Fiori M (2018) Use of encapsulated natural compounds as antimicrobial
additives in food packaging: a brief review. Trends Food Sci Technol 81:51. https://doi.org/10.
1016/j.tifs.2018.09.003
36. Makwana S, Choudhary R, Kohli P (2015) Advances in antimicrobial food packaging with
nanotechnology and natural antimicrobials. Int J Food Sci Nutr Eng 5(4):169. https://doi.org/10.
5923/j.food.20150504.02
37. Sung SY, Sin LT, Tee TT, Bee ST, Rahmat AR, Rahman WAWA, Tan AC, Vikhraman M
(2013) Antimicrobial agents for food packaging applications. https://doi.org/10.1016/j.tifs.
2013.08.001
38. Kim D, Seo J (2018) A review: breathable films for packaging applications. Trends Food Sci
Technol 76:15. https://doi.org/10.1016/j.tifs.2018.03.020
39. Saha N, Benlikaya R, Slobodian P, Saha P (2015) Breathable and polyol based hydrogel food
packaging. J Biobased Mater Bioenergy 9(2):136. https://doi.org/10.1166/jbmb.2015.1515
40. Del Nobile MA, Buonocore GG, Limbo S, Fava P (2003) Shelf life prediction of cereal-based
dry foods packed in moisture-sensitive films. J Food Sci 68(4):1292. https://doi.org/10.1111/j.
1365-2621.2003.tb09642.x
41. Krokida MK, Zogzas NP, Maroulis ZB (2001) Mass transfer coefficient in food processing:
compilation of literature data. Int J Food Prop 4(3):373. https://doi.org/10.1081/JFP-100108643
42. Rovera C, Ghaani M, Farris S (2020) Nano-inspired oxygen barrier coatings for food packaging
applications: an overview. Trends Food Sci Technol 97:210. https://doi.org/10.1016/j.tifs.2020.
01.024
43. Mostafavi FS, Zaeim D (2020) Agar-based edible films for food packaging applications – a
review. Int J Biol Macromol 159:1165. https://doi.org/10.1016/j.ijbiomac.2020.05.123
44. Marsin AM, Muhamad II, Anis SNS, Lazim NAM, Ching LW, Dolhaji NH (2020) Essential
oils as insect repellent agents in food packaging: a review. Eur Food Res Technol 246(8):1519.
https://doi.org/10.1007/s00217-020-03511-1
45. Alpaslan D (2019) Use of colorimetric hydrogel as an indicator for food packaging applications.
Bulletin of Materials Science 42(5):247. https://doi.org/10.1007/s12034-019-1908-z
46. Batista RA, Espitia PJP, Quintans JSS, Freitas MM, Cerqueira MA, Teixeira JA, Cardoso JC
(2019) Hydrogel as an alternative structure for food packaging systems. Carbohydr Polym 205:
106. https://doi.org/10.1016/j.carbpol.2018.10.006
47. Singh S, Gaikwad KK, Lee M, Lee YS (2018) Temperature sensitive smart packaging for
monitoring the shelf life of fresh beef. J Food Eng 234:41. https://doi.org/10.1016/j.jfoodeng.
2018.04.014
48. Meritaine R, de Souza MM, Prentice C (2018) Biodegradable films: an alternative food
packaging:307–342. https://doi.org/10.1016/B978-0-12-811516-9.00009-9
49. Liu XW, Wang LQ, Qiao YF, Sun XX, Ma SF, Cheng XY, Qi WF, Huang WQ, Li YH (2018)
Adhesion of liquid food to packaging surfaces: mechanisms, test methods, influencing factors
and anti-adhesion methods. J Food Eng 228:102. https://doi.org/10.1016/j.jfoodeng.2018.
02.002
50. Ahmed I, Lin H, Zou L, Brody AL, Li ZX, Qazi IM, Pavase TR, Lv LT (2017) A comprehen-
sive review on the application of active packaging technologies to muscle foods. Food Control
82:163. https://doi.org/10.1016/j.foodcont.2017.06.009
51. Acerbi F, Guillard V, Guillaume C, Gontard N (2016) Assessment of gas permeability of the
whole packaging system mimicking industrial conditions. Food Packaging Shelf Life 8:81.
https://doi.org/10.1016/j.fpsl.2016.04.003
52. Vu CH, Won K (2013) Novel water-resistant UV-activated oxygen indicator for intelligent food
packaging. Food Chem 140(1–2):52. https://doi.org/10.1016/j.foodchem.2013.02.056
310 7 Permeability

53. Lee SJ, Rahman ATMM (2014) In: Han JH (ed) Intelligent packaging for food products.
Academic Press, San Diego, pp 171–209. https://doi.org/10.1016/b978-0-12-394601-0.
00008-4
54. Vanderroost M, Ragaert P, Devlieghere F, De Meulenaer B (2014) Intelligent food packaging:
the next generation. Trends Food Sci Technol 39(1):47. https://doi.org/10.1016/j.tifs.2014.
06.009
55. Müller P, Schmid M (2019) Intelligent packaging in the food sector: a brief overview. Foods
8(1). https://doi.org/10.3390/foods8010016
56. Finnegan E, Mahajan PV, O’Connell M, Francis GA, O’Beirne D (2013) Modelling respiration
in fresh-cut pineapple and prediction of gas permeability needs for optimal modified atmosphere
packaging. Postharvest Biol Technol 79:47. https://doi.org/10.1016/j.postharvbio.2012.12.015
57. Dumas C, Mittal GS (2002) Heat and mass transfer properties of pizza during baking. Int J Food
Prop 5(1):161. https://doi.org/10.1081/Jfp-120015599
58. Muthu SS (2021) Sustainable packaging. Springer, Singapore. https://doi.org/10.1007/978-
981-16-4609-6
Thermal Properties
8

Most food processing operations used to prolong the shelf life of foods involve
heating foods to temperatures capable of inactivating microbial and enzymatic
activity. These heat treatments are based on controlled heat transfer that depends
upon thermal properties of the food materials. To increase the internal temperature of
a solid body food product, such as food sealed in a can or jar, heat must first be
transferred to the outer surface of the food container and then conducted through the
food material in order to reach the center of the food product. This is an example of
heat transfer by means of conduction. In Sect. 8.6 heat transfer is described in more
detail. When heat is added to a material (heating), the temperature of that material
will increase so long as it is not undergoing a change in phase. The extent of
temperature rise is governed by the heat capacity of the material (see Sect. 8.4).
When heat is removed from a material (cooling) and transferred to a surrounding
heat exchange medium at a lower temperature, the temperature of the material will
decrease. Figure 8.1 illustrates these different directions of heat flow. In food
processing, thermal process operations are very important for food safety. Some
examples of thermal process operations are listed in Table 8.1.
In this chapter we want to focus on thermal properties of foods, such as heat
capacity, temperature and enthalpy of phase transition points (melting, freezing,
glass transition), chemical reactions, evaporation, etc.), as well as the caloric value of
foods. We will also introduce some methods and techniques for measuring some of
these thermal properties.

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 311
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_8
312 8 Thermal Properties

Fig. 8.1 Heat transfer across

the interface of a food

heat exchange medium food

heat heating

heat cooling

8.1 Temperature

The temperature of a system is a measure of the kinetic energy of its molecular

components. The product of absolute temperature T and the Boltzmann constant k is
called the thermal energy E of a system or energy of activation:
E=k
T ð8:1Þ

E energy in J
k Boltzmann’s constant in J
K-1
T temperature in K

The SI unit Kelvin is defined by natural constants [1]. The exact definitions of the
SI basic units are set out in the Appendix.
When referring to the term thermal energy, we can say that a Kelvin degree
corresponds to a change in thermodynamic temperature, which leads to a change in
thermal energy of a system (kT) by 1.380649
10-23 J [2]. Before the introduction of
the fundamental constant-based international system of units (SI), 1 K was defined as
the 273.16th part of the triple point temperature of the water.
In the case of an ideal gas, increasing thermal energy can be imagined as an
increasing mean velocity of the gas molecules. At that level, heat is nothing more
than the kinetic energy of atoms or molecules of a system. This heat (the so-called
sensitive heat) can be measured externally by measuring the temperature of the
system. An elevated temperature indicates high kinetic energy of the molecules,
whereas near the zero point of absolute temperature, the molecules are at rest. The
8.1 Temperature 313

Table 8.1 Thermal process operations important in food engineering

Food Heat exchange Temperature
Operation Purpose examples medium Range
Pasteurization/ Inactivation of Meat, fish, Electricity, hot 63–135 °C
sterilization pathogenic soup, water, steam
microorganisms vegetables, (direct or
for increasing fruit, cream indirect steam)
shelf life.
– Batch – Canned solids Milk, cream, Steam
and liquids custard,
desserts,
soup, fruit,
juice, beer,
egg
– Long-life aseptic Long-life Steam
Continuous packaging milk, cream,
flow/liquids fruit juices
(UHT/HTST)
Evaporation Removal of Milk, fruit Steam 40–100 °C
water, and
production of vegetables,
liquid coffee, cheese
concentrate whey
Dehydration Removal of Milk, potato, Hot air, steam, 150–250 °C
water, vegetables, hot water,
production of fruit, meat, electricity
dried material fish
with low water
activity
Frying Reactions French fries, Hot oil 100–150 °C
(proteins, potatoes,
carbohydrates), doughnuts
evaporation of
water
Cooking/ Reactions Catering Steam, hot air, 100–200 °C
baking (proteins, operations, microwaves
carbohydrates), bread, meat
evaporation of pies, cakes
water
Chilling/ Reducing Dairy Cold air, 10–0 °C
freezing spoilage products, refrigerants, -18 to -30 °C
reactions, meat, fish, cryogenic fluids -100 to -200 °C
microbial fruit, (liquid nitrogen
activity vegetables, or carbon
frozen dioxide)
desserts

absolute temperature of a system is given in the unit K (Kelvin). The scale of

absolute temperature starts at 0 K (absolute zero point of temperature) and is open
to the top.
Because of historic reasons, there are other temperature scales (°C, °F, °R) having
other units and fixed points. For example, the Celsius scale is based on the fixed
314 8 Thermal Properties

points for the temperatures at which water will freeze (freezing point) and boil
(boiling point) at standard atmospheric pressure. The temperature difference
between those fixed points was defined to be 100 degrees. In a similar manner the
Fahrenheit scale was based on two fixed points that could be recognized at the time.
Zero on the Fahrenheit scale (-17.8 °C = 0 °F) was the lowest temperature that
could be reached at that time, and the high point of the scale (100 °F) was set at what
was believed to be body blood temperature of a healthy person (37 °C = 100 °F).
The temperature difference between those fixed points was defined to be 100 degrees.
Likewise, there exists an absolute temperature scale based on each degree being the
same as a Fahrenheit degree. This is called the Rankine temperature scale (R), and is
a counterpart to the Kelvin temperature scale, but in Fahrenheit degree units (instead
of Celsius degree units). Table 8.2 shows an overview. Formulas for conversion of
temperatures between these different scales are given in the appendix.
For industrial use, there is an international temperature scale called ITS 90 (inter-
national temperature scale of 1990). It is based on fixed points in the range between
0.7 K and 2500 K, which can be reproduced by many laboratories. In Table 8.3 there
are some fixed points of ITS-90 shown, which are of interest for food engineers. The
number of fixed points and their values are adjusted occasionally by international
conventions between national metrological institutes.

8.2 Heat and Enthalpy

Heat is a form of energy. Energy per se, however, can neither be created nor
destroyed. This is known as the first law of thermodynamics, and is often used by
engineers as the rule of energy conservation in carrying out energy balance
calculations on a system. Table 8.4 shows some examples of forms of energy.
That part of heat that leads to a temperature increase (or decrease) of the system is
called sensitive heat. That part of heat that is absorbed (or emitted) by a system
without changing the temperature of the system is called latent heat. Latent heat is
transferred, for example, when water evaporates, condenses, freezes, or melts.
While all other forms of energy can theoretically be transferred into each other
with an efficiency of 100% (in reality, however, the efficiency is always less than
100%), this does not apply to heat: While all forms of energy can be completely
converted into heat, heat cannot be completely converted into other forms of energy
such as work or kinetic energy. The maximum part that can be converted into work is
limited for thermodynamic reasons and is called Carnot efficiency. This means heat
has a special position among all forms of energy.
The internal energy U of a thermodynamic system exists in the forms of both heat
and work. Therefore, two transformations are possible for internal energy. Transfer
of heat Q and/or transfer of work W′. So, we can express internal energy in the
following way:
dU = dQ þ dW 0 ð8:2Þ

In a thermodynamic system, we treat work only in the form of displacement work

W (force-displacement or pressure-volume). We assume that other forms of work
 8.2 Heat and Enthalpy

Table 8.2 Some fixed points for temperature, and related scales
Absolute Boiling Freezing Triple point Temperature of human Boiling temperature Temperature
zero temperature of N2 temperature of H2O of H2O body blood of H2O scale name
T 0 77.4 273.15 273.16 310.15 373.15 Kelvin
(K)
ϑ (° -273.15 -195.8 0.0 0.01 37 100.0 Celsius
C)
ϑ (° -459.67 -320.4 32.0 32.02 100 212.0 Fahrenheit
F)
315
316 8 Thermal Properties

Table 8.3 Some fixed points of the international temperature scale, ITS-90, which are in the
temperature range of food processes
Equilibrium state ϑ (°C) after ITS-90
Triple point of water 0.01
Melting point of Gallium 29.7646
Freezing point of Indium 156.5985

Table 8.4 Different forms of energy

Energy form Example: energy in . . .
Mechanic energy a loaded spring (potential energy)
a moving body (kinetic energy)
Electrical energy electric power networks
Light energy light of sun, light from incandescent bulb
Chemical energy vegetable oil, mineral oil, potato starch, fuel
Nuclear energy nucleus of atom emitting radioactivity
Heat energy cooking/heating stove, home heating system

like electric, magnetic, elastic and frictional are not involved. Displacement work
W is the product of pressure and volume change.
dW = - pdV ð8:3Þ

The negative sign in Eq. (8.3) accounts for negative displacement, while dV
represents energy uptake of a system, and must be counted as a positive contribution
(and vice versa).
So it is
dU = dQ þ dW ð8:4Þ

and
dU = dQ - pdV ð8:5Þ

or
dQ = dU þ pdV ð8:6Þ

U internal energy in J
Q heat in J
p pressure in Pa
W′ work in J
W displacement work in J
V volume in m3
H enthalpy in J

The term enthalpy H now is used for the sum of internal energy and the product
pV:
8.2 Heat and Enthalpy 317

Table 8.5 Heat dQ transferred to/from a system (isobaric cases)

General process dQ = dU + pdV + Vdp dQ = dH
Isobaric process with displacement work dQ = dU + pdV + 0 dQ = dH
Isobaric process with displacement work being zero dQ = dU + 0 + 0 dQ = dU

H = U þ pV ð8:7Þ

So, for a change in enthalpy, we have

dH = dU þ pdV þ Vdp ð8:8Þ

If we consider only cases where the pressure is constant (dp = 0), then we have:
dH = dU þ pdV ð8:9Þ

together with Eq. (8.6) this means:

dQ = dU þ pdV = dH ð8:10Þ

or
dQ = dH ð8:11Þ

This means that the amount of heat dQ which occurs during an isobaric (constant
pressure) process is the same as the change in enthalpy of the system. So, when we
investigate material properties in a laboratory under constant pressure (normally
atmospheric), we talk about enthalpy instead of energy of a system. Thermal analysis
also measures amounts of heat to obtain information about the enthalpy of a system.

Definition
Enthalpy is the combination of internal energy U and the product pV of a
system.

The difference between change in internal energy dU of a system and change in

its enthalpy dH lies in the work dW′ or under the above-mentioned simplifications in
the displacement work. If in an isobaric process, there is no displacement or it is zero,
then the displacement work plays no role in the system, and the distinction between
internal energy and enthalpy is no longer important. The values of dH and dU are the
same. Table 8.5 summarizes the cases.

Bottom Line
Heat generated in isobaric processes corresponds to the enthalpy change of the
system. It results from the change in internal energy plus displacement work. If
displacement work is zero, the change in enthalpy is equal to the change in
internal energy.
318 8 Thermal Properties

Definition
The calorie (cal) is a historic unit for heat. The official SI unit for heat and
energy is the joule (J), conversion: 1 cal = 4.1868 J. The definition of calorie
(cal) was based on the heating of water: it was the amount of heat needed to
raise the temperature of 1 g of water by 1 K. Depending on the measurement
technology (the heat capacity of the water is temperature-dependent), slightly
different values resulted and thus half a dozen different values for a calorie
[3]. The indication of the energy value of food in accordance with the
European Food Information Regulation [4] the unit cal may only be used in
the form of the kcal after an indication in kJ. Example: “The caloric value of
100 g is 250 kJ (60 kcal).”

8.3 Thermodynamics: Basis Principles

Thermodynamics is the body of science in which we study the way in which

substances are affected by heat, either when being heated or cooled, and especially
when heat addition or removal causes a phase change. It is no surprise therefore that
thermodynamics is an essential topic that must be well understood by most
engineers, especially food engineers. Normally, entire textbooks are devoted solely
to a basic primer in thermodynamics. Since thermodynamics is not the main topic of
this book, only a brief discussion of basic principles will be presented to appreciate
the importance of thermal properties.
Measuring thermal properties of materials requires that we conduct experiments
that cause thermal effects to occur and record the results of these effects. Most often,
temperature or quantity of heat is measured and monitored. Observing the tempera-
ture dependency (like the pressure dependency) of a physical quantity is a common
way to study the energetic behavior of a material on a molecular scale.
The energy of a system is often determined with calorimeters. Since only thermal
effects or temperature changes are accessible from a metrological point of view, the
contributions of internal energy, work and heat to the thermal effect must be
calculated. In addition, information can be obtained from temperature dependence
and pressure dependence of thermodynamic variables.

8.3.1 Laws of Thermodynamics

In the previous section, we just learned that the law of energy conservation stems
directly from thermodynamics, and is called the first law of thermodynamics
(Eq. 8.6). Recall this equation was derived with the assumption that no work other
than displacement work would occur.
8.3 Thermodynamics: Basis Principles 319

dU = dQ - pdV ð8:6Þ

It states that in a closed system, changes in the internal energy dU of a system can
only be achieved in two ways: exchange of heat dQ or exchange of work pdV.
If a system is thermally insulated, it can be assumed to be adiabatic (isentropic),
then dQ = 0. This means that an exchange of heat does not take place. If the volume
is constant, dV = 0, then the change in internal energy is zero, dU = 0 i.e. U = const.
The first law of thermodynamics is thus a form of the law of conservation of energy.
Entropy S is the quotient of heat and temperature. For a reversible operation in a
closed system, the entropy change dS is

dQrev
dS  ð8:12Þ
T
This definition is valid for a closed system and a completely reversible process.
When the process of interest is partly or completely irreversible, then we have:

dQ
dS > ð8:13Þ
T
Equation 8.13 is an expression of the second law of thermodynamics, stating that
if a system is not in equilibrium the entropy S tends to increase and to reach a
maximum.
It can remain constant in the best case (for reversible operations). In all spontane-
ous processes, i.e., in natural processes, the entropy increases: dS > 0.
With the help of entropy, state functions can be formulated. For example, free
energy F, as the difference between the internal energy and the product of tempera-
ture and entropy is also known as Helmholtz energy.
F=U -T
S ð8:14Þ

Also, free enthalpy, G, s the difference between enthalpy and the product of
temperature and entropy, and is known as Gibbs energy.
G=H -T
S ð8:15Þ

Both F and G indicate the maximum amount of work a system can deliver.
Maximum means that the conversion would be completely reversible. In this
understanding F and G have the property of a potential.
We already know the concept of potential from the section on transport variables
in Chap. 7. A potential curve as in Fig. 8.2 has curve sections with positive and
negative slope as well as extreme points. From the derivative, i.e., the slope of the
potential curve, it is possible to read how a system will behave when the variable
x changes.
In mechanics, the concept of potential is used, for example, in connection with the
position energy (potential energy) dependent on the height of a body. As long as the
320 8 Thermal Properties

Fig. 8.2 Potential of a

system—schematically—as a
function of a variable φx. The
point O indicates the position
of a stable equilibrium

body is not at the minimum point of its potential energy, a force acts on it that drives
it towards this minimum. The direction and magnitude of this force can be calculated
from the derivation of the potential curve. At the minimum of the curve (point O in
Fig. 8.2), the slope of the potential curve and thus the force is zero.

Definition
The concept of potential can be illustrated by the position energy of a body in
the Earth’s gravitational field: The position energy is calculated from the scalar
→ →
product E = m g
h . If the height is increased by dh, the position energy
changes by dE = mg
dh. The term dE dh = mg is the potential of the body in
the gravitational field. The product of potential and dh provides the energy
change dE = dE dh
dh of the body.

Analogously, in thermodynamics we have dG = dG dx
dx , this means that the

change in the free enthalpy G is obtained from the product of thermodynamic
potential dG
dx and the change dx.
In order to familiarize ourselves with the use of thermodynamic potential, we first
consider the dependence of enthalpy H on pressure and temperature.with
dH = T
dS þ Vdp ð8:16Þ

or

∂H ∂H
dH =
dp þ
dT ð8:17Þ
∂p T ∂T p

∂H
in an isothermal case with dS = 0, the partial differential dH becomes ∂p T
and
with Eq. (8.16)

∂H ∂ðV
dpÞ
= =V ð8:18Þ
∂p T ∂p T
8.3 Thermodynamics: Basis Principles 321

and the partial differential is

∂H
= cp ð8:19Þ
∂T p

The first derivative of the enthalpy function after the variable p, i.e., the partial
differential ∂H
∂p
indicates that the thermodynamic potential is the volume V,
T
which is a state variable. The first derivative of the enthalpy function after the
variable T, i.e., the partial differential ∂H∂T p
indicates that the thermodynamic
potential is the heat capacity cp, which is a material property.

→ Attention
While a differential is usually dG
dx is written, a partial differential is written ∂G
∂x
∂G
Example: The expression ∂p T
means “partial differential from G with
respect to p at constant temperature T”.

Let us now consider the dependence of the free enthalpy G on pressure p and
temperature T. Here, too, we want to describe the dependence of p and T as
thermodynamic potentials.
The differential of Eq. (8.15) is
dG = dH - T
dS - S
dT ð8:20Þ

with
dH = T
dS þ Vdp ð8:21Þ

we get
dG = T
dS þ Vdp - T
dS - S
dT ð8:22Þ

i.e.
dG = Vdp - S
dT ð8:23Þ

if we write the p-T dependence of dG as partial differentials, this is

∂G ∂G
dG =
dp þ
dT ð8:24Þ
∂p T ∂T p

By comparing of Eq. (8.23) with Eq. (8.24) we see

∂G
=V ð8:25Þ
∂p T
322 8 Thermal Properties

and

∂G
- =S ð8:26Þ
∂T p

The partial differential of G, i.e., the first derivative, provides the potential of
G with respect to p or T. The partial differential of the free enthalpy G with respect to
pressure is identical to the volume V of the system. The negative partial differential
of the free enthalpy G with respect to temperature is identical to the entropy of the
system. Let us compare again with the case of mechanical position energy. There,
the potential was the mechanical force that drives the body into equilibrium.
Therefore, V and S take on the role of “mechanical force” with respect to G, which
drives the system towards thermodynamic equilibrium.
As a simple case, let’s consider a food that emits water vapor and is on its way to
equilibrium with its ambient atmosphere. We leave the temperature constant at room
temperature. If we express the volume of water vapor by the volume of an ideal gas,
we obtain for the potential

∂G R
T
=V = ð8:27Þ
∂p T
p

or
dp
dG = R
T
ð8:28Þ
p

or

dG = R
T
d ln p ð8:29Þ

The change in free enthalpy in this simple system can be calculated after
measuring the water vapor pressure. If we consider that water vapor is not an ideal
gas, we can replace the pressure p with the so-called fugacity f of the water vapor.
dG = R
T
d ln f ð8:30Þ

The dimensionless relative fugacity is the so-called activity of a substance. In case

of water, we have already know this quantity as water activity aW (refer Chap. 1).
Up to this point, we have focused on the variables p and T. In the case of chemical
reactions, dG is dependent on the amount of substance n in the reactants and
products, and is generally dependent on ni. The partial differential of free enthalpy
with respect to the amount of substance ni is again a thermodynamic potential called
chemical potential μ.

∂G
= μi ð8:31Þ
∂ni p,T,njðj ≠ iÞ

The chemical potential describes the potential of a substance to undergo a

reaction and thereby to release reaction energy.
8.4 Heat capacity 323

G Gibbs energy in J
n substance in mol
f fugacity in Pa
μ chemical potential in J
mol-1
S entropy in J
K-1
aW water activity

In this section, we have seen the physical significance of the first derivative of
enthalpy and free enthalpy. In Sect. 8.5, we will see that the second derivative of
these functions is also physical quantities.

8.4 Heat capacity

Heat capacity of a material is a thermal property that indicates the ability of the
material to hold and store heat. It can be quantified by specifying the amount of heat
that is needed to raise the temperature by a specified amount. Mathematically, it is
the quotient of heat ΔQ divided by change in temperature ΔT:

ΔQ
C= ð8:32Þ
ΔT
For infinitesimal changes
dQ
C= ð8:33Þ
dT
When heat capacity is defined only in this way, it will also depend upon the mass
or size of the material sample, and serves as a property only of the specific sample
size measured. For this reason, we normally measure and report the heat capacity
bases on a common unit of mass. When we do this, we call it the specific heat
capacity. Sometimes this property is called specific heat of the material but this
should be avoided because dQ/dm = q is specific heat.
C
c= ð8:34Þ
m

1 dQ dq
c=
= ð8:35Þ
m dT dT
In order to help better understand heat capacity, let us assume we wish to
determine how much heat is needed to raise the temperature of 1 L of water at
21 °C up to 23 °C (by 2 degrees K). When we do this, we measure the heat required
to be 8.36 kJ.
When heat is added to a system like this (liter of water), the water molecules
experience an increase in their kinetic energy. They move in both rotational and
translational motion at faster rates. If we insert our finger (or a thermometer) into this
324 8 Thermal Properties

Table 8.6 Heat capacity p = const. V = const.

terms at constant pressure
and volume C p = dH
dT CV = dU
dT
cp = m1
dH
dT cV = m
dT
1 dU

cp = dT
dh
cV = du
dT
-1
C heat capacity in J
K
c specific heat capacity in J
kg-1
K-1
Q heat in J
q specific heat in J
kg-1
T temperature in K
m mass in kg
H enthalpy in J
h specific enthalpy in J
kg-1
U internal energy in J
u specific internal energy in J
kg-1

liter of water, we can sense this increased thermal energy level by a warming
sensation on our finger, and a rise in the reading of the temperature scale of the
thermometer. Therefore, we use temperature as a measure of increased thermal
energy. In this case, the temperature increase was ΔT = 2 K.

Example
Calculation of the specific heat capacity of water. 1 kg of water was supplied
with 8.36 kJ of heat, after which the temperature of the water rose by 2 K:

1 ΔQ
cp =

m ΔT
8:36 kJ
cp =
1 kg
2 K
cp = 4:18 kJ
kg - 1
K - 1

Often in thermodynamics, it is necessary to distinguish between internal energy

terms in which displacement work is present or not present. When there is no
displacement work, the volume of the system remains constant (V = const., dQ =
dU ), and the subscript V is used with the heat capacity term. When displacement is
present, then the pressure of the system remains constant ( p = const., dQ = dH ), and
the subscript p is used with the heat capacity term. Table 8.6 summarizes these
designations.
8.4 Heat capacity 325

8.4.1 Ideal Gases and Ideal Solids

For ideal gases and solids, the movement of molecules in response to thermal energy
can be predicted from theory. Therefore, the heat capacity of such ideal substances
can also be predicted from theory. In an ideal gas, the molecules are free to have
translational motion in the three directions of three-dimensional space (back and
forth, side-to-side, and up and down). Thus, we say, they have three degrees of
freedom for translational motion, f = 3.
In addition to translational motion, molecules are also free to have rotational
motion. Gas molecules consisting of only two atoms, such as oxygen, O2, or
nitrogen N2, are bonded linearly. This makes them look like a rigid body which
can rotate about two axes. This gives them two more additional degrees of freedom.
Thus, they have a total of five degrees of freedom, f = 5.
In the case of solids, the molecules are fixed in place, and can have no transla-
tional or rotational motion. Therefore, they have no translational or rotational
degrees of freedom. However, they are free to oscillate in three directions. Thus,
they have six degrees of freedom in response to thermal energy. Table 8.7 lists the
degrees of freedom for atoms or molecules in some simple ideal systems.
In the mechanical model of an ideal gas, the kinetic energy is equal to the thermal
energy:

1 f
m
v2 = kT ð8:36Þ
2 2

f
E= kT ð8:37Þ
2
The internal energy U for N particles (atoms or molecules) is

f
U = N
kT ð8:38Þ
2

N f f
U=
k
NA
T = n
R
T ð8:39Þ
NA 2 2

f R f
U = n
M


T = m
Rs
T ð8:40Þ
2 M 2
With the heat capacity for constant volume

Table 8.7 Degrees of freedom for atoms and molecules in simple, ideal systems
Translational Rotational Oscillatory Total
One atom gas 3 – – f=3
Two-atom gas molecule 3 2 0 f=5
Molecules in a solid – – 6 f=6
326 8 Thermal Properties

dQ dU
CV = = ð8:41Þ
dT dT
we get

f
C V = m
Rs ð8:42Þ
2
or

f
cV = R ð8:43Þ
2 s
In the isobaric case we get

dQ dH dU p
dV dU m
Rs
dT
Cp = = = þ = þ ð8:44Þ
dT dT dT dT dT dT
So, we see
C p = CV þ m
Rs ð8:45Þ

or

cp = c v þ R s ð8:46Þ

k Boltzmann’s constant
NA Avogadro’s constant
m mass in kg
v velocity in m
s-1
N number of particles
U internal energy in J
M molecular mass in kg
mol-1
R universal gas constant in J
K-1
mol-1
Rs specific gas constant in J
K-1
kg-1
f degrees of freedom
p pressure in Pa
V volume in m3
Cp heat capacity ( p = constant) in J
K-1
CV heat capacity (V = constant) in J
K-1
cp specific heat capacity ( p = constant) in J
kg-1
K-1
cV specific heat capacity (V = constant) in J
kg-1
K-1

In the case of solids without very high pressure, it is

cp ≈ cV ð8:47Þ
8.4 Heat capacity 327

Table 8.8 Theoretical and experimental values of specific heat capacity for selected systems (from
[5])
Substance f Rs (kJ
kg-1
K-1) cp (theory) (kJ
kg-1
K-1) cp (exp.) (kJ
kg-1
K-1)
He 3 2.078 5.195 5.23
N2 5 0.297 1.039 1.03
Air 5 0.287 1.0045 1.005
Pb 6 0.0401 0.1203 0.129

with f = 6, we have:
cp ≈ 3
Rs ð8:48Þ

This is called the rule of Dulong-Petit for the heat capacity of solids. Table 8.8
illustrates how theoretical values based on thermodynamic considerations and
experimental data for heat capacity fit.
From Table 8.8 we can see that the agreements for estimates are quite sufficient,
especially in the area of gases, which can be approached as ideal. In the case of
solids, however, there are also numerous major deviations from the Dulong-Petit
rule at room temperature. To estimate the heat capacity of solid foods, the following
method should preferably be used. It is based on an additivity of the heat capacities
of the material components.

8.4.2 Heat Capacity of Real Solids

If specific heat capacities of the chemical constituents of a substance and its

composition are known, the heat capacity of the substance can be roughly estimated
by adding up the heat capacities of the individual components:

cp = xi
cp,i ð8:49Þ
i

with

mi
xi = ð8:50Þ
m

cp specific heat capacity in J
K-1
kg-1

xi mass fraction of component i
cp,i specific heat capacity of component i in J
K-1
kg-1
mi mass of component i in kg
m total mass in kg

Table 8.9 lists the values for specific heat capacities of food components.
328 8 Thermal Properties

Table 8.9 Some data for cp in kJ
kg-1
K-1

specific heat capacities of
Water ≈ 4.2
food constituents [6]
Carbohydrates ≈ 1.4
Proteins ≈ 1.6
Fats ≈ 1.7
Ash (Minerals) ≈ 0.8
Ice ≈ 2.1
Non-fat solids (from animal) ≈ 1.34 . . . 1.68
Non-fat solids (from plants) ≈ 1.21

Example
Calculation of the heat capacity of a food
The composition of a food product is given as follows:

Mass fraction
xi in% (m/m)
84 Water
12 Fat
3 Protein
1 Minerals

Solution:
The heat capacities of the individual components are added to the total heat
capacity.

cp kJ
kg - 1
K - 1 = 0:84
4:2 þ 0:12
1:7 þ 0:03
1:6 þ 0:01
0:8

-1 -1
cp = 3:8 kJ
kg
K

8.5 Classification of Phase Transitions

When we look at a simple phase transition, such as when a substance undergoes

freezing from a liquid to a solid phase, then we normally see significant changes in
enthalpy, entropy and volume. When we start in the liquid phase toward transition to
the solid phase, we will observe these changes at the phase transition point, which is
a point on the p-T-diagram (phase diagram) of the material. The chemical potential
of both phases is equal at the transition point, but the enthalpies and entropies of the
different phases (solid and liquid) are different.
In other situations, we can observe phase transitions where a solid is melting to
become a viscous liquid without sudden change in enthalpy. This type of phase
transition is called a glass transition, and is observed when we deal with
8.5 Classification of Phase Transitions 329

non-crystalline materials. For example, formulations with solid carbohydrates like

confectionery or powders can exhibit glass transitions [7].
According to Ehrenfest phase transitions are called first-order (n = 1) phase
transitions when the first derivative of Gibbs’ enthalpy over temperature has a
discontinuity (a “jump” in the curve). In the same manner, a phase transition is of
second order (n = 2) when the discontinuity appears in the curve for the second
derivative of Gibbs’ enthalpy over temperature.
Let us consider a simple example of a liquid–solid phase transition upon cooling
of a material. Then, the thermal effect observed during this transition will be the
release of an enthalpy ΔtrsGm which is the difference in the enthalpy of the liquid
phase G00m and the solid phase G0m .

Δtrs Gm = G00m - G0m = 0 ð8:51Þ

The partial derivate over temperature is:

∂Δtrs Gm ∂G00m ∂G0m

= - = - S00m þ S0m = - Δtrs Sm ≠ 0 ð8:52Þ
∂T ∂T p ∂T p

with the help of the second law of thermodynamics this means for the phase
transition enthalpy ΔtrsHm
T
Δtrs Sm = Δtrs H m ≠ 0 ð8:53Þ

From Sect. 8.4 we know Cp is the derivative of H over T, so it is the second

derivative of ΔtrsGm over temperature. Having this in mind, let us now look at the
graphs in Fig. 8.3. Let us imagine moving along on a curve in the diagram from
higher to lower temperature (in the direction from liquid to a solid state). In the upper
left figure, we see the G-T-graphs of the liquid phase (2) and the solid phase (1),
which meet at the transition temperature. Upon decreasing the temperature, the
system will normally change at Ttrs from curve 2 to curve 1 to reach a low energy
state. Plotting the derivative of these curves, we get the H-T-curve immediately
below (middle left), which has a distinct discontinuity (“jump”) at the transition
temperature Ttrs. This is a first-order phase transition.
Let us now look at the graphs for second-order transitions shown on the right side
in Fig. 8.3.
In the upper right graph, we can see that, at the transition temperature Ttrs, the
G-T-curve of phase 2 and phase 1 do not intersect, but are tangent to each other, and
simply touch at a point of tangency. Upon decreasing the temperature, we will enter
the transition from phase 2 to 1. The first derivative is shown in the graph immedi-
ately below (middle right). In this case, the first derivative curve, which is H, simply
changes direction abruptly at the transition point Ttrs, but shows no discontinuity or
“jump.” The second derivative (bottom right) does show such a discontinuity “jump”
at Ttrs: This transition is a second-order phase transition. A glass transition is a
typical example of a second-order phase transition. In a thermal analysis experiment,
330 8 Thermal Properties

first order phase transition (n = 1) second order phase transition (n = 2)

2
2

1
1

Gp,m Gp,m
1
2
1
2

Ttrs Ttrs
T T

Hm Hm

Ttrs Ttrs
T T

c p,m c p,m

Ttrs Ttrs
T T

Fig. 8.3 Classification of phase transitions after Ehrenfest, temperature scan

Table 8.10 Classification of phase transition based on the derivatives of Gibb’s enthalpy
First order phase Second order phase
transition transition
1. Derivative of von ∂Δtrs Gm
= - Δtrs Sm ≠0 =0
∂T
G over T p

2. Derivative of von 2
∂ Δtrs Gm
=-
Δtrs cp,m ≠0 ≠0
∂T 2 T
G over T p

these orders of transition can be identified by having a “jump” in the heat capacity
Cp,m, but not in the enthalpy ΔHm. Table 8.10 summarizes that.
Ehrenfest’s classification based on the derivative of Gibbs’ enthalpy generally
formulated is: A phase transformation of a pure substance of n-th order has an
unsteadiness of the n-th derivative over temperature of Gibbs’ enthalpy.
8.5 Classification of Phase Transitions 331

∂ G0m ∂ G00m
n n
≠
∂T n p ∂T n p

Ehrenfest’s classification of phase transitions works analogously with the deriva-

tive of Gibbs’ enthalpy over pressure instead of temperature:

∂ G0m ∂ G00m
n n
≠
∂pn T ∂pn T

We had already learned about the first derivative of Gibbs’ enthalpy with respect
to pressure when using the thermodynamic potential for the water activity of a food
(refer Eq. 8.25)

∂Δtrs Gm
= Δtrs V m
∂p T

The second derivative of Gibbs’ enthalpy over pressure is

2
∂ Δtrs Gm
= - V m
Δtrs κ
∂p2 T

So the order of a phase transformation can be determined experimentally by a

temperature scan or by a pressure scan. In a temperature scan, one will pay attention
to jumps in enthalpy and heat capacity, in a pressure scan to jumps in molar volume
and compressibility κ T (Fig. 8.4).

C heat capacity in J
kg-1
K-1

c specific heat capacity in J
kg-1
K-1
Q heat in J
T temperature in K
p pressure in Pa
H enthalpy in J
h specific enthalpy in J∙kg-1
U internal energy in J
u specific internal energy in J∙kg-1
G Gibbs’ enthalpy in J
S entropy in J∙K-1
κT isothermal compressibility in K-1
V volume in m3
332 8 Thermal Properties

Fig. 8.4 Classification of phase transitions after Ehrenfest, pressure scan

Indices
trs transition
m molar
n order

With these excursions into the thermodynamics of phase transitions, which are
not simply freezing/melting or evaporating/condensing, we want to leave this section
and return to thermal properties. We might keep in mind that glass transitions, as
well as some solid–solid transitions are related to the quality and the stability of the
food substance in which they occur [8]. There are two approaches through which to
obtain data for constructing phase diagrams: One approach is by scanning the
temperature (thermal analysis) and observing the related quantities that result from
the scan. The other approach is to study materials under an isothermal pressure scan.

8.6 Heat Transfer in Food

Spontaneous heat transfer always takes place from a region of higher temperature to
a surrounding region of lower temperature. There are four mechanisms by which
heat can transfer: radiation, conduction, convection, and phase transitions. In
8.6 Heat Transfer in Food 333

Table 8.11 Mechanisms of heat transfer

Mechanism Description Example
Heat Electromagnetic radiation with Broiling, grilling, infra-red heat
radiation wavelengths from 1 μm to 1 mm lamps, solar heating.
Heat Transport of heat by excited molecules Solid being heated at one end,
conduction “bumping” against each other warming opposite end
Heat Transport of heat being carried by a flowing Flowing hot water or air in central
convection fluid home heating system
Phase Uptake/release of latent heat Condensing of water vapor
transitions

Table 8.11 they are listed and described briefly. In this chapter section and those that
follow, we will discuss each of these mechanisms and the thermal properties that are
needed in each case.

8.6.1 Heat Radiation

Heat radiation is electromagnetic radiation with frequencies below that of visible

light. It is also called infrared radiation. All bodies with a temperature above 0 K
emit heat by radiation. As with all heat transfer mechanisms, heat will flow from the
body at higher temperature to the one at lower temperature. However, there is no
need for the bodies to contact each other, nor is there any need for any substance to
exist between the two bodies. Therefore, heat radiation can occur in a perfect
vacuum and over great distances. For example, this is how we receive heat from
the sun. For heat radiation to occur, the bodies must simply be able to “see” each
other. The Stefan–Boltzmann law allows us to calculate the heat flow under
radiation:

Q_ = A
ε
σ
T 4 ð8:54Þ

ε emissivity of a body
σ Stefan–Boltzmann constant
T thermodynamic temperature of the body in K
A area which is emitting or absorbing radiation in m2

When we have two bodies of different temperatures T1 and T2, each one is
emitting heat radiation according to Stefan–Boltzmann’s law, and at the same time
they are receiving radiation from the other body. So there is a net heat flow of:
334 8 Thermal Properties

ΔQ_ 12 = A
C12
T 42 - T 41 ð8:55Þ

ΔQ_ heat flow from/to body 1 to/from body 2 in J∙s-1

A area emitting radiation in m2
T1 temperature of body 1 in K
T2 temperature of body 2 in K
C12 radiation exchange factor in W
K-4
m-2
ε1 emissivity of body 1
ε2 emissivity of body 2

The heat radiation exchange factor C12 is a quantity combining the emissivity of
the bodies as well as their geometries (shape factors) and relative surface absorptiv-
ity. If, for example, we have two parallel plates with identical geometry and surface
absorptivity emitting/absorbing heat radiation, then the heat radiation exchange
factor would consist of the following expression, C12 [9]:
σ
C 12 = ð8:56Þ
1
ε1 þ ε12 - 1

8.6.1.1 Emissivity
The emissivity of a material gives an indication of its ability to emit electromagnetic
radiation (in this case, heat radiation). A corollary property is absorptivity, which
indicates the ability of a material to absorb heat radiation. When a material is in
thermal equilibrium, we can assume that its emissivity and absorptivity are equal.
This is known as Kirchhoff’s Law of thermal radiation.
An ideal black body is defined as a body having maximum absorptivity, with a
value for absorptivity of ε = 1. An ideal reflective body is incapable of absorbing any
heat radiation, and has a value for absorptivity of ε = 0. All real bodies are known as
“grey bodies,” and have values for absorptivity between zero and one.
If we know the emissivity ε of a material body and its surface temperature, we can
calculate the energy lost from the body caused by radiation heat transfer. Likewise, if
we do not know the temperature but can measure the quantity of radiant heat energy
emitted, we can calculate the surface temperature of the body. This is the principle
for measuring surface temperatures by infra-red thermometry (IR thermometers, see
also Chap. 16).
If we want to tabulate values for the emissivity of materials, the surface condition
of the material must be specified. Thus, a polished metal surface can have ε below
0.05 while the corroded surface of the same material has a value of ε > 0.75. With
the help of suitable coatings, the emissivity of materials can be controlled [10]. This
is especially important and useful with packaging materials. The emissivity of liquid
water is around ε = 0.95 and unpackaged foods also have values of this
magnitude [11].
8.6 Heat Transfer in Food 335

8.6.2 Heat Conduction Transfer

Heat transfer by conduction was first mentioned in Chap. 6 as an example of other

diffusion-like transport phenomena that can be described with the same type of
mathematical equation as that used to describe molecular diffusion through a
material substance.
When we first approach the study of heat conduction, it is important to make a
distinction between heat conduction under steady state conditions (temperatures
remain constant at any point over time while heat is flowing) and unsteady state,
or transient, conditions (temperatures at any point change over time while heat is
flowing). In this section, we will limit our study to steady state heat conduction. This
is the situation encountered most frequently in food processing involving the heating
and cooling of liquid food products through heat exchangers and holding tubes.
These heating and cooling systems operate under steady state conditions because
product and heat exchange medium temperatures at the entrance and exit of these
systems remain constant over time while conduction heat transfers at a steady
(constant) rate. Most heat exchangers used in food processing operations are either
a plate-type (made up of flat plates) or tubular-type (made up of cylindrical tubes). In
a plate heat exchanger, liquid product flows through a narrow space between two
parallel stainless-steel plates. On the other side of these plates, a heat exchange
medium at a different temperature is flowing (e.g., hot water or condensing steam).
This situation causes heat to flow from the high-temperature side to the
low-temperature side causing the temperature of the cool incoming product to rise
as it exchanges heat with the hot heat exchange medium fluid on the other side. In a
tubular heat exchanger, the product flows through a narrow cylindrical tube to
exchange heat with a flowing fluid heat exchange medium on the other side of the
tubular wall. For this reason, we will first consider heat conduction across a flat plate
“wall” barrier. Then, we will consider heat conduction across a cylindrical “wall”
barrier.
Heat transfer through a material can occur in all three directions (dimensions) of
space. However, in heat exchanger applications, the temperature gradient across the
thin stainless steel wall drives the heat transfer predominantly in the one direction
crossing the wall, and there is little or no temperature gradient (potential) causing
heat to travel along the metal wall material in either of the other two directions.
Therefore, we can limit our analysis of heat conduction to the simplest case of
one-dimensional steady state heat conduction across a flat plate wall.

8.6.2.1 One-Dimensional Steady State Heat Conduction Across a Flat

Plate
Let us consider heat travelling across a solid flat plate by conduction like in Fig. 8.5
we can express the quantity of heat flow by Fourier’s first law:
336 8 Thermal Properties

Fig. 8.5 Temperature profile

across a solid flat plate O
T0

T1
T

G x

Table 8.12 Sign of temperature gradient and resulting heat flow in Fig. 8.5
Case Temperature gradient Description
x>0 Positive T is increasing with increasing x
T1 > T0
x>0 Negative T is decreasing with increasing x
T1 < T0

Q_ dT
= -λ
ð8:57Þ
A dx
with the symbols of Fig. 8.5

Q_ T - T1
=λ
0 ð8:58Þ
A δ
so

T - T1
Q_ = A
λ
0 ð8:59Þ
δ

dT
dx
temperature gradient in K
m-1
λ heat conductivity in W
K-1
m-1
Q_ heat flow in W
Q_ heat flow density in W
m-2
A

The negative sign in Fourier’s law is necessary in order to have a positive heat
flow in response to a negative temperature gradient. This is because heat will only
flow from high to low temperature, which is “down hill,” or having a negative slope.
Table 8.12 shows the type of responding heat flow to positive and negative temper-
ature gradients.
8.6 Heat Transfer in Food 337

The heat transferred can be calculated by integration over time:

dQ
Q_ = ð8:60Þ
dt
or

ðT 1 - T 0 Þ
dQ = - A
λ

dt ð8:61Þ
δ
it is

t
λ
Q= A

ðT 0- T 1 Þ
d t ð8:62Þ
δ
t=0

The λ is the thermal conductivity of the material of the solid wall is the same. The
reciprocal quantity is a resistance to heat conduction, it is called specific thermal
resistivity. Materials designed to minimize heat conduction must have low thermal
conductivity (e.g., materials for thermal insulation on buildings or pipelines).
Materials with high thermal conductivity are suitable for achieving rapid heat
transfer in heat exchangers for heating or cooling.

8.6.2.2 Three-Dimensional Steady State Heat Conduction

In the most general case, when temperature gradients exist in all three dimensions of
space, heat will transfer through a material in all three directions. In this situation,
Fourier’s law must be expressed in the form of a partial differential equation to
include terms for the temperature change in each of the three directions x,y,z:

Q_ ∂T ∂T ∂T
= -λ
, , ð8:63Þ
A ∂x ∂y ∂z

or

Q_ ∂ ∂ ∂
= -λ
, , T = - λ
grad T ð8:64Þ
A ∂x ∂y ∂z

or with the abbreviation

∂ ∂ ∂
∇T = grad T = , , T ð8:65Þ
∂x ∂y ∂z

shortly

Q_
= - λ
∇T ð8:66Þ
A
where ∇ is called Nabla or dell operator in matrix algebra.
338 8 Thermal Properties

Fig. 8.6 Temperature profile O1 O3

across a multilayer solid
flat wall
T0

T1
T T2
T3

G1 G2 G3
x

8.6.2.3 One-Dimensional Steady State Heat Conduction Across Multiple

Layers
Figure 8.6 illustrates the case of conduction heat transfer occurring across a flat wall
that is made up of several layers of different materials, with each material having a
different thermal conductivity λ and different layer thickness δ. We can calculate the
total heat flow through the wall with help of the concept of additive resistors, refer to
Chap. 6.
In this case, the mathematical expression for the quantity of heat flow can be
derived as follows. Because the heat flowing through all three layers is the same:

Q_ 1 = Q_ 2 = Q_ 3 ð8:67Þ

Q_ 1 Q_ 2 Q_ 3
= = ð8:68Þ
A A A
with

Q_ 1 δ1
T0 - T1 =

A λ1

Q_ 1 T - T0
= -λ
1 ð8:69Þ
A δ1
and

Q_ 2 δ2
T1 - T2 =

A λ2

Q_ 2 T - T1
= - λ2
2 ð8:70Þ
A δ2
8.6 Heat Transfer in Food 339

and

Q_ 3 δ3
T2 - T3 =

A λ3

Q_ 3 T - T2
= - λ3
3 ð8:71Þ
A δ3
By adding the temperature differences from Eqs. (8.69–8.71) we get the total
temperature difference Δ T

Q_ δ δ δ
Δ T = T0 - T3 =
1þ 2þ 3 ð8:72Þ
A λ1 λ2 λ3

So, we get for heat flow density:

Q_ 1
=
ΔT ð8:73Þ
A δ1
þ δ2
þ δλ33
λ1 λ2

Writing a more general expression for the steady state case with n layers we have:

for the heat flow density for the heat flow for the heat quantity
Q_
= 1
n
ΔT Q_ = A
n1
Δ T Q = A
n1
Δ T
t
δi
A δi δi
λi λi λi
i=1 i=1 i=1

When we have different sizes of areas for each layer, we write for the heat flow
1
Q_ = n
ΔT ð8:74Þ
δi
λi
Ai
i=1

A comparison of Eq. (8.74) with the related expressions in Table 7.7 shows that
the term in the denominator is equal to the sum of the heat resistances R of each
layer.
The heat conduction resistance R results from the reciprocal thermal conductivity
and the geometric ratio of each layer Aδ :

δ
R= ð8:75Þ
λ
A
The concept of addition of resistances can also be used here in the case of steady
state multilayer conduction of heat. We can write this in a more abbreviated way as
follows:
n n
δi
= Ri = Rges ð8:76Þ
i=1
λi
A i i=1

With Rtotal First Fourier’s Law gets the appearance of Ohm’s Law:
340 8 Thermal Properties

1
Q_ =
ΔT ð8:77Þ
Rtotal
Respectively

Δ T = Rtotal
Q_ ð8:78Þ

λ heat conductivity in W
K-1
m-1

δ thickness in m
Q_ heat flow in W
Q heat in J
A area in m2
R heat resistance in K
W-1
i layer index
n total number of layers

8.6.2.4 One-Dimensional Steady State Conduction Across a Single Layer

Cylindrical Wall
Figure 8.7 illustrates the case when heat must transfer across a cylindrical wall, such
as when tubular heat exchangers are used. In this case, the same expression for
Fourier’s law still applies except that the total wall thickness must be expressed as
the difference between outer and inner tube radii, and the incremental distance along
the tube wall thickness must be expressed as radius r.
Expressing the Fourier law for an infinitesimally thin cylinder wall

Q_ dT
= -λ
ð8:79Þ
A dr
we get the heat flow by integration over the cylindrical wall dr:

Fig. 8.7 Temperature profile

across a single-layer
cylindrical wall, schematic

T0

T1
r0
r1
8.6 Heat Transfer in Food 341

Q_ Q_
dT = - dr = - dr ð8:80Þ
A
λ 2π
l
r
λ

T1 r1 r1
Q_ Q_ dr
dT = - dr = - ð8:81Þ
2π
l
r
λ 2π
l
λ r
T0 r0 r0

Q_ r
T1 - T0 = - ln 1 ð8:82Þ
2π
l
λ r 0
so

2π
l
Q_ =
λ
ðT 0- T 1 Þ ð8:83Þ
ln rr10

An alternative calculation for the heat conducted in Fig. 8.7 is to use an average
area Am. The value of Am can be calculated by use of an average radius rm. For the
simplest case in a thin-walled tube, we can use the arithmetic mean between outer
and inner diameter for the average radius rm

r1 þ r0
rm = ð8:84Þ
2
But, let us first calculate a general average radius rm between r0 and r1. Equation
(8.79) is going to be:

Q_ Q_ λ
- = =
ðT 1- T 0 Þ ð8:85Þ
A m 2 π r m
l ðr 1 - r 0 Þ

For the heat flow, we get:

2 π
rm
l
Q_ =
λ
ðT 0- T 1 Þ ð8:86Þ
ðr 1 - r 0 Þ

Comparing this with Eq. (8.83) provides:

2π
l 2 π
rm
l

λ
ðT 0- T 1 Þ =
λ
ðT 0- T 1 Þ ð8:87Þ
ln rr10 ðr 1 - r 0 Þ

so

r1 - r0
rm = ð8:88Þ
ln rr10

So, now we have an algorithm to calculate rm for all type of tubes. This type of
average radius rm is called the logarithmic mean radius. It can be used for all cases of
342 8 Thermal Properties

ra 2˜r i r a  2˜ r i

I II III

Fig. 8.8 Thick walled tube (I) and thin walled tube (III). II is borderline between I and III

tubes. In case of a thin-walled tube, we learned we could use the arithmetic average.
But how can we decide what is a thin-walled or thick-walled tube? Figure 8.8
illustrates that a thin-walled tube is a tube with outer radius smaller than twice the
inner radius (ra < 2 ri). The opposite case is a thick-walled tube when the outer
radius is greater than twice the inner radius (ra > 2 ri), and we call a tube thick
walled. Between both cases we have the case in which the outer radius is precisely
twice the inner radius (ra = 2 ri), which presents a borderline situation, and either
approach can be used.
At this point, we have learned that usage of the arithmetic mean is an approxima-
tion, and usage of the logarithmic mean is exact. But how big is the error we might
get by using the approximation from the arithmetic mean instead of the exact result
from the logarithmic mean? Let us compare: We calculate for case II in Fig. 8.8

Logarithmic mean Arithmetic mean

r m = raln-rari rm = r a þri
2
ri

with
ra = 2ri
is
2 ri - ri 2 r i þr i
rm = 2r rm = 2
ln r i
i

rm = ri
ln 2 r m = 32 r i
rm = 1.44
ri rm = 1.5
ri

Therefore, relative error is:Δrmrm = 1:5
r1:44

i - 1:44
r i

r i = 0:06
1:44 = 4%
This means that if, for reasons of simplification, the arithmetic mean value for the
mean radius I m is used instead of the logarithmic mean, the value calculated in this
way for the radius of thin-walled pipes is a maximum of 4% below the correct value.
The thinner the pipe wall, the smaller the difference between arithmetic and loga-
rithmic mean. In the case of thick-walled pipes, the thicker the pipe wall, the greater
the difference, which is always greater than 4%. The approximation “arithmetic
mean instead of the logarithmic mean” is therefore suitable for thin-walled tubes and
8.6 Heat Transfer in Food 343

rather not suitable for thick-walled pipes. Considerations of this kind allow the
quantification of the error caused using an approximation. By knowing the error
quantitatively, better decisions can be made. There are cases in which an error of 4%
is not tolerable, but also cases in which, for example, due to other uncertainties, an
error as great as 10% can be tolerated.
For the cylindrical, one-dimensional, steady state, single-layer case we get:

Heat flow density Heat flow Heat

Q_
= λ

ðT i- T a Þ Q_ = Am
ra -λ ri
ðT i- T a Þ Q = Am
ra -λ ri
ðT i- T a Þ
t
Am r a - ri

8.6.2.5 One-Dimensional Steady State Conduction with Multi-layer

Cylindrical Wall
Figure 8.9 illustrates the case of conduction heat transfer occurring across a cylin-
drical wall that is made up of several layers of different materials, with each material
having a different thermal conductivity λ and different layer thickness δ. In this case,
the general mathematical expression for the quantity of heat flow can be expressed
the same as in the case for a flat plate wall, refer to Eq. (8.74).
Starting with the equation from the flat plate multilayer case:

1
Q_ = n
ΔT ð8:89Þ
δi
λi
Ai
i=1

The values of the areas Ai depend on the inner and outer radius of each layer.
Therefore, they are not the same for all the layers. They must to be added separately.
The denominator of Eq. (8.89) is the sum of all n heat resistances.
n n
δi
= Ri ð8:90Þ
i=1
λi
Ai i=1

This is the sum of the n involved thermal conduction resistances Ri. Again, we
have a case in which to use the concept of adding resistances.

Fig. 8.9 Multilayer

cylindrical solid wall

r2
r1
r0
344 8 Thermal Properties

However, the areas Ai depend on their respective radii ri and therefore are not the
same size for all layers. For this reason, the areas Ai must be included individually in
calculating of the heat flow. With the designations rn and rn - 1 for outer and inner
radii and rm, n for the average radius of each layer, and with Eq. (8.88) in the form of

rn - rn - 1
rm = ð8:91Þ
ln rnr-n 1

we get
1 2π l
Q_ = rn - rn - 1
ΔT = rn - rn - 1
ΔT ð8:92Þ
λn
2 π rm,n
l λn
rm,n

the denominator alone is:

rn - rn - 1 ðr n - r n - 1 Þ
ln rnr-n 1 ln rnr-n 1
= = ð8:93Þ
λn
r m,n λ n
ðr n - r n - 1 Þ λn

So, for the heat flow we get:

2π l
Q_ = ln r rn

ΔT ð8:94Þ
n-1
λn

Written as sum of heat conduction resistances Ri

1
n ln rnr-n 1 n

= Ri = Rtotal ð8:95Þ
2π
l i=1 λn i=1

we get again
ΔT
Q_ = ð8:77Þ
Rtotal
Recall that in addition to heat transfer by radiation and conduction, we also have
heat transfer by convection and phase transition as alternative mechanisms by which
to transfer heat (Table 8.11). So, to calculate heat transfer in general we have more
properties and parameters yet to study, which are only mentioned briefly here.

8.6.3 Convection Heat Transfer

Convection heat transfer occurs when heat energy is carried along by a moving fluid
in contact with a solid surface. The fluid can be either liquid or gas. For example, we
feel warm when we enter a heated room from the cold outdoors because the surface
of our body encounters warm heated air that circulates in the room. Likewise, the air
in the room receives heat when it encounters the heated surface of a metal radiator.
8.6 Heat Transfer in Food 345

The inside metal walls of the radiator become heated when hot water flows in contact
with the inner surface of these walls. The means by which heat transfers from the
flowing hot water to the inside metal surface of the radiator, as well as from the outer
metal surface of the radiator to the circulating air in the room, is heat transfer by
convection.
In the case of convection heat transfer, the fluid experiencing heating or cooling
also moves. This movement may be due to the natural buoyancy effect of decreased
density with increased temperature (natural convection), or it may be caused artifi-
cially by imparting mechanical energy to the fluid, such as with pumps or blowers
for liquids and gases, respectively (forced convection).
Determining the rate of heat transfer from convection is complicated because of
this fluid motion. When a fluid is flowing over a solid surface, shear stresses occur in
the fluid near the surface because of the viscous properties of the fluid. Molecules at
the surface try to attach themselves to the surface while neighboring molecules are
trying to pass them by within the bulk fluid flow. This causes a velocity profile to
develop near the surface. The fluid next to the surface does not move but sticks to it,
while neighboring fluid flow is slowed down by the friction of trying to pass the
stationery molecules. Therefore, fluid velocity gradually increases with distance
away from the surface until the region of bulk fluid flow is reached, where the
fluid velocity is all the same at the maximum. This region is known as a boundary
layer.
Along with the velocity profile, a temperature profile develops in this boundary
layer near the wall. This is because the rate of convective heat transfer from a fluid to
a solid surface depends in part on the relative velocity of the fluid in contact with the
surface. Since this relative velocity is near zero at the surface because of the sticky
viscous effects, transfer of heat is also poor. This means the surface does not readily
sense the temperature in the bulk fluid flow, and the temperature will change with
distance away from the surface as it moves from surface bulk fluid temperature.
Therefore, this invisible boundary layer region near the surface acts as though it were
an additional layer of insulating material, adding further resistance to the heat
transfer between the fluid and the surface (Fig. 8.10).

8.6.3.1 Heat Transfer Coefficient

Mathematically, we can account for this convective heat transfer resistance by
assuming it represents another layer of material across which heat must transfer in
the expression for steady state heat conduction from Fourier’s first law. We can do
this by assigning a value for the “thermal conductance” of this invisible boundary
layer (Fig. 8.10), which is known as the surface heat transfer coefficient α. Then we
could express the simple case of heat transfer across such a boundary layer from a
hot surface temperature Ts to a cool bulk fluid temperature T1 in the following way:
346 8 Thermal Properties

Fig. 8.10 Convection heat 1

transfer. Showing boundary
layer where velocity and 2
temperature profiles exist near
wall surface, 1 hot wall,
2 boundary layer velocity
TS

temperature

Fluid Flow

Q_ = α
A
ðT S- T 1 Þ ð8:96Þ

Q_ heat flow in J
s-1

α heat transfer coefficient in W
m-2
K-1
A area in m2
TS temperature of wall in K
T1 temperature of bulk fluid in K

It is important to note that the heat transfer coefficient is a numerical value that
represents the overall “thermal conductance” of the invisible boundary layer. Its
value will depend on the combined effects of the physical, thermal and viscous
properties of the fluid in contact with the surface, relative velocity of the fluid at the
surface, as well as system geometry at the contacting surface, among other things.
Therefore, the heat transfer coefficient is not a material property, and cannot be
looked up in a handbook. It is a parameter (coefficient) in heat transfer equations that
depends on conditions of the process system under study. Normally, it is determined
experimentally for heat exchanger systems by running experiments under controlled
conditions, where all temperatures, surface areas, material properties and flow rates
are known, and the heat transfer coefficient is the only unknown. There are also
various approaches for attempting to estimate values for the heat transfer coefficient
under different specified sets of heating and cooling conditions and surface contact
geometries. These approaches can be found in references devoted more completely
to heat transfer and fluid mechanics [12, 13].
8.6 Heat Transfer in Food 347

8.6.3.2 Overall Heat Transfer Coefficient

The convective boundary layers are only part of the multi-layer systems across
which heat must transfer in more realistic heating and cooling situations. Recall
our radiator for heating the room. The air in the room receives heat when it
encounters the heated surface of the metal radiator. The inside metal walls of the
radiator become heated when hot water flows in contact with the inner surface of
these walls. Let us take a close look at just a small section of radiator wall that we
could imagine as a flat plate (thermal conductivity λ and thickness δ) with hot water
flowing along the inside surface, and cool room air flowing along the outside surface.
In this case, we have a boundary layer on the inside caused by the flowing hot water
represented by a heat transfer coefficient α1. We also have a boundary layer on the
outside caused by the flowing cool air represented by a heat transfer coefficient α2.
This combination of heat transfer 1, heat conduction and heat transfer 2 can be
characterized by a single parameter called overall heat transfer coefficient k.
To understand the idea of the overall heat transfer coefficient let us look again in
the concept of adding resistors. The heat has to pass resistances of boundary 1, of the
flat plate and of boundary 2 one after the other. That means we can add the
resistances like in a serial circuit to get the total resistance against heat flow:
Rtotal = R1 þ Rλ þ R2 ð8:97Þ

The heat transfer resistance R is the reciprocal of the heat transfer resistance
coefficient α. The heat conduction resistance R is the reciprocal of the quotient of
heat conductivity λ and thickness δ of the solid material. So we can write
1 δ 1
Rtotal = þ þ ð8:98Þ
α1 λ α2
So Rtotal is the total resistance against heat flow causes by serial circuit of
boundary 1, plate and boundary 2.
The reciprocal of the total resistance is the overall heat transfer coefficient k.
1
Rtotal = ð8:99Þ
k
so

1 1 δ 1
= þ þ ð8:100Þ
k α1 λ α2
respectively
1
k= ð8:101Þ
1
α1 þ δλ þ α12

To calculate the heat flow through several heat resistances, we can use the overall
heat transfer coefficient k in an equation analogous to Fourier’s first law:
348 8 Thermal Properties

Fig. 8.11 Multilayer heat

transfer with convection on
both sides. Example: Material
2 is steel, 1 is rust and 3 is D D2
paint. The overall heat transfer 1
coefficient k combines α1, α2,
λ1, λ2, λ3,δ1,δ2 andδ3 in a
single parameter
hot O1 O2 O3 cold
water air

G1 G2 G3

Q_ = A
k
ΔT ð8:102Þ

Q_ heat flow in J
s-1

A area in m2
α heat transfer coefficient in W
m-2
K-1
k overall heat transfer coefficient in W
m-2
K-1
ΔT temperature difference in K
λ thermal conductivity in W
m-1
K-1
δ thickness in m

Let us imagine further that the radiator wall has a layer of rust on the inside and a
layer of paint on the outside in addition to the metal core between these rust and paint
layers. Each of these material layers would have its own thermal conductivity and
thickness. Then we would have a multilayer situation consisting of five layers. This
would be a situation similar to that shown earlier in Fig. 8.6 but with two additional
boundary layers with their respective heat transfer coefficients on each side, as
shown in Fig. 8.11.
So the total resistance is
Rtotal = R1 þ Rλ1 þ Rλ2 þ Rλ3 þ R2 ð8:103Þ

i.e.

1 δ1 δ2 δ3 1
Rtotal = þ þ þ þ ð8:104Þ
α1 λ1 λ2 λ3 α2
8.6 Heat Transfer in Food 349

the overall heat transfer coefficient therefore is

1
k= ð8:105Þ
1
α1 þ δ1
λ1 þ δ2
λ2 þ δλ33 þ α12

For an n-layer solid the overall heat transfer coefficient is

1
k= n ð8:106Þ
δi
αouter þ þ αinner
1 1
λi
i=1

Process engineers often use Eqs. (8.106) and (8.102) for heat transfer calculations
involving heat exchangers. Often it is much too difficult to estimate or measure the
exact thermal conductivity and thickness of various material layers in a system,
especially if they consist of poorly defined rust and scale deposits or fouling layers. It
is also difficult to correctly characterize the fluid boundary layers from convective
heat transfer in order to estimate convective surface heat transfer coefficients. The
overall heat transfer coefficient k considers all these factors. It is most often deter-
mined experimentally for heat exchanger systems by running experiments under
controlled conditions, where all temperatures, surface areas, and flow rates are
known, and the overall heat transfer coefficient k is the only unknown. For more
details, see hydrodynamics and process engineering [6, 14–16]. In the field of
building construction and thermal insulation, there are regulations and standardized
measurement methods for the heat transfer coefficient, too [17–19].

8.6.4 Heat Transfer by Phase Transition

We learned earlier from our review of basic thermodynamics that when we look at a
simple phase transition, such as when a substance undergoes evaporation from a
liquid to a vapor, freezing from a liquid to a solid, or condensing from a vapor to a
liquid, then we normally see significant changes in enthalpy, entropy and volume.
These are caused by the change of enthalpy ΔH in the form of latent heat of fusion or
latent heat of vaporization. These enthalpy changes involve significant quantities of
heat transfer to occur in support of the phase change. Because of these significant
quantities of latent heat energy transferred during phase change, process engineers
prefer to use condensing steam (water vapor) whenever practical as a heat exchange
medium in heat exchanger systems.

Example
When 1 kg of saturated steam at 100 °C condenses onto the cool metal surface
of a heat exchanger, it immediately gives up its latent heat of vaporization as
the enthalpy changes from the value at saturated vapor (2676 kJ
kg-1) to that

(continued)
350 8 Thermal Properties

of saturated liquid (419 kJ
kg-1). This represents a net heat transfer of

(at normal pressure) 2257 kJ
kg-1.
If instead, we used 1 kg of liquid hot water at 100 °C as the heat exchange
medium, it would probably experience a change in temperature of about 30 K,
and the heat given up would be the sensible heat loss. We would get
Q = cp
ΔT = 4.2 kJ
K
kg-1
30 K = 126 kJ
kg-1 and thus less than
1
10 of the heat of condensing water vapor.

Clearly in the case of condensing water vapor, the amount of heat transferred
from the same quantity of steam or water is two orders of magnitude greater for
condensing steam (phase transition) than when water is used alone without any
phase change.
The enthalpy change occurring during condensation corresponds to the enthalpy
change occurring during evaporation (with the opposite sign). Therefore, the heat
released during condensation can be calculated from the enthalpy of evaporation:
jQsteam j = m
Δhvap ð8:107Þ

Q heat in J
h specific enthalpy in J
kg-1
Δhvap specific evaporation enthalpy in J
kg-1
m mass in kg
cp specific heat capacity in J
kg-1
K-1
T temperature in K

In addition to the condensation of water vapor, the evaporation of water is also

used for heat transfer—i.e., for cooling. In so-called adiabatic cooling, the tempera-
ture of the air is lowered by spraying water with nozzles into small droplets and let it
evaporate. In addition, many power plants use the evaporation of water to cool their
cooling water. The solid-liquid phase change of water has also long been used for
heat transfer, e.g., in ice cellars or cold baths. Modern concepts of buildings use ice
storage as a latent heat source [20]. A pumpable ice-water mixture can be used to
cool buildings or products such as food [21].

Remark
The enthalpy of evaporation of water at normal pressure is 2257 kJ
kg-1.
Methane (CH4), a substance with a molecular weight similar to that of water,
has an enthalpy of evaporation of only 510 kJ
kg-1. The cause of the high
value of water lies in its molecular structure. Due to the dipolar character of the
H2O molecule, numerous hydrogen bonds are formed in the liquid phase,
which must be dissolved during evaporation. This is not the case with a
nonpolar molecule such as CH4.
8.6 Heat Transfer in Food 351

In order to carry out heat transfer calculations involving phase change, it is

necessary to know the thermodynamic properties of the material and which phase
transition it is undergoing. In the case of condensing steam, we would need the
thermodynamic properties of saturated steam. These properties include enthalpy,
entropy and specific volume of both saturated liquid and saturated vapor as a
function of temperature and pressure. These are material properties, but they depend
on temperature and pressure. Therefore, they are normally published in hand books
and text books in the form of tables or charts for easy reference [22]. An example of
such a steam table can be found in the Appendix.

8.6.5 Thermal Conductivity

In previous chapter sections, we described conduction heat transfer through solid

materials (across heat exchanger walls), and how we use Fourier’s law to derive
mathematical expressions that characterizes conduction heat transfer. You will recall
that the thermal conductivity λ was the singular coefficient representing material
properties in Fourier’s law. In the case of conduction heat transfer, this material
property is called thermal conductivity, and indicates how easily heat will pass
through the material. Different materials have different thermal conductivities.
This difference helps to explain why a metal spoon gets too hot to hold in your
hand when stirring a boiling liquid in a pan, but a wooden spoon of the same size and
dimensions does not. We have already been using the thermal conductivity as the
coefficient λ in Fourier’s first law:

Q_ dT
= -λ
ð8:108Þ
A dx
The reason different materials have different thermal conductivities is because
they have different chemical and physical compositions. Recall that heat is
conducted through solid materials because the increasing kinetic energy imparted
to the molecules at the point where the heat is entering the material excites them to
oscillate at greater speeds and amplitudes. This excitation is driven by higher
electron mobility within the atomic structure of each molecule, and causes the
molecules to experience increased “bumping” into neighboring molecules. This
molecular “bumping” propagates along the material, and further continues this
process of heat transmission (conduction) along the material substance. Therefore,
thermal conductivity must depend on both molecular structures at the atomic
electron level (the chemical composition), as well as the physical lattice structure
by which the molecules are held in place within the material substance (physical
structure). In the case of liquid and gaseous substances, heat convection occurs in
parallel in addition to heat conduction. In solids where convection cannot take place,
heat can only be transported by heat conduction. For this reason, we first consider
solid substances. These can be foodstuffs and materials for packaging technology,
e.g., polymers or materials for mechanical engineering, e.g., stainless steel.
352 8 Thermal Properties

8.6.5.1 Solids
In metal solids, much of the molecular excitation that occurs in response to heat is
due to the relatively high mobility of the electrons in metals. Therefore, this electron
mobility accounts for the fact that in metals, electron conductivity is the major
contributing mechanism for the relatively high thermal conductivity found in metals.
The ratio of electric conductivity to thermal conductivity is a linear function of
temperature (Wiedemann–Franz-law):

λ
=L
T ð8:109Þ
σ

λ thermal conductivity in W
K-1
m-1

σ electric conductivity in S
m-1
L Lorenz-coefficient (for metals at room temperature is L ≈ 2.4
10-8 V2
K2)
T temperature in K

Disturbances of crystal structure such as degradations and defects reduce the

thermal conductivity. Therefore Aluminum alloys or stainless steel have a lower
thermal conductivity than pure aluminum or pure iron. Among the metals, silver and
copper have the highest values for thermal conductivity. Selection of a material for
critically functioning components or equipment such as heat exchangers is always
made as a compromise between mechanical and thermal properties, corrosion
resistance and price. For this reason, stainless steels are often chosen for food
processing equipment over pure metals. When packaged food is heated for the
purpose of preservation (refer Table 8.1), the heat transfer to the core of the food
is also influenced by the thermal conductivity of the packaging material. Therefore,
different processes are necessary for glass or plastic packages than for sheet metal
packages. Even with non-metallic materials and polymers, thermal conductivity
decreases with decreasing crystallinity of the material.
In solid foods, thermal conductivity is strongly determined by their water content.
Empirical calculation equations for the thermal conductivity of solid foods therefore
contain the water content of the material as a key variable. Water in solid foods is
more or less immobilized (refer to Chap. 1). For this reason, heat transport in solid
foods does not take place by convection but by heat conduction.

Example
We have made a jelly from a fruit juice (water content 90% wb) by adding
gelatin (The water content is also 90% wb). We put a glass of jelly and a glass
of the original fruit juice in the refrigerator and record the core temperature of
both samples. It turns out that the temperature in the juice drops faster. The
juice seems to transport the heat better.

(continued)
8.6 Heat Transfer in Food 353

Explanation: In addition to heat conduction, in the juice heat convection

also occurs, i.e., the liquid in the glass is mixed by natural convection as
increasing density of the cold juice at the surface causes it to sink to the bottom
while the less dense warmer juice at the center rises upward. In jelly, heat is
transported by heat conduction only. Both samples have the same geometry
and the same mass. The only difference is that in the jelly sample there is a gel
network that immobilizes the water molecules making it behave as a
solid body.

In the Annex values for the thermal conductivity of some foods and materials can
be found. For further values see, e.g. [23–26].
Since gases have a much lower thermal conductivity than liquid phases, the value
of the thermal conductivity of food drops sharply during the formation of foam.

Definition
Foam is a disperse system. The dispersed phase is gaseous. Foamed foods are
produced by creating gas bubbles in the food matrix. Examples are whipped
cream, ice cream, sugar foam but also baked goods. The gas pores are usually
filled with air, but also with N2O (whipped cream) or CO2 (by baking soda or
yeast in baked goods).

For foamed materials air content is therefore another variable for the calculation
of thermal conductivity. If a material contains pores, a decrease in thermal conduc-
tivity is observed with increasing porosity. The pore radius distribution also plays a
role in affecting thermal conductivity. A distinction is made here between open-pore
systems, in which the pores have connection to the ambient atmosphere and closed-
pore systems. In open-pored systems, there is a clear dependence of the thermal
conductivity on the relative humidity of the environment, as this determines the
surface cover and capillary condensation, i.e., the water content in the pores (refer
Chap. 1). For these reasons, solid materials for thermal insulation often consist of
closed-porous, non-crystalline polymers with high porosity.
In isotropic materials, the physical properties are the same in all spatial directions.
For non-isotropic materials, the physical properties depend on the spatial direction.
In the case of thermal conductivity, this is observed with fibrous materials based on
plants or animals. As a rule of thumb for meat, the thermal conductivity parallel to
the fiber is 7–10% higher than perpendicular to the fiber (Fig. 8.12).
354 8 Thermal Properties

Fig. 8.12 Thermal

conductivity of water (H2O),
butter (B) and turkey meat
(T). T: upper curve indicates λ
parallel to fiber, lower curve λ H2O
perpendicular to meat fiber [6] O

T
B

-10 0 10
-/ qC

Definition
A material is called isotropic if the physical properties are the same in all
spatial directions. Anisotropic materials are non-isotropic, i.e., some physical
properties, e.g., optical or mechanical properties depend on the spatial direc-
tion. Fibers and fibrous materials are examples of anisotropic materials.

8.6.5.2 Multilayer Solids

The thermal conductivity of solids which consist of several layers with different
thermal conductivities can once again be calculated using the concept of additive
resistors again. Heat conduction resistivity is the reciprocal of thermal conductivity.
In the case of series connection, thermal conduction resistances are added to the total
resistance. If the thermal conductive resistances are parallel to each other, not the
resistances R are added, but their reciprocal size R1, the so-called heat conductance is
added to get the total heat conductance Gtotal. In Table 8.13 the cases are listed side
by side. In Chap. 7 we already learned about the analogous mathematical treatment
of steady state heat conduction and conduction of electricity. To get used to work

Table 8.13 Total resistance to heat conduction at serial connection and parallel connection
Serial connection Parallel connection
Rtotal = Ri 1
Rtotal = Gtotal Gtotal = Gi
i i
addition of resistances addition of conductances
Rtotal = di
Ai
λ1i Gtotal = d i λi
Ai
i i
when A1 = A2 = A3...
and d1 = d2 = d3...
it is
λtotal =
1 1
λi
λtotal = λi
i i
addition of reciprocal thermal conductivities addition of thermal conductivities
8.6 Heat Transfer in Food 355

Table 8.14 Analogy of heat conduction electrical conduction: terms

Heat conduction Electrical conduction
heat conductivity λ in W
K-1
m-1 Electrical conductivity κ in S
m-1
λ = Rheat
1

Ad κ = 1ρ = R1el
Ad
Q_ = κ
ΔU Ohm’ s law
= λ
ΔT Fourier’ s first law
I
A d A d
Heat conductance in W
K-1 Electrical conductance in S
Rheat = λ
d Rheat = Gheat Rel = ρ
d Rel = Gel
1 A 1 1 1 A 1

Heat conduction resistivity in K
m
W-1 Electrical resistivity in Ω
m

λ = Rheat
d ρ = Rheat
Ad
1 1 A 1

Heat conduction resistance in K
W-1 Electrical resistance in Ω

Rheat = 1λ
Ad Rel = ρ
Ad
Rheat = ΔTQ_
Fourier’ s first law Rel = UI Ohm’ s law
Q_ heat flow in W
λ specific heat conductivity in W
K-1
m-1
A area in m2
d thickness in m
ΔT temperature difference in K
Rheat heat conduction resistance in K
W-1
Gheat heat conductance in W
K-1
κ electrical conductivity in S
m-1
ρ electrical resistivity in Ω
m
U voltage in V
I electrical current in A
Rel electrical resistance in Ω
Gel electrical conductance in S
xi mass fraction of component i

with heat conduction resistances, it is good to recall the analogy between Ohm’s law
and Fourier’s first law. Table 8.14 can help to remember the related terms.
This principle can also be applied to mixtures of solids, such as foods made of
several different solid ingredients with different thermal conductivities λi. For
mixtures of solids the parallel connection applies, i.e., the thermal conductivities
are added together and weighted according to their relative mass fractions xi. In
Table 8.15 some Indicative values for λi are listed.

λges = xi
λ i ð8:110Þ
i

If the ingredient formulation of a food is imagined to be simplified to only water

and dry matter, the thermal conductivity of such composite systems can be quickly
added.
With the mass fractions
356 8 Thermal Properties

Table 8.15 Indicative λ (W
K-1
m-1)

values for the thermal con-
Air 0.025
ductivity of food
components [6] Protein 0.20
Carbohydrate 0.245
Fat 0.15
Water, liquid 0.55
Ice 2.21
Non-fat dry matter from animals 0.26
Non-fat dry matter from plants 0.22

xdm þ xH2 O = 1 ð8:111Þ

the thermal conductivity is

λtotal = xdm
λdm þ xH2 O ð8:112Þ

Values for λdm can be found in Table 8.15. In the case of frozen foods, the water
content of the food is split into a frozen fraction α and a non-frozen fraction (1 - α).
Then we have
λtotal = xdm
λdm þ xH2 O ð1- αÞ
λH2 O þ xH2 O
α
λice ð8:113Þ

The determination of the frozen fraction α can be carried out experimentally,

e.g. by DSC, see Sect. 8.8.2 or by NMR, see Chap. 11.

Example
A sample of fruit concentrate (dry matter: 47% (m/m) has at -20 °C a frozen
fraction of 80% The thermal conductivity of the frozen fruit concentrate can be
estimated to

λtotal = xdm
λdm þ xH2 O ð1- αÞ
λH2 O þ xH2 O
α
λice

λtotal = 0:47
0:22 W
K - 1
m - 1 þ 0:53
ð1 - 0:8Þ
0:55 W
K - 1
m - 1

þ 0:53
0:8
2:21 W
K - 1
m - 1
λtotal = 1:1 W
K - 1
m - 1

8.6.5.3 Temperature Dependency of Thermal Conductivity

Thermal conductivity of solid food materials is only mildly dependent on tempera-
ture, and increases slightly with increasing temperature. This mild dependency is
overshadowed by the effect of gross composition, such as content of water or air.
However, at temperatures near the freezing point of water, solid foods with high
water content will experience much stronger temperature dependency. The thermal
8.6 Heat Transfer in Food 357

conductivity of water at its freezing point undergoes a dramatic change in value as

water undergoes phase transition from liquid into solid ice. Because of the orderly
crystalline structure of solid ice, its thermal conductivity is much greater than that of
liquid water. Therefore, while water, or food with high water content, is freezing, the
thermal conductivity will become a strong function of the ice fraction α present in the
water-ice mixture, as shown in Fig. 8.12.

xdm mass fraction of dry matter in kg
kg-1

xH2 O mass fraction of water in kg
kg-1
α frozen fraction in kg
kg-1
λdm thermal conductivity of dry matter in W
K-1
m-1
λH2 O thermal conductivity of water in W
K-1
m-1
λice thermal conductivity of ice in W
K-1
m-1

8.6.5.4 Liquids
In liquids, heat transfer occurs mostly by convection, the mechanism by which heat
is carried along with the flowing fluid. So, it is difficult to measure pure thermal
conductivity in liquids. Pure heat conduction can only take place in liquids if they
can be completely prevented from flowing in some way, such as when they are
formed into a gel. In the case of water, the thermal conductivity of water can be
estimated as follows [9]:

λH2 O W
K - 1
m - 1 = 0:100 þ 0:00166
T ðKÞ ð8:114Þ

λH2 O thermal conductivity of water in W
K-1
m-1

T temperature in K
ϑ temperature in °C

More accurate values are provided by the following polynomial equations [6]:

for water

λ mW
K - 1
m - 1 = 568:96 þ 188
ϑ ð ° CÞ - 8:2
10 - 3
½ϑð ° CÞ2 þ 6:02

10 - 6
½ϑð ° CÞ3

for dairy products with a fat content xf between 3 and 62%

λ mW
K - 1
m - 1 = 0:5406 - 0:0055
xf with xf in%ðm=mÞ

358 8 Thermal Properties

for dairy products with a fat content xf between 62 and 100% (m/m):

λ mW
K - 1
m - 1 = 0:2309 - 0:00051
xf with xf in%ðm=mÞ

8.6.5.5 Gases
In the case of gases, molecules are not fixed in place, but are free to have translational
motion as they become excited with increased kinetic energy from heat input.
Therefore, heat conduction in gases evidences itself by the propagation of kinetic
energy of the molecules as they “bump” into each other in response to receiving heat
energy. This transmission of kinetic energy occurs at a faster rate with increasing
velocity of the molecules. Therefore, thermal conductivity in gases increases with
increasing temperature, and with decreasing molecular weight of the gas molecule.
This is why we use low molecular weight gases like water vapor and helium for
heating and cooling when we want high thermal conductivity, and we use heavier
molecular weight gas like Xenon for thermal insulation when we want low thermal
conductivity.
The thermal conductivity of air at atmospheric pressure in the temperature range
from 0 °C to 500 °C can be estimated from the following empirical expression:

λ mW
K - 1
m - 1 = 0:701 þ 0:0629
T ðKÞ

More exact values for air at different relative humidity can be found in
reference [13].
Recall that thermal conductivity in gases will increase as molecules are
“bumping” against each other with increased kinetic energy. At the same energy
level, this “bumping” will occur more frequently when the molecules are located
more closely together. Therefore, thermal conductivity in gases will also increase as
the mean distance between molecules decreases.
At low pressures or at low concentrations of gas molecules in a given space, the
mean distance between molecules can be very large. When such a gas is confined to
limited volume of space, such as when held in a closed tank or vessel, this mean
distance between molecules may extend beyond the interior space of the enclose
tank or vessel. When this occurs, the molecules are forced to be closer together than
they would normally be, and pressure develops. As pressure increases the molecules
are forced to be closer together. Therefore, in the case of such gases, thermal
conductivity will also increase with increasing pressure.
For this reason, the thermal conductivity in a vacuum is very low and essentially
non-existent. That is why thermally insulated bottles are sometimes lined with a
glass-enclosed partial vacuum space for very effective thermal insulation. The
temperature dependency of thermal conductivity in gases is shown in Fig. 8.13.
We take advantage of this low thermal conductivity in a vacuum by carrying out
certain food processes under vacuum, such as in freeze drying. The vacuum in the
freeze dryer does not permit any heat conduction from the gaseous phase to reach the
8.6 Heat Transfer in Food 359

Fig. 8.13 Pressure

dependency of thermal
conductivity of gases,
schematic

food product. Similarly to the thermally insulated vacuum bottle, thermally insulated
windows are made of two panes of glass just 1 or 2 cm apart trapping a “dead” space
of partial vacuum between them to serve as very effective thermal insulation.

8.6.5.6 Apparent Thermal Conductivity

In systems containing both solids and liquids, heat will transfer by convection as
well as conduction (and radiation in some cases). In these types of systems, the
overall rate of heat transfer is greater than if it were to occur only by pure conduction.
If we assume this overall heat transfer is coming only from conduction, then we
observe an “apparent” thermal conductivity that is greater than the pure thermal
conductivity of any of the component substances in the system.
Let’s imagine a pile of vegetable bulk materials. It contains numerous cavities
filled with air. These cavities are connected to the surrounding atmosphere, this is
called an open-pore system. With a temperature difference between the environment
and the vegetables, heat is transported by both heat conduction and convection of the
air in the cavities. The convective part increases with increasing temperature gradient
and with increasing pore diameter. Natural convection has a predetermined direc-
tion, it is opposite to gravity. As a result, the apparent thermal conductivity is
direction-dependent. Figure 8.14 shows the different apparent thermal conductivities
of a bulk material pile in different spatial directions. The pure thermal conductivity

Fig. 8.14 Increase of

apparent thermal conductivity
in a bulk good with increasing
temperature gradient: Heat
going from bottom to top (1), O 1
from top to bottom (2). 2
Thermal conductivity of air,
alone, is independent of
3
direction (3)

T
360 8 Thermal Properties

of air, i.e., without any convective component, has only a low-temperature depen-
dence and shows no directional dependence.

8.6.6 Thermal Diffusivity

Recall at the beginning of the chapter, we mentioned the need to distinguish between
steady state and unsteady state (transient) heat transfer. In steady state heat conduc-
tion, the temperature gradient is constant with time, and we used Fourier’s first law
to describe it mathematically (Fig. 8.15). In transient (unsteady state) heat conduc-
tion, the temperature gradient is not constant and varies as a function of time
(Fig. 8.16). We can mathematically describe unsteady state heat conduction with
Fourier’s second law, Eq. (8.116). The coefficient a in Fourier’s second law is
known as the thermal diffusivity of a material, and it plays a similar role to the
thermal conductivity in his first law.
Fourier’s second law for transient heat

Fig. 8.15 Steady state heat

conduction: temperature -2
gradient is constant over time

-
-1

Fig. 8.16 Transient heat

conduction: temperature
gradient is a function of time
t

x
8.6 Heat Transfer in Food 361

one dimensional:

dT d2 T
=a
2 ð8:115Þ
dt dx

three dimensional:

dT
= a
∇2 T ð8:116Þ
dt
d2 T 2 2
whereas ∇2 T = dx2 þ ddyT2 þ ddzT2
Thermal diffusivity differs from thermal conductivity because it must also
account for the other physical and thermal properties (density and specific heat
capacity) that govern the rate of temperature response to a given rate of heat flow.
Therefore, thermal diffusivity is made up of the quotient of thermal conductivity
divided by the product of density and specific heat capacity.
λ
a= ð8:117Þ
ρ
cp

a thermal diffusivity in m2
s-1

T temperature in K
t time in s
x length in m
λ thermal conductivity in W
K-1
m-1
ρ density in kg
m-3
cp specific heat capacity in J
K-1
kg-1
a thermal diffusivity in m2
s-1

Magnitude of Thermal Diffusivity:

Most foods have thermal conductivities in the range of 0.2–0.6 W
K-1
m-1 (above
0 °C), and densities from 900 to 1500 kg
m-3 and heat capacities between 1.2 and
4.2 kJ
kg-1
K-1. So, the values of thermal diffusivity according to Eq. (8.117) can
range from 0.02
10-6 to 0.6
10-6 m2
s-1. Very often we find values in the range
0.1
10-6–0.6
10-6 m2
s-1. More exact values can be found by laboratory
experiments involving heat penetration tests.
362 8 Thermal Properties

8.6.7 Measurement of Thermal Conductivity and Thermal

Diffusivity

Thermal conductivity and thermal diffusivity are material properties that govern the
rate at which heat will conduct through the material. Both of these properties pertain
to conduction heat transfer. Thermal conductivity is the governing property in steady
state heat conduction, while thermal diffusivity is the governing property in transient
(unsteady-state) heat conduction. Recall that steady state heat conduction occurs in
response to a constant temperature gradient that does not change with time. There-
fore, measurement techniques used to measure thermal conductivity rely on a
constant temperature gradient and constant flow of heat. Techniques for measuring
thermal diffusivity are carried out under a transient temperature gradient that usually
decreases with time causing flow of heat to also decrease with time.

8.6.7.1 Steady State Techniques

The primary advantage of measurement under steady state conditions is that
experiments can be performed with relatively simple and inexpensive laboratory
equipment and straight forward data analysis following Fourier’s first law. The
disadvantage is that measurements must be taken under steady state conditions,
and it can be very time consuming to reach a controlled steady state condition in
many laboratory situations.
A basic type of experiment for measuring thermal conductivity of a material
involves arranging a sample of known geometry (area and thickness) so that it will
be subjected to heat transfer in only one direction. Once steady state is reached
(temperatures at opposite ends of the sample remain constant), it is only necessary to
record the constant temperatures and measure the flow of heat. Then, thermal
conductivity can be calculated from Fourier’s first Law.
A schematic drawing of the laboratory apparatus used for such an experiment is
shown in Fig. 8.17. A sample of material of known geometry is sandwiched on both
sides of a heat source between the heat source (hot plate) and a heat sink (cold plate),
with the edges thermally insulated to prevent any heat transfer in a transverse
direction [27].
When we use the type of apparatus shown in Fig. 8.17 it is possible to put only the
sample material of unknown thermal conductivity on one side of the hot plate, and a
reference material sample of known thermal conductivity on the other side of the hot
plate. This allows determination of the thermal conductivity in the sample material
on a relative basis to that of the reference material, and eliminates the need to

Fig. 8.17 Guarded hot plate C

method: measurement of
thermal conductivity. I P I
P sample, C cooled plate, H
H heater, I thermal insulation P
C
8.6 Heat Transfer in Food 363

Fig. 8.18 Two-plate

Q&
technique with reference
material for measurement of
thermal conductivity relative
to a reference material. dP P 'TP
P sample material, R reference
material
dR R 'TR

measure heat flow, as shown in Fig. 8.18. It is only necessary to measure the steady
state temperatures. Then we have:
λp d p ΔT R
=
ð8:118Þ
λR dR ΔT p

and

d p ΔT R
λp = λR

ð8:119Þ
dR ΔT p

For measurement of thermal conductivity in liquid or fluid samples, a flat plate

method is not appropriate, and a concentric cylinder method is used, as shown in
Fig. 8.19. In this method the heat source is a cylindrical rod as the axis inside a
hollow cylindrical tube leaving a narrow annular space in which the fluid sample is
held. In order to avoid development of any natural convection, the thickness of the
annular space must be kept very small (1–3 mm) depending on fluid viscosity, and
the temperature differences must also be small (1–5 K).

8.6.7.2 Transient State Techniques

The principle of transient measurement methods is to introduce a defined heat flow
into the sample and record the resulting temporal course of the temperature. When
using a measuring device as in Fig. 8.19, the material to be examined is filled into a
cylinder in which an axially tensioned platinum wire of length L is located. At the
beginning of the experiment (t = 0), the wire and sample are at the same temperature.
After switching on the heating current Q_ with a constant heating power, the temporal
course of the temperatures is recorded.
In the laser flash method, a test specimen is uniformly heated on the underside by
a short-term laser pulse and the resulting temperature increase on the top of the
sample is measured with the help of an infrared detector [28]. The transient hot
bridge method (THB) uses a setup like in Fig. 8.17 with a strip-shaped electrical
conductor which is used for heating and acts simultaneously as a resistance
thermometer [29].
364 8 Thermal Properties

Fig. 8.19 Concentric

cylinder method for
measurement of thermal
conductivity of fluid sample
materials, schematic. A thin
layer of sample P is in the C
annular space between heater C
P P
rod H and cooled outer
cylindrical jacket C

8.7 Caloric Value of Foods

8.7.1 Caloric (Energy) Requirement of the Human Body

Foods contain energy, which is taken up by the body when they are consumed. This
energy is used first to sustain basic metabolism of the cells in the body tissues and
organs, which is necessary even when the body is at rest (basal metabolism). Next,
additional energy is converted into work energy by the muscle tissues of the body in
contracting and expanding for the body to undergo movement and do work. Finally,
more energy is also given off as excess heat from the body after maintaining normal
body temperature. This total energy requirement can be calculated from the follow-
ing energy balance:
EA = EM þ EN ð8:120Þ

EA energy taken up in J
EM energy metabolized in J
EN energy which is not metabolized in J
EB energy excreted with body waste in J

Very little energy is normally excreted with body waste (urine and feces) from a
healthy person because these waste materials contain very little protein,
carbohydrates or fat. Also, the carbon dioxide exhaled from breathing contains no
combustion energy.
In the case of carbohydrates in our foods, only the digestible part, which can be
absorbed from the intestinal track, is taken up as energy by the body. The remainder
is called dietary fiber that is made up largely of cellulose fibers, and is excreted with
the body waste. We now know that this undigested dietary fiber still plays an
important role in sustaining a healthy balance of natural microbial populations in
8.7 Caloric Value of Foods 365

Table 8.16 Power

E_ (W) Type of activity
required by a human body
of a 70 kg person at differ- 80 Sleeping
ent activities, from [32] in 265 Walking at 5 km
h-1
[31] 400 Cycling at 15 km
h-1
700 Cycling at 21 km
h-1
100 24 h—average

the intestinal track. These cellulose fibers are further broken down by enzymes
produced by intestinal microbes to provide important nutrients for sustaining micro-
bial metabolism in the intestines. Therefore, this activity provides an important
physiological effect for sustaining good health, but it does not contribute to any
remarkable further uptake of energy by the body. So, looking at the energy balance
once again, we can say:
EA = EM þ EB ð8:121Þ

The energy turnover E_ of the body is the energy converted per unit of time.

dE
E_ = ð8:122Þ
dt
The energy turnover in J/s thus has the physical meaning of power in watts.
The power required by the human body E_ depends on many factors, such as age,
sex, work load, surrounding temperature that may fluctuate strongly during a day.
As a rule of thumb, we require a magnitude of approximately 100 W over a 24-h
period as a measure of human power intake. It is remarkable that about 70% of this
power is needed just to sustain basal metabolism. The basal metabolic rate, BMR of
a 70 kg person is ca. 80 W [30]. The metabolic basal rate is the result of basal
functions such as cell metabolism, synthesis of enzymes, hormones and other
proteids, maintenance of body temperature, uninterrupted work of muscles for
cardiological and respiratory functions, as well as brain function—in other words,
just for “being there“. If we add the energy turnover for physical activities to the
basal metabolic rate, you get the total metabolic rate, MR. The average total
metabolic rate (MR) of a 70 kg person in leisure time is 110 W, while a cyclist for
short periods can reach 1600 W [31]. Table 8.16 lists more examples of physical
activities.

Example
A rule of thumb for metabolic basal rate of our body is: 1 kcal per hour for each
kg body weight we carry. At a body weight of 65 kg we get E_ =
65 kg
1 kcal=hso

(continued)
366 8 Thermal Properties

65
103 cal 65
103
4:18 J

E_ = 65
1000 cal=h = = = 75 J
s - 1 = 75 W
3600 s 3600 s

Example
Expressing “1 kcal per hour for each kg body weight” in SI units we get a
further rule of thumb:
1 kcal 4:2 kJ 24
4:2 J
E_ = = = ≈ 100 kJ
day - 1
kg - 1
h
kg h
kg day
kg

The metabolic basal rate of for a human body is about 100 kJ per kg body
weight per day.

Example
The daily energy requirement E of a man with an a 24 h—average total
metabolic rate of 116 W is:

E = E_
t

E = 116 W
24 h = 116 W
24
3600 s = 10, 022, 400 J = 10 MJ

The approximate value of 10 MJ per person per day obtained in this example
corresponds to around 2400 kcal day-1 and is thus at the center of frequently
mentioned dietary recommendations, such as 1900–2600 kcal day-1.

→ Attention
Colloquially, calories are often spoken of, where kilo calories (kcal) are meant.
For example, the recommendation to consume 2000 calories a day through
food means 2000 kcal. Therefore sometimes food calories are expressed in the
unit Cal (C in upper case). Meaning: 1 Cal = 1000 cal = 1 kcal (with c in
lower case).
8.7 Caloric Value of Foods 367

Example
The global demand for food can be estimated by multiplying the daily energy
requirement of 10 MJ daily by a population of 9 billion:

E = 9
109
10 MJ = 9
109
10
106 J = 90
1015 J = 90 PJ

Multiplying this daily value by 365 days results to

E = 9
109
10 MJ
365 = 32
1018 J = 23 EJ

The wide range of recommendations of daily energy intake through food, such as
1900–2600 kcal day-1, is due to the fact that the energy needs of humans are
individually different, and factors such as physical activity, weight, age, gender
must be taken into account. An individualized calculation of the energy requirement
can be carried by multiplying the individual basal metabolic rate (BMR) with a
factor for the physical activity level (PAL). Let us look onto this type of calculation
more in detail.

Basal Metabolic Rate

The individual basal metabolic rate (BMR) can be measured experimentally, e.g., by
calorimetry. These are extensive long-term measurements with numerous boundary
conditions such as body position, inactive non-sleeping state, time distance to the
last meal, etc. Alternatively there are numerous suggestions for the calculation of the
BMR, e.g., based on body weight, body surface area, body mass index, BMI [33]. In
Table 8.17 some equations of body weight mK are shown for the calculation of the
basal daily energy requirement of adults. Further equations are listed [33] and are
based on empirical data of [34]:

Example
A 22-year-old woman with a body weight of 65 kg has a basal metabolic rate
of

BMR kcal
day - 1 = 14:818
mK ðkgÞ þ 486:6 = 14:818
65 þ 486:6

= 1449:8

Table 8.17 BMR calculation equations based on body weight (examples from [33])
Age: 18–30 years Age: 30–60 years
Males BMR (kcal
day-1) = 15.057
mK BMR (kcal
day-1) = 11.472
mK
(kg) + 692.2 (kg) + 873.1
Females BMR (kcal
day-1) = 14.818
mK BMR (kcal
day-1) = 8.126
mK
(kg) + 486.6 (kg) + 845.6
368 8 Thermal Properties

Table 8.18 Values of Lifestyle PAL

PAL for different lifestyles
Light activity 1.4–1.7
[33]
Moderate activity 1.7–2.0
Vigorous activity 2.0–2.4

Calculation formulas, such as those in Table 8.17 always apply to a narrowly

defined group of people such as healthy, not pregnant, not breastfeeding, etc. and
nevertheless have statistical uncertainties.

Physical Activity
The physical activity level (PAL) is a number representing the daily average of the
individual activities a person performs. It results from the weighted addition of the
factors of the individual activities (PAR) which are tabulated for a variety of
activities in different countries [35]. Resulting values of PAL are in the range
between 1.0 and 2.4 [33]. In Table 8.18 magnitudes are given for different lifestyles.

The total metabolic rate of a person can be calculated from the product of
individual basal metabolic rate and the factor for physical activity:

E_ = BMR
PAL ð8:123Þ

The energy requirement for a specified period of time Δt, e.g. a day, then results
from
E = BMR
PAL
Δt ð8:124Þ

Example
A female person with a basal metabolic rate of 1450 kcal
day-1 and a PAL of
1.5 (8 h sleep, 8 h screen work, low leisure activity). With E = BMR
PAL we
get

E_ = 1450 kcal
day - 1
1:5 = = 2175 kcal
day - 1

The daily energy requirement by food is 2175 kcal.

A male person (50 years, 65 kg) with a basal metabolic rate of 1600 kcal
day-1
and a PAL of 1.5 has E_ = 1600 kcal
day - 1
1:5 = 2400 kcal
day - 1 .
The daily energy requirement by food is 2175 kcal.
8.7 Caloric Value of Foods 369

Example
For comparison: With the rule of thumb “basal metabolic rate = 100 kJ/kg
d”
we get for a person with 65 kg body weight

kJ
E = 65 kg
100
1 day
1:5 = 65
100
1:5 kJ
kg
day
E = 9750 kJ = 2333 kcal:

It can be seen that the value is close to the previously calculated values, but
that gender-specific and activity-related influences cannot be taken into
account in this calculation.

In addition to measuring the emitted body heat (calorimetry) to determine the

metabolic rate it is possible to calculate the total metabolic rate from oxygen
consumption of a test person. A unit that is often used here is the metabolic
equivalent, MET.
Thus, the total metabolic rate for “sitting, standing” is referred to as one metabolic
equivalent. The oxygen consumption during physical activities then is given as a
multiple of this metabolic equivalent. Table 8.19 gives some examples.

8.7.2 Caloric Value of Food

Since the energy gain from food is mainly due to oxidation, the reaction energy,
i.e. the energy supply from ingested foods, can be calculated in a simple way. The
physiological caloric value of a food indicates the specific energy that humans can
obtain from these foods through physiological, chemical transformations. The phys-
iological caloric values of some groups of substances are listed in Table 8.21.
The physical measurement of the caloric value of a substance is carried out
calorimetrically by complete oxidation, i.e., combustion of this substance. The
value determined in this way is called physical caloric value. Since the human
body does not completely oxidize proteins, but excretes the contained nitrogen as
urea, the physiological caloric value of proteins is slightly lower than the physical
caloric value. This difference does not exist in carbohydrates, fats and alcohol, as
these substances are also completely oxidized in the human metabolism, i.e., the

Table 8.19 Metabolic Physical activity MET

equivalents (MET) of
Sitting, standing 1
physical activities,
examples from [31] Driving car 1.5
Walking, 4 km h-1 3.5
Running, 8 km h-1 7.5
Swimming, crawl 45 m min-1 15
370 8 Thermal Properties

Table 8.20 Comparison of physiological and physical caloric values, from [36]
Physiological combustion value Physical combustion value
Group e (kJ
g-1) e (kcal
g-1) Δhcomb (kJ
g-1) Δhcomb (kcal
g-1)
Proteins 17.2 4.1 22.6 5.4
Carbohydrates 17.2 4.1 17.2 4.1
Fats 38.9 9.3 38.9 9.3
Ethanol 29.7 7.1 29.7 7.1

Table 8.21 Physiological Group ei (kJ
g-1) e (kcal
g-1)

caloric values of the main
Proteins 17 4
energy sources of food [4]
Carbohydrates 17 4
Fats 37 9
Ethanol 29 7
Dietary fiber 8 2
Organic acids 13 3

carbon contained leaves the body as CO2. The differences between physiological
caloric value and physical caloric value can be seen from Table 8.20.
According to the official European Food Information Regulation [4] the physio-
logical caloric values given in Table 8.21 are used in declaration of food
composition.
According to this regulation, the physiological energy value of a food must be
indicated on the packaging [4]. The values to be reported are not obtained by
calorimetric measurements of production samples but mathematically using the
composition and data as shown in Table 8.21. The sum of the products of relative
mass fraction xi and associated specific physiological caloric value ei gives the
specific physiological caloric value of the food (in J/g). By multiplying by 100 we
obtain the indication for 100 g of food.

etotal = xi
ei ð8:125Þ
i

with

mi
xi = ð8:126Þ
mtotal

E energy contained in food in J (physiological caloric value of the food)

ei specific physiological caloric value of the component i in J∙g-1
(Atwater coefficient)
etotal specific physiological caloric value of the food in J∙g-1
(continued)
8.7 Caloric Value of Foods 371

mi mass of component i in g
mtotal mass of food in g
xi mass fraction of component i
Δhcomb specific enthalpy of combustion i in J∙g-1 (physical caloric value)

Remark
This calculation method is also called the Atwater system, after Wilbur
Atwater (1844–1907), who early worked on the combustion calorimetry of
food. The specific physiological caloric values are also referred to as Atwater
factors or Atwater coefficients.

Example
Physiological caloric value of pea, Pisum sativum L. (hull and seed):
Composition of 100 g is [37]

mi (g)
75.2 Water
6.6 Protein
0.5 Fat
12.3 Carbohydrate
4.3 Dietary fiber
0.3 Organic acids

1
According to etotal = xi
ei with that values we get e = ð6:6 g

i 100 g
17:2 kJ
g - 1 þ 0:5 g
38:9 kJ
g - 1 þ 12:37 g
17:2 kJ
g - 1 þ 4:3 g
8 kJ

1
g - 1 þ 0:3 g
13 kJ
g - 1 Þshorter :e = ð6:6
17:2 kJ þ 0:5
38:9 kJ þ
100 g
1
12:3
17:2 kJ þ 4:3
8 kJ þ 0:3
13 kJÞe =
382:8 kJ = 382:8 kJ=100
100 g
ge = 91:6 kcal=100 g
with the official values of EU, see Table 8.21, we get a slightly different
result

(continued)
372 8 Thermal Properties

1
e= ð6:6
17 kJ þ 0:5
37 kJ þ 12:3
17 kJ þ 4:3
8 kJ þ 0:3
13 kJÞ
100 g
e = 378:1 kJ=100 g = 90:2 kcal=100 g

For foods with a negligible content of fiber or organic acids, the caloric value can
be roughly calculated by focussing on proteins, carbohydrates and fats. Knowing the
mass fractions xi these food components, respectively, their mass in 100 g of the food
under consideration is first multiplied their related Atwater coefficients and then,
added together to obtain the caloric value of 100 g of that food.

Example
Physiological caloric value of peeled potato. Composition of 100 g edible
substance:

mi (g)
79.8 Water
2.1 Protein
0.1 Fat
17.7 Carbohydrates

According to etotal = xi
e i
i

2:1 g 0:1 g 17:7 g

e=
17:2 kJ
g - 1 þ
38:9 kJ
g - 1 þ
17:2 kJ
g - 1
100 g 100 g 100 g
1
e= 2:1 g
17:2 kJ
g - 1 þ 0:1 g
38:9 kJ
g - 1 þ 17:7 g
17:2 kJ
g - 1
100 g
written shorter :
1
e= ð2:1
17:2 kJ þ 0:1
38:9 kJ þ 17:7
17:2 kJÞ
100 g
1
e= ð36:1 þ 3:9 þ 304:4Þ kJ
100 g
1
e=
344:4 kJ = 344:4 kJ=100 g
100 g
e = 82:4 kcal=100 g

If we need the caloric value of a meal or of a certain portion that is not exactly
100 g, it is possible to directly add the products of mass mi and associated specific
physiological caloric value of component i. This is how you get the caloric value of
the given portion in J.
8.7 Caloric Value of Foods 373

Etotal = mi
ei ð8:127Þ
i

Example
One serving of nougat cream contains 4.8 g of fat, 8.5 g of carbohydrates and
0.1 g of protein. With the Atwater coefficients from Table 8.21 we get the
physiological caloric value of the serving to

E = 4:8 g
37 kJ
g - 1 þ 8:5 g
17 kJ
g - 1 þ 0:1 g
17 kJ
g - 1 = 324 kJ

It has already been pointed out that the caloric values given on the food packaging
are not laboratory-measured values of the respective production batch, but compu-
tational values calculated with the Atwater system. For this purpose, the average
composition of the product is used. This average composition may be subject to crop
or seasonal variations. In addition, the Atwater coefficients are themselves average
values for the specific physiological caloric value of main energy sources in food.
Strictly speaking, not all monosaccharides, starches, amino acids, animal and vege-
table proteins have the same value of 17.2 kJ
g-1 but differ slightly. In addition, the
digestive system of humans works differently, i.e. not every person can get the
energy from a food component that the Atwater coefficient indicates. There is
evidence that the food matrix itself also influences the energetic usability of
ingredients [38], which is not taken into account in the Atwater system. These
limitations should be taken into account when interpreting and applying the calcu-
lated caloric values. Not only the energy demand of humans but also the energy
utilization of food is subject to individual influences.

8.7.3 Measurement of Caloric (Combustion) Values

The caloric value (energy content) of a food sample can be measured experimentally
in a laboratory with the use of a combustion calorimeter. The sample is placed in the
combustion calorimeter and completely oxidized. The heat evolved from this oxida-
tion reaction causes a rise in temperature that is measured and recorded, from which
the caloric value can be calculated. Standard methods for measuring caloric food
values with combustion calorimeters can be found in [39, 40]. In a combustion
calorimeter all the protein, carbohydrate and fat are completely oxidized into NO2,
CO2, and H2O. However, the human body has a different need for the nitrogen in the
protein, and does not oxidize the nitrogen into NO2 as does the calorimeter. Much of
the nitrogen coming from proteins in the human body is metabolized in the
following way: 80–90% to urea (H2N–CO–NH2), 3–5% to ureic acid and <1% to
creatinine. Therefore, the physiological caloric value of proteins is below the
374 8 Thermal Properties

physical caloric value. Therefore, the energy value obtained from the calorimeter is
not the same as the energy value taken up by the body, and some adjustment is
needed. For this conversion, knowledge of the protein content of the food is
necessary.

8.8 Thermal Analysis

The International Confederation for Thermal Analysis and Calorimetry (ICTAC) has
defined Thermal Analysis (TA) as a group of techniques that allow us to study the
relationship between material properties of a sample and its temperature Examples of
such techniques include thermal gravimetric analysis, thermal mechanical analysis,
dielectric thermal analysis, thermometry, etc. [39]. Table 8.22 gives a listing of such
examples.
Performance of these analytical methods for thermal analysis [41, 42] can often
be enhanced by coupling different methods of thermal analysis, such as DSC-TG,
DSC-TOA, and TG-EGA. In the following section, we will focus on TG and DSC.

8.8.1 Thermogravimetry (TG)

A thermogravimetric measuring unit consists mainly of a sensitive weighing balance

that is placed within an oven (Fig. 8.20). During a temperature-time program in the
oven, the weight of the sample is measured and recorded over time.
The temperature-time program can be a simple linear ramp, but can also be a step-
wise increasing temperature function with each step increase consisting of a short
isothermal time interval. The program can also be a single-step increase to a
specified oven temperature held constant over a specified time. This is called an
isothermal analysis.
The oven atmosphere can be dry air or just nitrogen. At high temperatures,
volatile substances are driven off, and a decrease in sample mass can be observed
and measured, as shown in Fig. 8.22.
The various reactions that can be studied in this way are listed in Table 8.23.
Those most used in food technology are involved in food drying (desorption and
evaporation of water). Figure 8.21 shows some typical TG signals for these types of
transitions.
In food process engineering, drying and oven roasting processes can be studied
by thermogravimetric analyses. By observing the moisture loss as a function of
temperature over time, the water binding characteristics of the food sample can be
studied, as well as the enthalpy of desorption or adsorption, see Chap. 1.
We can also use TG to validate the performance of techniques used to measure
the moisture content in foods, such as oven drying or infra-red balance methods. The
use of TG analyses in these applications also allows us to determine at what
temperatures and after what times drying is completed, and at what point thermal
degradation of the food material will occur.
8.8 Thermal Analysis 375

Table 8.22 Examples of techniques used in thermal analysis

Property
measured Name of technique Abbreviation Note
Mass Thermogravimetry or tg mass of sample is recorded during a
Thermogravimetric tga given temperature-time program
Analysis
Dimensional Dynamic Mechanical DMA Deformation or dimensions are
and Analysis TMA measured to get mechanical moduli
mechanical Thermomechanical TD (storage/loss)
Analysis
Thermodilatometry
Electrical Dielectric Thermal DEA Permittivity/dielectric loss measured
Analysis
Electrical Thermally Stimulated TSC
Current
Magnetic Thermomagnetometry Often combined with TGA
Gas flow Evolved Gas Analysis EGA Nature and/or amount of gas/vapor is
determined
Gas flow Emanation Thermal ETA Trapped radioactive gas within the
Analysis sample is released and measured
Pressure Thermomanometry Evolution of gas is detected by
pressure change
Pressure Thermobarometry Pressure exerted by a dense sample
on the walls of a constant volume cell
is studied
Optical Thermoptometry TO Family of techniques in which an
optical characteristic or property of a
sample is studied
Optical Thermoluminescense TL Emitted light measured
Acoustic Thermosonimetry or Techniques where the sound emitted
Thermoacoustimetry (sonimetry) or absorbed
(acoustimetry) by the sample is
studied

The output signal from a TG analysis is a smooth curve on a graph of mass versus
temperature, called a TG-signal, as shown in the upper graph of Fig. 8.22. The first
derivative of this TG-signal curve is the mass flow. When it is plotted versus
temperature this is called a DTG plot as shown in the lower graph of Fig. 8.22.
When we examine the DTG-Signal curve, we can see that a negative peak will
represent a decrease in mass, while a positive peak would represent an increase in
mass. The integration of the DTG peak (area within the peak profile) will result in the
total mass that was lost or gained during the analysis.
Making a DTG-Signal plot instead of a TG-signal plot allows us to apply standard
mathematical techniques to analyze the important features. For example, we can
determine the temperatures at the beginning and end of a given peak, as well as the
maximum temperature of each peak, as shown in the lower curve in Fig. 8.22. In
Fig. 8.23 we can see an example of both TG and DTG signal plots from the thermal
376 8 Thermal Properties

Fig. 8.20 Thermogravimetric measuring system, schematic: the sample P is in an oven O. The
sample carrier is coupled to a balance W which is outside of the oven

Table 8.23 Reactions which can be investigated by TG

Sample behavior TG signal on heating
Desorption, evaporation (e.g., drying) Mass decrease
Sublimation Mass decrease
Dissociation into gaseous substances Mass decrease
Reaction with gases (e.g., oxidation) Mass increase
Curie transition of metals Apparent mass change

Fig. 8.21 TG-signals,

schematic: (a) evaporation, a
drying, sublimation;
(b) boiling in pan with small b
hole; (c) sample oxidation;
(d) Curie transition
m

c
d

degradation (roasting) of calcium oxalate dehydrate, showing three steps of mass

drop due to dehydration and dissociation:
Another example is shown in Fig. 8.24 where we have both TG and DTG signal
plots from the thermal degradation of aspartame (artificial sweetener used in many
food formulations). In this case, we see a distinct peak occurring in the temperature
8.8 Thermal Analysis 377

Fig. 8.22 TG and

DTG-signal curves.
Beginning Te and end Tf of the
transition can be read out as
extrapolated temperatures. Tp
is the peak temperature, it is m
the temperature where the
mass flux has its maximum

Te Tf Tc
Ti T

Te' TP Tf'

dm
dt

range between 180 and 200 °C. This is the temperature range normally used in most
food baking and roasting processes, and poses a problem with the stability of
aspartame during such processes.
In many food ingredients the mass loss during oven roasting processes is not only
due to loss of moisture, but many gaseous components, such as CO2-emissions and
volatile flavor and aroma compounds. In order to study such compound losses by
thermal analysis, it is necessary to use TG analysis coupled with other appropriate
methods of thermal analysis, such as EGA. The roasting of coffee falls into this
category, and was examined further by Cammenga et al. [43, 44].
A thermal gravimetric method must be calibrated first for the weighing system,
using methods discussed earlier in Chap. 2. Then, it must be calibrated for the
temperature measuring system by using standard materials with defined transition
points at known temperatures at well. Standard materials that can be used for this
purpose are various hydrates, which give up their water of hydration at exactly
known temperatures (Fig. 8.23). Alternatively, materials which exhibit a well-
defined Curie transition can be used as standards. A Curie transition is when
ferromagnetic materials lose their magnetization and therefore lose their magnetic
attraction to a magnet. Therefore, if a permanent magnet is placed near to the sample
in a thermogravimetric system, the balance will show a signal. When a permanent
magnet is mounted in the region just below the sample on the weighing balance, the
sample will be attracted to the magnet, and show more than its true weight while
378 8 Thermal Properties

Fig. 8.23 TG- and DTG plot:

thermal degradation of TG
Calcium-Oxalate-Dihydrate:
three steps of mass drop due to
dehydration and dissociation m dm
[43]
DTG dt

200 400 600 800 1000

- / qC

Fig. 8.24 TG of Aspartame.

Note the mass drop on heating
TG
above baking temperatures of
180 °C
m, m
DTG

60 100 140 180 220

- / qC

initially resting on the balance. Then upon heating, the sample will show a Curie
transition (loss of its magnetic attraction to the magnet). This will be observed as an
apparent sudden mass loss by the sample resting on the balance because of the loss of
magnetic attraction This will happen at a specific defined temperature that is
precisely known for the standard material, and is called the Curie temperature
(Fig. 8.25).
Consideration must also be given to the precision with which temperature is
measured in thermogravimetric analysis. For example, the temperature measuring
sensor or probe used in TG analysis is not located within the sample material itself,
but in close proximity to the sample. Therefore, the temperatures measured and
recorded during a test are not precisely those experienced by the sample material.
The difference between displayed and sample temperature will depend on the
thermal conductivity and size of the material sample. This difference will increase
as thermal conductivity decreases and sample size increases. This will cause large
differences that may be problematic with large sample size and low thermal conduc-
tivity. Therefore, it is recommended that sample sizes (as well as rate of heating) be
as small (or low) as practically possible to minimize this lack of precision in
temperature measurement with TG analysis.
Additional information can be obtained from TG analysis by modification of the
gaseous atmosphere surrounding the sample in the heating oven chamber. For
example, by changing the atmosphere from low to high relative humidity in the
chamber, we can study first desorption, then adsorption. Or, by changing from
aerobic to anaerobic atmosphere, we can study what part of the reaction depends
8.8 Thermal Analysis 379

Fig. 8.25 TG calibration

with a Nickel standard. On
reaching the Curie
temperature (for Nickel
351.4 °C ± 4.8 °C) the
magnetic attraction disappears m
and the mass signal shows a
sudden decrease

on oxygen. We can also study applications under near vacuum [45]. When extreme
levels of accuracy and precision are important, then we must also consider buoyancy
effects of the atmospheric gases or fluids surrounding the sample in the heating oven
chamber, and the fact that different gases will have different buoyancy effects. In
those types of situations, correction factors for these buoyancy effects must be
applied (see Sect. 2.1., weighing and buoyancy effects).

8.8.2 Heat Flow Calorimetry

Heat flow calorimetry is useful for relating temperature and heat flow measurements
with heat capacity of a material. Heat flow calorimeters display a heat flow signal by
sensing a temperature difference between the sample and a reference material during
application of a temperature-time program. The use of a reference material is key to
the relative simplicity of this method. Only the signal representing the difference
between sample and reference temperature is needed. Measurement of absolute
temperature is not required anywhere. However, initial and boundary test conditions
surrounding both sample and reference materials must be identically the same, and
handled like twin sister samples during the complete procedure. For example, the
pans used for sample and reference must have exactly the same heat capacity.
Most heat flow calorimeters in modern use today are known as differential
scanning calorimeters (DSC). There are two general principles of DSC hardware.
First is the temperature difference ΔT measuring system, and second is the power
compensation system.

8.8.2.1 ΔT-Measuring System

Figure 8.26 shows a schematic diagram of a DSC oven chamber containing both
sample and reference materials. On the left side is the sample in a pan and on the
right side is the reference material in an identical pan. When the oven is turned on,
heat will flow into both pans at a given controlled rate. This will cause the tempera-
ture to rise in both sample and reference material in accordance with the thermal
properties of each material.
380 8 Thermal Properties

TP TR

Q&OP TO Q&OR

Fig. 8.26 DSC oven schematic. TP sample temperature, TR reference temperature, TO oven
temperature, Q_ heat flow

The DSC oven chamber is designed in such a way, that heat transfer by radiation
and convection is negligible. Therefore, heat transfers from the oven into both
sample and reference material by pure conduction, and we can apply Fourier’s fist
law to calculate the heat flow into the sample material using the temperature
difference (TO - TP) between oven and sample. In the same way, we can calculate
the heat flow from the oven to the reference material. Then, we can determine the
difference in heat flow between the sample and the reference material by subtracting
the heat flow into the reference from the heat flow into the sample:
heat flow into the sample:

Q_ OP = k
A
ðT O- T P Þ ð8:128Þ

heat flow into reference material:

Q_ OR = k
A
ðT O- T R Þ ð8:129Þ

difference:

Q_ = Q_ OP - Q_ OR ð8:130Þ

or

Q_ = k
A
ðT O- T P - ðT O- T R ÞÞ ð8:131Þ

i.e.

Q_ = k
A
ðT R- T P Þ ð8:132Þ

as a short cut, we can say:

Q_ = k
A
ΔT ð8:133Þ

using the hardware constant

K =k
A ð8:134Þ
8.8 Thermal Analysis 381

we get for the heat flow

Q_ = K
ΔT ð8:135Þ

Q_ heat flow in W
T temperature in K
ΔT temperature difference in K
A area in m2
k overall heat transfer coefficient in W∙m-2∙K-1
K hardware constant in W∙K-1
Indices:
O oven
P sample
R reference

The temperature-time plots for the oven, sample and reference material are shown
in the upper graph of Fig. 8.27, while the temperature difference between sample and
reference over time (ΔT-Signal) can be seen on the lower graph in Fig. 8.27. As can
be seen from Eq. (8.135), this ΔT-Signal is proportional to the heat flow into the
sample.

Fig. 8.27 ΔT= TR - TP

TR
versus time during a DSC run T0

TP

TR – TP
0

t
382 8 Thermal Properties

Fig. 8.28 DSC disc system

[43]
P R
˜ ˜
QOP QOR

'T

We can distinguish two types of hardware systems used in DSC heat flow
calorimetry. One system is known as the DSC disc system in which the sample
and reference pans are placed upon a disc. The thermocouples used for measuring
ΔT are incorporated into the disc, as shown in Fig. 8.28. The other type is a
cylindrical system, in which the temperature probes are in the form of cylindrical
cups, as shown in Fig. 8.29. This type is appropriate for larger volume sample sizes.

8.8.2.2 Power Compensation System

Besides the ΔT-measuring system, there are DSC instruments that are based upon
the principle of power compensation. The general construction of such a system is
illustrated in Fig. 8.30. The main characteristic of these systems is the use of small
individual heaters beneath each separate pan. These small heaters allow independent
control of heat flow so that the temperature in both sample pan and reference pan is
kept the same, and the difference between them kept at zero all the time. In this case,
the power needed to keep this difference at zero is what is measured and recorded
during the test.

8.8.2.3 Evaluation of a DSC Plot

A typical DSC plot is shown in Fig. 8.31, in which we see a graph of heat flow
plotted against temperature. Such a DSC plot is sometimes called a thermogram. In
thermodynamic systems analysis, heat taken up by a system (gain) is considered
positive, while heat given up (loss) is considered negative. Likewise with heat flow,
in which heat flowing into a system is positive (endothermic), and heat flowing out is
negative (exothermic). This same convention is used in thermal analysis and DSC
peaks. In cases where different rules of convention are used, the terms “endother-
mic” or “exothermic” heat flow in the DSC plot must be specified.
The DSC plot shown in Fig. 8.31 has a positive peak, meaning we are looking at
an endothermic reaction. The figure shows how the quantity of heat delivered can be
calculated by integration of heat flow over temperature within the peak (area under
the peak), and how the onset temperature is determined by reading the temperature at
which the peak begins to develop. To calculate the quantity of heat delivered, the
incremental heat flow is expressed as:
Δq = cp
ΔT ð8:136Þ
8.8 Thermal Analysis 383

Fig. 8.29 DSC cylindrical

system [43]

˜ P ˜
QOP R QOR

'T

Fig. 8.30 Power

compensation calorimeter,
schematic. TP sample
temperature, PP heating power P R
for the sample, TR temperature
of reference, PR heating
power for the reference [43]
PP TP TR PR

where

Q
q= ð8:137Þ
m
so

dq = cp
dT ð8:138Þ

and

Δq = cp
dT ð8:139Þ
T

with the heating rate β

dT
β= ð8:140Þ
dt
we get

Δq = cp
β
dt ð8:141Þ
t
384 8 Thermal Properties

Fig. 8.31 Simple

endothermic DSC plot.
Illustration of onset
temperature and area of a
˜
DSC peak Q

Tonset

When we use Eq. (8.11) then it is dh = dq and the integral provides

Δh = cp
β
dt ð8:142Þ
t

During a DSC experiment, the temperature of the sample changes over time.
From this profile, we can integrate the heat flow over time to obtain the enthalpy that
is taken up or given up by the sample.

cp specific heat capacity in J
kg-1
K-1

q specific heat in J
kg-1
T temperature in K
t time in s
β heating rate in K-1∙min-1
h specific enthalpy in J
kg-1

Process engineers often need to work with thermodynamic properties like heat
and enthalpy on the basis of a unit mass of material (mass-specific quantities). These
mass-specific quantities are designated by using lower case letters, such as q, cp, h,
instead of upper case letters Q, Cp, H normally used for the global (total) values.
Table 8.24 presents a comparison of the total and mass-specific quantities for
specific heat, heat, and enthalpy.
Figure 8.32 shows an enthalpy-temperature diagram for ice cream mix, on which
the enthalpy is plotted as a function of temperature. Enthalpy is decreasing with
decreasing temperature, and we should want to know at what temperature it will
reach zero. We know it is not possible to measure enthalpy at zero absolute
temperature. Therefore, it will be useful to designate a reference point where the
enthalpy is assumed to be zero for subsequent calculations. For practical reasons,
this reference point has been chosen to be at zero degrees on the centigrade
temperature scale, h(0 ° C) = 0. The following example shows that the choice of
datum point when calculating Δh is irrelevant to the result.
8.8 Thermal Analysis 385

Table 8.24 Calculation of enthalpy from a DSC peak. Comparison of total and mass-specific
nomenclature
Total Mass-specific
dQ dq
Cp = cp =
dT dT
dQ = C p
dT dq = cp
dT
with β = dT
dt
dQ = C p
β
dt dq = cp
β
dt

Q= C p
β
dt q= cp
β
dt

When V
dp = 0 the heat measured is identical to the change of enthalpy of the system
dH = dQ dh = dq
ΔH = dQ Δh = dq

ΔH = C p
β
dt Δh = cp
β
dt

Example
How much heat must be drawn out of 500 kg ice mix to chill it from -5 °C to
-20 °C

Δh = h2 - h1 = hð- 20 ° CÞ - hð- 5 ° CÞ

Solution:
From Fig. 8.32 we read:

Δh = - 200 kJ
kg‐1 - - 100 kJ
kg‐1

Δh = - 100 kJ
kg‐1

ΔH = m
Δh
ΔH = 500 kg
- 100 kJ
kg‐1 = - 5000 kJ = - 5 MJ

So a heat of 50 MJ has to be drawn out of the material sample. The negative

sign in the result indicates that this heat leaves the system, it is an exothermic
process. 5 MJ of heat leave the ice mix system.

When we integrate completely under the peak in a thermogram like that in

Fig. 8.32 we obtain the heat or enthalpy of the transition that occurred during the
thermal analysis, such as the heat of vaporization or the heat of fusion. It is also
possible to carry out the integration only part way through the peak, resulting in a
partial integration. This partial integration can be quantified as a ratio of the partial
integration over the whole integration, as illustrated in Fig. 8.33. This ratio will have
a value between zero and one. This quantity can be called the turnover of the peak α,
386 8 Thermal Properties

Fig. 8.32 Enthalpy of ice 50

cream mix [16] 0
-20
-50
h
-100
kJ ˜kg 1
-150
-200
-250
-40 -30 -20 -10 0 10 20
- / qC

Fig. 8.33 Partial integration

of a DSC peak
A
D
AB

&
Q

A B

-20 -15 -10 -5 0 5

- / qC

Fig. 8.34 DSC turnover

plots of the melting transition 1.0
for different ice cream mix 0.8
recipes D 0.6 1 2 3

0.4

0.2

-20 -15 -10 -5 0 5

- / qC

and can also be plotted against temperature. This is known as a turnover plot, and is
shown in Fig. 8.34 for the case of ice cream.
Figure 8.35 shows some typical signals which may occur during a DSC run.
Looking at the graph from left to right we see a step in the baseline which can
indicate a glass transition followed by an exotherm peak which can be the signal of a
recrystallization. The following endotherm peak might be due to a melting transition,
vaporization, sublimation, or another type of transition. At high temperatures, many
materials will undergo some type of thermal degradation in the form of decomposi-
tion or oxidation. This will appear as a strong exothermal signal near the end of the
DSC plot with very poor reproducibility.
Figure 8.36 is a DSC plot (thermogram) of a plastic film consisting of poly-
ethylene-terephthalate (PET). It is an example of an amorphous plastic material out
of which most soft drink bottles are made. In the DSC plot, we see a glass transition
with a step in the base line at about 60 °C. Upon further heating, we see an
8.8 Thermal Analysis 387

Fig. 8.35 Schematic of

hypothetical DSC signal
options

&
Q

Fig. 8.36 DSC plot

(thermogram) of PET

q˜
W˜g-1

-50 0 50 100 150 200

- / °C

exothermic peak just after 100 °C due to recrystallization of the polymer. Finally, we
see an endothermic peak due to the melting of the crystalline structure at about 180 °
C. Glass transition is a second-order transition (refer Sect. 8.3.1) and can be detected
by shifting of the baseline in a thermogram. This is because, at the glass transition
temperature, the heat capacity of the material changes dramatically. Glass transitions
can occur only with solids that are not crystalline. There are some applications in
which food products are in the form of a solid glass state material. In these types of
products, the moisture content has strong influence on the glass transition
temperature.
In food, water content often has significant influence on the glass transition [7, 8,
46]. Table 8.25 lists phenomena from food science, which are examined by thermal
analysis [47, 48]. A good introduction to thermal analysis of foods can be found in
reference [49].

8.8.2.4 Temperature Modulated DSC

Temperature modulated DSC (MDSC) is based upon the superposition of a periodic
cycling temperature program onto a linear temperature-time program. Sometimes
this technique is called alternating DSC (ADSC). In Fig. 8.37 we see an example of a
periodic modulated temperature-time profile superimposed upon a linear ramp
temperature-time profile. If the periodic profile is a sinusoidal function, then we
can express the temperature as a function of time as follows:
388 8 Thermal Properties

Table 8.25 Transformations and quantities that are examined in the food sector by means of
thermal analysis [47]
Material Transitions and properties
Food, general Heat capacity, supercooling, evaporation, reactions, roasting
Carbohydrates Crystallization, melting, decomposition, gelation, retrogradation, glass
transition, recrystallization of amorphous substances
Fats Crystallization, melting, polymorphism, oxidation
Proteins Denaturation, oxidation, glass transition in powders
Microorganisms Growth under aerobic or anerobic conditions

Fig. 8.37 Temperature

A-
modulated heating rate

-
T

ϑ = ϑ0 þ β0
t þ Aϑ
sinðωt Þ ð8:143Þ

The period T of the modulation can be obtained from

2π
ω = 2π
f = ð8:144Þ
T
to

2π
T= ð8:145Þ
ω
The heating rate can be obtained as the derivative of the temperature
dϑ
β= ð8:146Þ
dt
Then, we get:

ϑ = ϑ0 þ β0
t þ Aϑ
sinðωt Þ ð8:147Þ
8.8 Thermal Analysis 389

so

β = β0 þ Aϑ
ω
cosðωt Þ ð8:148Þ

The product Aϑ
ω we call amplitude Aβ of the heating rate. So, it is:

β = β0 þ Aβ
cosðωt Þ ð8:149Þ

β0 is the underlying linear heating rate, whereas β is the actual (momentary)

heating rate.

ϑ temperature in °C
ϑ0 initial temperature in °C
t time in s
β0 underlying heating rate in K
s-1
β actual heating rate in K
s-1
Aϑ amplitude of temperature in K
Aβ amplitude of heating rate in K
ωf angular frequency in s-1
f frequency in s-1
T period in s

When the heating rate is modulated, the resulting heat flow is modulated also.

dQ
Cp = ð8:150Þ
dϑ
the modulated heat flow is

dϑ
Q_ = C p
ð8:151Þ
dt
i.e.

Q_ = C p
β ð8:152Þ

When we place a sample into a MDSC instrument and apply a sinusoidal

temperature modulation, the system will respond with a sinusoidal modulated heat
flow Q_ . This heat flow is not proportional to the temperature, but to the first
derivative of temperature with respect to time, which is the heating ratedϑ dt = β .
Therefore, heat flow Q_ and heating rate β are said to be “in phase” with each
other, and there is zero phase shift between them, δ = 0°.
Zero phase shift represents a behavior in which a sample responds only based on
its heat capacity. When there are thermal transitions in the sample, this is different
from the behavior of a simple heat capacity because then heat can be “consumed” by
the sample or dissipated. In such cases the response to temperature-modulated
analysis in an MDSC will show a phase shift greater than zero. So the value of the
390 8 Thermal Properties

Table 8.26 MDSC: Cases of phase shifts between heating rate and heat flow and reasons for them
Phase shift angle δ Heating rate and heat flow are. . . The heat flow results from . . . .
δ = 0° in phase a heat capacity only
0 ° < δ < 90° shifted by δ both parts
δ = 90° shifted by 90° a kinetic part only

phase shift δ provides information about the extent to which a sample takes up heat
in a reversing or a non-reversing mode. Table 8.26 gives an overview of typical
phase shifts.
When the response of a sample to MDSC differs from ideal heat capacity, it
means some type of transition is occurring that requires heat in addition to the heat
capacity part. We call this added heat flow for transitions the kinetic part. In general,
we can say the modulated heat flow to a sample from MDSC is made up of a part due
to the heat capacity of the sample as well as a kinetic part due to transitions.
Therefore, the response signal from an MDSC thermogram can be separated by
Fourier analysis into two component heat flow signals. Conversely, when we add
these two components with the phase shift δ, we can obtain the total measured signal.
In mathematics, we make use of complex numbers to represent these types of
quantities. A complex number contains an imaginary part and a real part (refer
appendix). So, using complex numbers, we can express the heat flow mathematically
as follows:
 0 00
Q_ = Q_ þ i
Q_ ð8:153Þ


Q_ complex heat flow in W
0
Q_ real part of complex heat flow in W
00
Q_ imaginary part of complex heat flow in W
C p complex heat capacity in J
K-1
C 0p real part of heat capacity in J
K-1
C 00p imaginary part of heat capacity in J
K-1

Recall that we first worked with complex numbers in Chap. 4 on rheology. The
real part of the number is the storage part, and the imaginary part is called the loss
0
quantity. In the case of heat flow, the real part Q_ is that part which can be taken up
by the sample with its heat capacity, so the heat can be stored and subsequently
released. We call this real part reversible heat flow. The imaginary part, on the other
hand, is the heat flow for the kinetic effects that are consumed by the sample. This
part is non-reversing, and is lost from the sample. Table 8.27 summarizes these
definitions.
In the same way, we can think of the heat capacity of a sample as being a complex
quantity made up of a real part and an imaginary part. The terms used for this
purpose are listed in Table 8.28.
8.8 Thermal Analysis 391

Table 8.27 Complex heat flow. Synonymous terms

 0 00
Q_ Q_ Q_
Complex Real part Imaginary part
quantity
Total heat flow Reversing heat flow Non-reversing heat flow
Part from heat capacity of the Part from kinetic effects in the
sample sample
In-phase part (δ = 0°) Out-of-phase part (δ = 90°)

Table 8.28 Complex heat capacity terms

C p C 0p C00p
Complex quantity Real part Imaginary part
In-phase part (δ = 0°) Out-of-phase part (δ = 90°)

With

Q_
C p = ð8:154Þ
β
it is

C p = C 0p þ i
C00p ð8:155Þ

With phase shift δ known it is possible to calculate the parts of that complex heat
capacity:
real part:

C 0p = C p
cos δ ð8:156Þ

imaginary part:

C 00p = C p
sin δ ð8:157Þ

We can now appreciate how MDSC experiments can allow us to distinguish

between reversing and non-reversing heat flow. Figure 8.38 shows a MDSC ther-
mogram from a partly crystalline carbohydrate sample. We can see that the glass
transition can be detected in the reversing signal. Whereas the recrystallization
reaction (exothermic peak) is detectable in the non-reversing signal. The melting
signal seems to have both components (reversing and non-reversing). Note that this
type of information cannot be derived from conventional DSC analysis.

8.8.2.5 Experimental Conditions for MDSC Experiments

Recall that in carrying out a DSC experiment, only certain experimental conditions
and parameters are subject to being controlled by the operator of the instrument.
392 8 Thermal Properties

Fig. 8.38 MDSC plot of a

semi crystalline carbohydrate.
The glass transition can be
seen in the reversing signal rev
(rev). The classic DSC signal
(total) does not show the glass non-rev
transition because it is hidden Q
by another thermal event [50]
total

50 100 150 200 250

- / qC

These include heating rates, oven atmospheres, types of pans used, sample size, etc.
We also learned that very small sample sizes are always recommended for best
results. However, the size of the sample must be sufficiently large so that the limited
sensitivity of the instrument and of the sample weighing balance will not introduce
any significant error.
When we advance to modulated DSC (MDSC), we also introduce frequency and
amplitude of the oscillations as additional control variables. If the frequency is set at
zero, then we have no modulations at all, and the experiment reverts to a classic DSC
analysis. If the frequency is set too high, the hardware system may be incapable of
reversing the direction of heat flow at such a high speed. Therefore, it is always
necessary to perform a series of preliminary experiments for the purpose of deter-
mining the most appropriate frequencies and amplitudes to be used in order to meet
the intended objectives of the MDSC analysis. For details refer to reference [51, 52].

8.8.3 Combustion Calorimetry

In a combustion calorimeter the heat released during complete combustion, i.e., the
complete oxidation of a material can be measured. In Fig. 8.39 the setup of such
calorimeter is shown as a schematic. The sample is located in a tightly sealed
stainless steel pressure vessel. This pressure vessel is surrounded by a water jacket,
the so-called inner jacket. The pressure vessel and inner mantle form the measuring
system and are thermally separated from the outer jacket of the calorimeter. The
outer jacket is also a temperature-controlled water jacket. The heat capacity of the
measuring system is known or is to be determined by calibration. Before starting the
measurement, the pressure vessel is filled with oxygen by some MPa. This excess
oxygen causes the combustion of the sample to run off completely, i.e., no
non-oxidized residues will remain. The combustion of the sample is initiated by an
electric ignition spark. The heat generated by the combustion increases the tempera-
ture of the pressure vessel and the surrounding inner jacket. This temperature
increase is the measure of combustion calorimetry [53].
8.8 Thermal Analysis 393

Fig. 8.39 Combustion

calorimeter, schematic. A
pressure vessel filled with
oxygen B inner jacket filled
with water, C thermal
insulation, D outer jacket, E
case, F ignition wires, T
temperature sensor, P sample

The carbohydrates, lipids and proteins of the food sample are converted to CO2
and H2O during combustion. The gaseous water produced during combustion
condenses in the pressure vessel and forms carbonic acid with the CO2 from
combustion. The sulfur, nitrogen and phosphorus compounds contained in the
sample to a small extent are also oxidized. Their oxides form the associated acids
with the water in the pressure vessel. So after completion of the measurement, the
condensate in the pressure vessel thus contains carbonic acid and, depending on the
sample, traces of sulfuric acid, nitric acid, and phosphoric acid. The minerals and
water contained in the food sample do not provide any heat of combustion, as they
are usually already present in their oxidized form. In combustion calorimetry, a
distinction is made between adiabatic and isoperibolic mode of operation [54]. In
isoperibolic operation, the temperature of the outer jacket is kept constant and the
temperature of the measuring system (inner jacket) is evaluated relative to this
temperature. In adiabatic operation, the temperature of the outer jacket is kept at
the same value as the temperature of the measuring system (inner jacket) with the
help of an electric heater. This makes the heat losses of the measuring system
nearly zero.

Example
Minerals such as sodium, calcium, magnesium, iron, which are present in the
food as Na+, Ca2+, Mg2+ and Fe3+, are already in their highest oxidation state,
i.e., they do not contribute to the caloric value of a food.
394 8 Thermal Properties

Further Reading

Temperature
Temperature recording in logistics of frozen food [55]
Determination of added water in milk by freezing point determination [56]
Time-temperature indicators for monitoring food quality [57]

Thermal conductivity and diffusivity

Thermophysical properties of fruit [58]
Thermal diffusivity of foods using 1D Fourier cylindrical solution [59]
Thermal conductivity of a frozen dough [60]
Heat capacity measurement by photoacoustics [61]
Thermal conductivity and diffusivity measurement on food [62–66]
Thermal diffusivity of olive oil during high pressure treatment [67]
Heat transfer coefficients during deep-fat frying [68, 69]

Heat transfer during phase transition of water

Adiabatic chilling [70, 71]
Flow ice [21, 72–75]

Thermal Analysis
Denaturation of muscle protein by DSC [49]
Gelation and recrystallization of starch by DSC and MDSC [76–80]
Maize starch, examination by DSC [81]
DSC on fat modification [82]
Tapioca starch and trehalose—examination by DSC [83]
Glass transition by MDSC: optimization of test parameters [84]
Dehydration kinetics of dihydrate trehalose by DSC [85]
Thermal transition of spaghetti by DSC [86]
Honey crystallization, examination by DSC [87]
State diagram of bovine gelatin by DSC [88]
Glass transition temperature on Sea Bass during high pressure treatment [89]
Thermal transitions in potato starch, DSC [90]
Reactions and kinetics of green coffee during roasting [91]
Glass transition and crystallization of amorphous trehalose sucrose mixtures [92]
Glass transition temperatures of breakfast cereals, DSC [93]
Anhydrous lactose: determination of stability by DSC and x-ray diffraction [94]
Microcalorimety on ice slurries [95]
Glass transition phenomena in expanded cereal-based food extrudates, DSC [96]

(continued)
References 395

Gelation of starches, examination by DSC [77, 79]

Frozen food: glass transition and re-crystallization phenomena [97]

Standards
ASTM publications on DSC [98–102]
DIN publications on DSC [103–111]
DIN publications on combustion calorimetry [54, 112]

Summary
This chapter provides an understanding of thermal behavior in foods. Basic
concepts of thermodynamics, such as enthalpy, heat capacity and higher order
phase transitions are explained. Different heat transfer mechanisms and
geometries are presented step by step and illustrated with examples. The
caloric value of food, its influencing variables and measurement are described.
When discussing the methods of thermal analysis, Differential Scanning
Calorimetry (DSC) and Thermogravimetry (TG) are considered in detail. At
the end of the chapter, numerous application examples are listed, which can be
used for further study of methods for conducting thermal analysis.

References
1. BIPM (2018) On the revision of the International System of Units (SI). Bureau International de
Poids et Mesures, Versailles
2. Schmiermund T (2020) Die Avogadro-Konstante. Springer Nature, Wiesbaden. https://doi.
org/10.1007/978-3-658-29279-9
3. Kurzweil P (2013) Das Vieweg Einheiten-Lexikon : Begriffe, Formeln und Konstanten aus
Naturwissenschaften, Technik und Medizin. Vieweg, Wiesbaden. https://doi.org/10.1007/
978-3-322-92920-4
4. LMIV (2011) Lebensmittelinformationsverordnung Regulation No 1169/2011 of the
European Parliament. Brussels. http://data.europa.eu/eli/reg/2011/1169/oj
5. Kohlrausch F (1996) Praktische Physik Bd.3. Teubner, Stuttgart. https://doi.org/10.1002/piuz.
19870180109
6. Tscheuschner HD (2016) Grundzüge der Lebensmitteltechnik. Behrs, Hamburg
7. Roos Y (2010) Glass Transition Temperature and Its Relevance in Food Processing. Annu Rev
Food Sci Technol 1:469. https://doi.org/10.1146/annurev.food.102308.124139
8. Abiad MG, Carvajal MT, Campanella OH (2009) A review on methods and theories to
describe the glass transition phenomenon: applications in food and pharmaceutical products.
Food Eng Rev 1(2):105. https://doi.org/10.1007/s12393-009-9009-1
9. Kurzhals H-A (2003) Lexikon Lebensmitteltechnik. Behr’s Verlag, Hamburg
396 8 Thermal Properties

10. Davies GF, Man CMD, Andrews SD, Paurine A, Hutchins MG, Maidment GG (2012)
Potential life cycle carbon savings with low emissivity packaging for refrigerated food on
display. J Food Eng 109(2):202. https://doi.org/10.1016/j.jfoodeng.2011.10.018
11. Fellows PJ (2009) Food processing technology – principles and practice. Woodhead Publish-
ing, Sawston Cambridge
12. Singh RP, Heldman DR (2014) Introduction to food engineering. Academic Press, San Diego.
https://doi.org/10.1016/C2011-0-06101-X
13. VDI (2019) Wärmeatlas. Verein Deutscher Ingenieure, Düsseldorf. https://doi.org/10.1007/
978-3-662-52989-8
14. Toledo RT (2007) Fundamentals of food process engineering. Springer, Berlin
15. Singh RP, Heldman DR (2009) Introduction to food engineering. Academic Press/Elsevier,
San Diego
16. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, München
17. EN 673 (2011) Glas im Bauwesen - Bestimmung des Wärmedurchgangskoeffizienten
(U-Wert) - Berechnungsverfahren. Beuth, Berlin. https://doi.org/10.31030/3123472
18. EN 675 (2011) Glas im Bauwesen - Bestimmung des Wärmedurchgangskoeffizienten
(U-Wert) - Wärmestrommesser-Verfahren. Beuth, Berlin. https://doi.org/10.31030/3123472
19. EN ISO 6946 (2017) Bauteile - Wärmedurchlasswiderstand und Wärmedurchgangskoeffizient
- Berechnungsverfahren. Beuth, Berlin. https://doi.org/10.31030/3123472
20. DBU (2020) Heizen mit Eis. Deutsche Bundesstiftung Umwelt. https://www.dbu.de/123
ibook62425_28975_.html. Accessed 9 Jan 2021
21. Wang J, Battaglia F, Wang S, Zhang T, Ma Z (2019) Flow and heat transfer characteristics of
ice slurry in typical components of cooling systems: a review. Int J Heat Mass Transfer 141:
922. https://doi.org/10.1016/j.ijheatmasstransfer.2019.07.021
22. Grigull U (1989) Zustandsgrößen von Wasser und Wasserdampf in SI-Einheiten. 0-800 Grad
Celsius+ 0-1000 bar. Springer, Berlin
23. Tschubik IA, Maslow AM (1973) Wärmephysikalische Konstanten von Lebensmitteln und
Halbfabrikaten. VEB Fachbuchverlag, Leipzig
24. Hayes (1987) Food engineering data handbook. Wiley, New York, NY
25. Rao MA, Rizvi SSH, Datta AK (2005) Engineering properties of foods. Taylor & Francis,
Boca Raton
26. Rahman S (2009) Food properties handbook. CRC Press, Boca Raton. https://doi.org/10.1201/
9781420003093
27. Kohlrausch F (1996) Praktische Physik Bd.1. Teubner, Stuttgart. https://doi.org/10.1007/978-
3-322-87205-0
28. DIN EN ISO 22007 (2012) Kunststoffe - Bestimmung der Wärmeleitfähigkeit und der
Temperaturleitfähigkeit - Teil 4: Laserblitzverfahren. Beuth, Berlin. https://doi.org/10.
31030/3123472
29. Hammerschmidt U, Meier V (2006) New transient hot-bridge sensor to measure thermal
conductivity, thermal diffusivity, and volumetric specific heat. Int J Thermophys 27(3):840.
https://doi.org/10.1007/s10765-006-0061-2
30. Schmidt RF, Lang F, Heckmann M (2011) Physiologie des Menschen. Springer, Heidelberg.
https://doi.org/10.1007/978-3-642-01651-6
31. Herman IP (2016) Physics of the human body. Springer, New York. https://doi.org/10.1007/
978-3-319-23932-3
32. Benedek GB, Villars FMH (2000) Physics with illustrative examples from medicine and
biology. Springer, New York, NY
33. FAO (2001) Human energy requirements – Report of a Joint FAO/WHO/UNU Expert
Consultation. Food and Agriculture Organization of the United Nations, Rome
34. James WPT, Schofield EC (1990) Human energy requirements. A manual for planners and
nutritionists. FAO, Food and Agriculture Organization of the United Nations, New York, NY
35. Vaz M, Karaolis N, Draper A, Shetty P (2007) A compilation of energy costs of physical
activities. Public Health Nutr 8(7a):1153. https://doi.org/10.1079/PHN2005802
References 397

36. Lexikon der Ernährung (2001) Spektrum Akademischer Verlag. https://www.spektrum.de/

lexikon/ernaehrung/
37. Souci SW, Fachmann W, Kraut H (2016) Die Zusammensetzung der Lebensmittel, Nährwert-
Tabellen. Wissenschaftliche Verlagsgesellschaft (WVG), Stuttgart
38. Novotny JA, Gebauer SK, Baer DJ (2012) Discrepancy between the Atwater factor predicted
and empirically measured energy values of almonds in human diets. Am J Clin Nutr 96(2):296.
https://doi.org/10.3945/ajcn.112.035782
39. DIN 51005 (2019) Thermische Analyse (TA) - Begriffe; Text Deutsch und Englisch. Beuth,
Berlin. https://doi.org/10.31030/3045657
40. DIN EN ISO (2004) Futtermittel, tierische Produkte und Kot oder Urin - Bestimmung des
effektiven Brennwerts - Verfahren mit der kalorimetrischen Bombe, vol 9831. Beuth, Berlin.
https://doi.org/10.31030/9511774
41. Haines PJ, Kett V, Gaisford S (eds) (2016) Principles of thermal analysis and calorimetry.
Royal Society of Chemistry, Cambridge
42. Gabbott P (ed) (2008) Principles and applications of thermal analysis. Blackwell, Oxford.
https://doi.org/10.1002/9780470697702
43. Hemminger WF, Cammenga HK (1989) Methoden der Thermischen Analyse. Springer,
Berlin. [u.a.]
44. Fischer M, Wohlfahrt S, Varga J, Matuschek G, Saraji-Bozorgzad MR, Walte A, Denner T,
Zimmermann R (2016) Evolution of volatile flavor compounds during roasting of nut seeds by
thermogravimetry coupled to fast-cycling optical heating gas chromatography-mass spectrom-
etry with electron and photoionization. Food Anal Method 10(1):49. https://doi.org/10.1007/
s12161-016-0549-8
45. Czarnecki J (2015) Precision thermogravimetry. J Therm Anal Calorim 120(1):139. https://
doi.org/10.1007/s10973-014-4384-0
46. Li R, Lin D, Roos YH, Miao S (2019) Glass transition, structural relaxation and stability of
spray-dried amorphous food solids: a review. Dry Technol 37(3):287. https://doi.org/10.1080/
07373937.2018.1459680
47. Raemy A, Lambelet P, Rousset P (2005) Calorimetric information about food and food
constituents. In: Lörinczy D (ed) The nature of biological systems as revealed by thermal
methods. Springer, Dordrecht, pp 69–98. https://doi.org/10.1007/1-4020-2219-0_4
48. Raemy A (2003) Behavior of foods studied by thermal analysis: introduction. J Therm Anal
Calorim 71(1):273. https://doi.org/10.1023/A:1022299124618
49. Levitsky DI (ed) (2004) The nature of biological systems as revealed by thermal methods.
Kluwer Academic, Dordrecht
50. Figura LO (2003) Thermoanalytische Charakterisierung teilkristalliner Trehalose. In: Kunze
W (ed) Anwenderseminar Thermische Analyse in der pharmazeutischen Industrie und der
Lebensmitteltechnologie. Reuters, Alzenau, pp 125–137
51. Merzlyakov M, Schick C (2000) Optimization of experimental parameters in TMDSC. The
influence of non-linear and non-stationary thermal response. J Therm Anal Calorim 61(2):649.
https://doi.org/10.1023/A:1010106626114
52. Höhne GWH, Merzlyakov M, Schick C (2002) Calibration of magnitude and phase angle of
TMDSC: Part1: Basic considerations. Thermochim Acta 391(1):51. https://doi.org/10.1016/
S0040-6031(02)00163-6
53. Lamprecht I (1999) Chapter 4 – Combustion calorimetry. In: Kemp RB (ed) Handbook of
thermal analysis and calorimetry, vol 4. Elsevier Science B.V., pp 175–218. https://doi.org/10.
1016/S1573-4374(99)80007-7
54. DIN 51900 (2000) Prüfung fester und flüssiger Brennstoffe - Bestimmung des Brennwertes
mit dem Bomben-Kalorimeter und Berechnung des Heizwertes. Beuth, Berlin. https://doi.org/
10.31030/8511091
55. BVL L 00.00-5 (1982) Ambulante Temperaturmessung bei gefrorenen und tiefgefrorenen
Lebensmitteln; Referenzverfahren. Bundesamt für Verbraucherschutz und
Lebensmittelsicherheit, Berlin
398 8 Thermal Properties

56. Deutsches Institut für Normung (2009) DIN EN ISO 5764 Milch - Bestimmung des
Gefrierpunktes - Thermistor-Kryoskop-Verfahren, Referenzverfahren. Beuth, Berlin. https://
doi.org/10.31030/1508099
57. Gao T, Tian Y, Zhu Z, Sun D-W (2020) Modelling, responses and applications of time-
temperature indicators (TTIs) in monitoring fresh food quality. Trends Food Sci Technol 99:
311. https://doi.org/10.1016/j.tifs.2020.02.019
58. Mukama M, Ambaw A, Opara UL (2020) Thermophysical properties of fruit—a review with
reference to postharvest handling. J Food Meas Charact 14(5):2917. https://doi.org/10.1007/
s11694-020-00536-8
59. Baïri A, Laraqi N, de María JMG (2007) Determination of thermal diffusivity of foods using
1D Fourier cylindrical solution. J Food Eng 78(2):669. https://doi.org/10.1016/j.jfoodeng.
2005.11.004
60. Kumcuoglu S, Tavman S, Nesvadba P, Tavman IH (2007) Thermal conductivity
measurements of a traditional fermented dough in the frozen state. J Food Eng 78(3):1079.
https://doi.org/10.1016/j.jfoodeng.2005.12.020
61. Vera-Medina G, Marín E, Calderón A, Díaz Góngora JAI, Peña-Rodríguez G, Delgado-
Vasallo O (2013) A method for heat capacity measurement by photoacoustics. Measurement
46(3):1208. https://doi.org/10.1016/j.measurement.2012.10.032
62. Glavina MY, Di Scala KC, Ansorena R, del Valle CE (2006) Estimation of thermal diffusivity
of foods using transfer functions. LWT – Food Sci Technol 39(5):455. https://doi.org/10.1016/
j.lwt.2005.03.010
63. Mohamed IO (2010) Development of a simple and robust inverse method for determination of
thermal diffusivity of solid foods. J Food Eng 101(1):1. https://doi.org/10.1016/j.jfoodeng.
2010.05.002
64. Carson JK, Wang J, North MF, Cleland DJ (2016) Effective thermal conductivity prediction of
foods using composition and temperature data. J Food Eng 175:65. https://doi.org/10.1016/j.
jfoodeng.2015.12.006
65. Krokida MK, Michailidis PA, Maroulis ZB, Saravacos GD (2002) Literature data of thermal
conductivity of foodstuffs. Int J Food Properties 5(1):63. https://doi.org/10.1081/
Jfp-120015594
66. Krokida MK, Zogzas NP, Maroulis ZB (2001) Mass transfer coefficient in food processing:
compilation of literature data. Int J Food Properties 4(3):373. https://doi.org/10.1081/
JFP-100108643
67. Kubásek M, Houška M, Landfeld A, Strohalm J, Kamarád J, Žitný R (2006) Thermal
diffusivity estimation of the olive oil during its high-pressure treatment. J Food Eng 74(3):
286. https://doi.org/10.1016/j.jfoodeng.2005.03.019
68. Håkansson A (2019) Estimating convective heat transfer coefficients and uncertainty thereof
using the general uncertainty management (GUM) framework. J Food Eng 263:53. https://doi.
org/10.1016/j.jfoodeng.2019.05.031
69. Safari A, Salamat R, Baik OD (2018) A review on heat and mass transfer coefficients during
deep-fat frying: determination methods and influencing factors. J Food Eng 230:114. https://
doi.org/10.1016/j.jfoodeng.2018.01.022
70. Ulpiani G (2019) Water mist spray for outdoor cooling: a systematic review of technologies,
methods and impacts. Appl Energy 254:113647. https://doi.org/10.1016/j.apenergy.2019.
113647
71. Yang Y, Cui G, Lan CQ (2019) Developments in evaporative cooling and enhanced evapora-
tive cooling – a review. Renew Sustain Energy Rev 113:109230. https://doi.org/10.1016/j.rser.
2019.06.037
72. Zou L, Zhang X, Zheng Q (2020) Research progress on preparation of binary ice by vacuum
flash evaporation: a review. Int J Refrig. https://doi.org/10.1016/j.ijrefrig.2020.10.005
73. Brooks S, Quarini G, Tierney M, Yun X, Lucas E (2020) Conditions for continuous ice slurry
generation in a nylon helical coiled heat exchanger. Therm Sci Eng Prog 15:100427. https://
doi.org/10.1016/j.tsep.2019.100427
References 399

74. Kauffeld M, Wang MJ, Goldstein V, Kasza KE (2010) Ice slurry applications. Int J Refrig
33(8):1491. https://doi.org/10.1016/j.ijrefrig.2010.07.018
75. Kauffeld M, Kawaji M, Egolf PW (eds) (2005) Handbook on ice slurries – fundamentals and
engineering. Int Inst Refrig, Paris
76. De Meuter P, Rahier H, Van Mele B (2004) Recrystallisation of starch studied with MDSC. In:
Lörinczy D (ed) The nature of biological systems as revealed by thermal methods. Kluwer
Academic, Dordrecht
77. Rangelov A, Arnaudov L, Stoyanov S, Spassov T (2017) Gelatinization of industrial starches
studied by DSC and TO. Bulg Chem Commun 49(2):422
78. Kovrlija R, Rondeau-Mouro C (2017) Hydrothermal changes of starch monitored by com-
bined NMR and DSC methods. Food Bioproc Technol 10(3):445. https://doi.org/10.1007/
s11947-016-1832-9
79. Alamri MS, Al-Ruquie IM, Hussain S, Mohamed AA, Xu J (2015) Gelatinisation kinetics of
corn and chickpea starches using DSC, RVA and dynamic rheometry. Qual Assur Saf Crop
Food 7(4):459. https://doi.org/10.3920/Qas2013.0374
80. Xie FW, Liu WC, Liu P, Wang J, Halley PJ, Yu L (2010) Starch thermal transitions
comparatively studied by DSC and MTDSC. Starch-Starke 62(7):350
81. Genkina NK, Kozlov SS, Martirosyan VV, Kiseleva VI (2014) Thermal behavior of maize
starches with different amylose/amylopectin ratio studied by DSC analysis. Starch-Starke
66(7–8):700. https://doi.org/10.1002/star.201300220
82. De Oliveira GM, Masuchi MH, Basso RC, Zuliani Stroppa VL, Ribeiro APB, Kieckbusch TG
(2014) DSC application to lipid modification processes. CRC Press, pp 221–242. https://doi.
org/10.1201/b17739
83. Zhang XY, Zhu WX, Tong QY, Ren F (2012) Rheological, thermal properties, and gelatini-
zation kinetics of tapioca starch-trehalose blends studied by non-isothermal DSC technology.
Starch-Starke 64(12):996. https://doi.org/10.1002/star.201200111
84. Xiville NR, Lorente LT, Kordikowski A (2012) MDSC parameter optimization for the
determination of glass transitions using a design of experiments approach. Int J Pharm
422(1–2):271. https://doi.org/10.1016/j.ijpharm.2011.11.009. Epub 2011 Nov 15
85. Verhoeven N, Neoh TL, Furuta T, Yamamoto C, Ohashi T, Yoshii H (2012) Characteristics of
dehydration kinetics of dihydrate trehalose to its anhydrous form in ethanol by DSC. Food
Chem 132(4):1638. https://doi.org/10.1016/j.foodchem.2011.06.010
86. Rahman MS, Senadeera W, Al-Alawi A, Truong T, Bhandari B, Al-Saidi G (2011) Thermal
transition properties of spaghetti measured by differential scanning calorimetry (DSC) and
thermal mechanical compression test (TMCT). Food Bioproc Technol 4(8):1422. https://doi.
org/10.1007/s11947-009-0258-z
87. Venir E, Spaziani M, Maltini E (2010) Crystallization in “Tarassaco” Italian honey studied by
DSC. Food Chem 122(2):410. https://doi.org/10.1016/j.foodchem.2009.04.012
88. Rahman MS, Al-Saidi G, Guizani N, Abdullah A (2010) Development of state diagram of
bovine gelatin by measuring thermal characteristics using differential scanning calorimetry
(DSC) and cooling curve method. Thermochim Acta 509(1–2):111. https://doi.org/10.1016/j.
tca.2010.06.011
89. Tironi V, de Lamballerie-Anton M, Le-Bail A (2009) DSC determination of glass transition
temperature on sea bass (Dicentrarchus labrax) muscle: effect of high-pressure processing.
Food Bioproc Technol 2(4):374. https://doi.org/10.1007/s11947-007-0041-y
90. Steeneken PAM, Woortman AJJ (2009) Identification of the thermal transitions in potato
starch at a low water content as studied by preparative DSC. Carbohydr Polym 77(2):288
91. Strezov V, Evans TJ (2005) Thermal analysis of the reactions and kinetics of green coffee
during roasting. Int J Food Prop 8(1):101. https://doi.org/10.1081/JFP-200048060
92. Roe KD, Labuza TP (2005) Glass transition and crystallization of amorphous trehalose-
sucrose mixtures. Int J Food Prop 8:559. https://doi.org/10.1080/10942910500269824
93. Sandoval A, Nuñez M, Müller A, Della Valle G, Lourdin D (2009) Glass transition
temperatures of a ready to eat breakfast cereal formulation and its main components
400 8 Thermal Properties

determined by DSC and DMTA. Carbohydr Polym 76:528. https://doi.org/10.1016/j.carbpol.

2008.11.019
94. Figura LO, Epple M (1995) Anhydrous α-lactose. A study with DSC and TXRD. J Therm
Anal 44(1):45. https://doi.org/10.1007/BF02547132
95. Torré J-P, Plantier F, Marlin L, André R, Haillot D (2020) A novel stirred microcalorimetric
cell for DSC measurements applied to the study of ice slurries and clathrate hydrates. Chem
Eng Res Des 160:465. https://doi.org/10.1016/j.cherd.2020.06.019
96. Masavang S, Roudaut G, Champion D (2019) Identification of complex glass transition
phenomena by DSC in expanded cereal-based food extrudates: impact of plasticization by
water and sucrose. J Food Eng 245:43. https://doi.org/10.1016/j.jfoodeng.2018.10.008
97. Roos YH (2021) Glass transition and re-crystallization phenomena of frozen materials and
their effect on frozen food quality. Foods 10(2). https://doi.org/10.3390/foods10020447
98. ASTM D3418 (2015) Standard test method for transition temperatures and enthalpies of fusion
and crystallization of polymers by differential scanning calorimetry. ASTM International,
West Conshohocken, PA. https://doi.org/10.1520/D3418-15
99. ASTM E793 (2001) Standard test method for enthalpies of fusion and crystallization by
differential scanning calorimetry. ASTM International, West Conshohocken, PA. https://doi.
org/10.1520/E0793-01
100. ASTM E794 (2018) Standard test method for melting and crystallization temperatures by
thermal analysis. ASTM International, West Conshohocken, PA. https://doi.org/10.1520/
E0794-06R18
101. ASTM E1269 (2018) Standard test method for determining specific heat capacity by differen-
tial scanning calorimetry. ASTM International, West Conshohocken, PA. https://doi.org/10.
1520/E1269-11R18
102. ASTM E1356 (2014) Standard test method for assignment of the glass transition temperatures
by differential scanning calorimetry. ASTM International, West Conshohocken, PA. https://
doi.org/10.1520/E1356-08R14
103. DIN 51007 (2019) Thermische Analyse (TA) - Differenz-Thermoanalyse (DTA) und
Dynamische Differenzkalorimetrie (DSC) - Allgemeine Grundlagen. Beuth, Berlin. https://
doi.org/10.31030/3025544
104. DIN EN ISO 11357-1 (2017) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -
Teil 1: Allgemeine Grundlagen I. Beuth, Berlin. https://doi.org/10.31030/2589772
105. DIN EN ISO 11357-2 (2020) Kunststoffe - Dynamische Differenzkalorimetrie (DSC) - Teil 2:
Bestimmung der Glasübergangstemperatur und der Glasübergangsstufenhöhe. Beuth, Berlin.
https://doi.org/10.31030/3127571
106. DIN EN ISO 11357-3 (2018) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -
Teil 3: Bestimmung der Schmelz- und Kristallisationstemperatur und der Schmelz- und
Kristallisationsenthalpie. Beuth, Berlin. https://doi.org/10.31030/2798705
107. DIN EN ISO 11357-4 (2020) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -
Teil 4: Bestimmung der spezifischen Wärmekapazität. Beuth, Berlin. https://doi.org/10.31030/
3154215
108. DIN EN ISO 11357-5 (2014) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -
Teil 5: Bestimmung von charakteristischen Reaktionstemperaturen und -zeiten,
Reaktionsenthalpie und Umsatz. Beuth, Berlin. https://doi.org/10.31030/2143352
References 401

109. DIN EN ISO 11357-6 (2018) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -

Teil 6: Bestimmung der Oxidations-Induktionszeit (isothermische OIT) und Oxidations-
Induktionstemperatur (dynamische OIT). Beuth, Berlin. https://doi.org/10.31030/2798706
110. DIN EN ISO 11357-7 (2015) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -
Teil 7: Bestimmung der Kristallisationskinetik. Beuth, Berlin. https://doi.org/10.31030/
2309304
111. DIN EN ISO 11357-8 (2020) Kunststoffe - Dynamische Differenz-Thermoanalyse (DSC) -
Teil 8: Bestimmung der Wärmeleitfähigkeit. Beuth, Berlin. https://doi.org/10.31030/3123472
112. DIN EN ISO 9831 (2004) Futtermittel, tierische Produkte und Kot oder Urin - Bestimmung
des effektiven Brennwerts. Beuth, Berlin. https://doi.org/10.31030/9511774
Electrical Properties
9

The electrical properties of food play a role in electrical processes in food technology
such as the conductive heating of food, the treatment with electric fields (pulsed
electric pulses, PEF), or quality checks by means of metal detectors. While in the
field of direct current technology (DC), the electrical conductivity of the materials or
their electrical resistance is of main interest (cf. Table Table 9.1), in alternating
current technology (AC) additional variables like impedance and reactance are of
interest. In Table 9.2, some of these sizes are listed with their synonymous names.
Alternating current resistors have a frequency-dependent resistance, which makes
things more complicated, but also increases the possibilities of technical
applications. To begin this chapter, let us start with the electrical conductivity in
the DC case (i.e., the frequency is 0) before we deal with AC properties. The
following chapters then deal with the magnetic properties (Chap. 10) and the
electromagnetic properties (Chap. 11) of food.

9.1 Electrical Conductivity

Foods contain positively charged and negatively charged ions, molecules, or

macromolecules and are therefore in principle able to conduct electric current. Positively
charged ions are called cations and negatively charged anions. In addition to the
presence of charge carriers, the prerequisite for the electrical conduction is their mobility
in the surrounding matrix. In a homogeneous electric field with the field strength E, on
an electrically charged particle with the charge qC the Coulomb force acts:
F C = E
qC ð9:1Þ

When the charge carriers show mobility, the Coulomb force causes a migration of
the charged particles according to the sign of the charge and the direction of field E.
The current density is proportional to the potential gradient (refer transport equation
in Chap. 7). Written in non-vectorial form, it is:

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 403
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_9
404 9 Electrical Properties

Table 9.1 Electrical variables, terminology

Symbol Name Unit Remark
R Electrical resistance Ω (Ohm) Ω = V
A-1
G Electrical conductance S (Siemens) S = Ω-1 = A
V-1
κ Electrical conductivity S
m-1 κ = 1ρ
ρ Electrical resistivity Ω
m-1 ρ = 1κ
C Capacitance F (Farad) 1 F = 1 C
V-1
I Inductance H (Henry) 1 H = 1 kg
m2
s-2
A-2

Table 9.2 Electrical quantities in alternating current technology, terminology

Name Synonym Unit
Impedance Complex electrical resistance Ω
Resistance Real part of impedance Ω
Ohmic resistance
Reactance Imaginary part of impedance Ω
Admittance Complex electrical conductance S
Reciprocal of impedance
Conductance Real part of conductance S
Susceptance Imaginary part of conductance S
Inductive reactance Inductive part of reactance Ω
Inductive resistance
Capacitive reactance Capacitive part of reactance Ω
Capacitive resistance

1 dqC dφ

=κ
ð9:2Þ
A dt dx
so
I U
=κ
ð9:3Þ
A d
or

A
I =κ

U ð9:4Þ
d
resp.

I =G
U ð9:5Þ

or

1
I=
U ð9:6Þ
R
9.1 Electrical Conductivity 405

E electric field strength in V
m-1

U electric voltage in V
φ electric potential in V
q electric charge in C
d distance in m
A area in m2
G conductance in S
κ specific electrical conductivity in S
m-1
t time in s
x length in m

In the presence of mobile charge carriers, after applying an electrical voltage U,

an electric current flows through a material, wherein the strength I of the current is
determined by the resistance R of the material under consideration. The reciprocal of
R is called electrical conductance G. A linear dependence on current and voltage is
known as Ohm’s law. Influencing variables on electrical conductivity are the
concentration of the charge carriers, their amount of charge (charge number), and,
above all, their mobility. The mobility of the charge carriers is the quotient of their
drift velocity in the electric field and the field strength E. The influencing variables
on the electrical conductivity of a food are summarized in Table 9.3.
To determine the electrical resistance of materials regardless of their geometry
(length l, cross-sectional area A, cf. Fig. 9.1) to be able to compare, the so-called
specific electrical resistance ρ is used.
A
ρ=R
ð9:7Þ
l

Table 9.3 Influencing factors on electrical conductivity of a food

Influencing factor Notes
Concentration of charge carriers Recipe, salinity, acidity
Charge number of charge carriers Monovalent or polyvalent ions
Mobility of the charge carriers Physical state, molecular weight, hydration, ion cloud

Fig. 9.1 Electric current U

flowing through a material,
schematic
I

l
406 9 Electrical Properties

resp.

l
R=ρ
ð9:8Þ
A
The reciprocal size of ρ is called specific electrical conductivity κ. The specific
conductivity κ of a food as well as the specific electrical resistance ρ is geometry-
independent, i.e., a pure material property. It depends on the composition of the
material (e.g., water content, salinity) and the state of the material (temperature,
pressure) and its physical properties, e.g., viscosity.

1
κ= ð9:9Þ
ρ

U electric voltage in V
I electric current in A
R electric resistance in Ω
G electric conductance in S
A area in m2
l distance in m

The SI unit of electrical resistance is V
A-1 = Ω(Ohm); the unit reciprocal to

Ohm is called Siemens: Ω-1 = S

→ Attention
A non-SI conform unit reciprocal to ohm Ω is mho.
Example: 1 mho
cm-1 = 100 S
m-1, 10 μmho = 10-5S.

In Table 9.4, values of the specific electrical resistance for different materials are
compiled.

Table 9.4 Specific electri- Material ϑ (°C) ρ (Ω
m) Ref.

cal resistance of some
KCl solution 0.01 M 25 7.08 [1]
liquids
KCl solution 0.1 M 25 0.78 [1]
KCl solution 1 M 25 0.09 [1]
Milk, 1.5% fat 20 0.57 [2]
Orange juice 20 0.44 [2]
Apple juice 20 0.29 [2]
Wheat beer 20 0.20 [2]
Tap water 20 0.06 [2]
Tomato juice 20 1.5 [2]
Sauerkraut juice 20 2.1 [2]
9.1 Electrical Conductivity 407

For applications such as ohmic heating of food, electrical conductivity is a

fundamental parameter. The electrical treatment of food, e.g., for heating, disinfec-
tion, or cell perforation, is usually carried out with alternating current fields. For
solid materials with a fibrous structure, the electrical conductivity can have different
values in different spatial directions. This is called anisotropy of electrical
conductivity.

9.1.1 Temperature Dependence of Electrical Conductivity

With increasing temperature, the viscosity of materials decreases, thus increasing the
mobility of dissolved ions. In the case of weak electrolytes, the degree of dissocia-
tion also increases with increasing temperature. For many liquid foods, the
temperature-related increase in electrical conductivity can be estimated with the
following linear relationship [2]:
κ = κ0 ð1 þ b0
ΔT Þ ð9:10Þ

κ electrical conductivity in S
m-1

κ0 electrical conductivity in S
m-1 at reference temperature
b0 temperature coefficient in K-1
ϑR reference temperature in °C
ϑ temperature in °C
ΔT = ϑ - ϑR

The linear relationship in Fig. 9.2 was quantified by Reitler [2] for a range of
foods. An influence of the measuring frequency in the range of 50–3000 Hz did not
occur. With tap water, however, there is an increasing loss of temporary water
hardness starting at about 100 °C. The cause of the temporary water hardness are
carbonate ions, which disappear with increasing CO2 outgassing. Therefore,
deviations from the linear behavior arise here.

Fig. 9.2 Approximately

linear temperature
dependence of the electrical M
conductivity of some liquid
foods [2]. M milk, W
W wheat beer N

- / qC
408 9 Electrical Properties

9.1.2 Electrolyte Solutions

The electrical conductivity of liquid foods is based on the presence of mobile ions.
Electrolyte solutions of suitable concentration can therefore serve as model
solutions. For the electrical conductivity of electrolyte solutions, the so-called
equivalent conductivity is used; this is the quotient of specific electrical conductivity
κ and electrolyte concentration c.
κ
Λ= ð9:11Þ
ν
c

Λ equivalent conductivity in S
m-2
mol-1

κ electric conductivity in S
m-1
c electrolyte concentration in mol
m-3
ν equivalent number of ions
n charge number
z stoichiometric factor

The concept of equivalent conductivity arises from the study of physical chemis-
try. It provides a means by which we can compare electrolyte solutions containing
the same concentration of charged ion carriers, although made of different
substances, but can exhibit different equivalent electrical conductivity. For example,
if we compare an aqueous solution of sodium chloride with one of acetic acid of
equal concentration, they will have different equivalent electrical conductivity. This
means that there must be a difference in the degree of mobility or dissociation of the
ions in the two electrolyte solutions.
Referring back to the definition of equivalent electrical conductivity in Eq. (9.11),
the equivalent number of ions ν is the product of the stoichiometric factor z of either
a cation or anion of a dissociating electrolyte and its charge number n. Some
examples are given in Table 9.5.
If we dissolve 1 mol Al2(SO4)3 in a liter of water, we have six-times the number
of charge carriers as with 1 mol of NaCl. So, in terms of electrical conductivity, we
can say a 1 M Al2(SO4)3 solution is equivalent to a 6 M NaCl solution.
With increasing electrolyte concentration, the degree of dissociation of
electrolytes decreases, as illustrated in Fig. 9.3. For strong electrolytes, Kohlrausch
found that there is a linear relationship between equivalent electrical conductivity
and the square root of the concentration. This is called Kohlrausch’s square root
law

Table 9.5 Equivalent Electrolyte system (z
n)cation = (z
n)anion = ν

number of electrolytes for
Al2(SO4)3 → 2 Al3+ + 3 SO42- 2
3 = 3
2 = 6
different systems, examples
H3PO4 → 3 H+ + PO43- 3
1 = 1
3 = 3
NaCl → Na+ + Cl- 1
1 = 1
1 = 1
CaCl2 → Ca2+ + 2 Cl- 1
2 = 2
1 = 2
9.1 Electrical Conductivity 409

Fig. 9.3 Concentration

dependency of equivalent
conductivity. Strong
electrolyte (upper curve, a NaCl
NaCl) and weak electrolyte
(lower curve, acetic acid)
Λ
HAc

p
Λ = Λ0 - a c ð9:12Þ

Λ equivalent conductivity
Λ0 equivalent conductivity at c → 0
c concentration
a constant

Figure 9.3 shows that the concentration dependency of equivalent electrical

conductivity for strong electrolytes, like sodium chloride, will give a straight line
(linear relationship) on a graph of equivalent conductivity versus square root of
concentration, and thus they follow Kohlrausch’s square root law. However, weak
electrolytes, such as acetic acid, do not give a straight line on this type of graph, and
they do not follow Kohlrausch’s square root law.
The extrapolation of the straight-line curve to c = 0 will intercept the vertical axis
at a value of equivalent electrical conductivity designated Λ0. This is the equivalent
conductivity of an electrolyte at an extreme dilution, in which the concentration
approaches zero where intermolecular interactions disappear. Values can be calcu-
lated (Eq. 9.12) with the help of tables where the equivalent conductivity for each ion
of the solution is listed; see Table 9.6.

Λ0 = λ0,i ðanionsÞ þ λ0,j ðcationsÞ ð9:13Þ

i j

Λ0 equivalent conductivity of electrolyte in S
cm2
mol-1

λ0 equivalent conductivity of ion S
cm2
mol-1
410 9 Electrical Properties

Table 9.6 Values of Ion λ0 (S
cm2
mol-1)

equivalent conductivity for
Temperature 25 °C 18 °C
some ions [3]
H+ 349.6 315.0
Na+ 50.1 43.5
NH4+ 73.5 64.0
1 2+
2Mg
50.3 45.0
1 2+
2Ca
59.0 56.0
-
OH 199.1 174.0
Cl- 76.4 65.5
NO3- 71.5 61.7
CH3COO- 40.9 34.5
CO32- 138.6 121.0

→ Attention
Whenever we take values of λ0 from tables, we have to be careful to avoid
confusion. For example, the equivalent conductivity of the calcium ion can be
given as λ0 12 Ca2þ = 59 S
cm2
mol - 1, or λ0(Ca2+) = 118 S
cm2
mol-1.
In the first case, the value is for a hypothetical half of a calcium ion ( 12 Ca2þ -
ion), and is for use in the following equation:
1 1
λ0 = λ þ λ ð9:14Þ
n 0 Cation n 0 Anion

Example
Calculation of the electric conductivity of an aqueous 0.5% (m/m) solution of
CaCl2
Solution:
Because of the dissociation CaCl2 → Ca2+ + 2 Cl the equivalent number is
ν=2
With the molar mass M(CaCl2) = 111 g
mol-1 we get a concentration of:
5g 5g 5g
c= ≈ = = 0:045mol
l - 1
1000g 1l 111g
mol - 1
1l
c = 4:5
10 - 5 mol
cm - 3

So, the equivalent conductivity is:

(continued)
9.1 Electrical Conductivity 411

Λ0 = λ0,Ca þ λ0,Cl
Λ0 = 59:0 S
cm2
mol - 1 þ 76:4 S
cm2
mol - 1 = 135:4 S
cm2
mol - 1

With use of Eq. (9.11) and without use of Kohlrausch’s square root law, we
get

κ = 135:4 S
cm2
mol - 1
4:5
10 - 5 mol
cm - 3
2 = 1:2 S
m - 1

Example
Calculation of the electric conductivity of an aqueous solution (20 °C) of 0.3%
(m/m) CaCl2 and 0.1% (m/m) NaCl.
Solution:
We have
M (CaCl2) = 111 g
mol-1, ν (CaCl2) = 2, M (NaCl) = 58.4 g
mol-1, ν
(NaCl) = 1
The concentration can be calculated as:

3 g 3 g 3g
cðCaCl2 Þ = ≈ = = 0:027mol
l - 1
1000 g 1l 111g
mol - 1
1l
c = 2:7
10 - 5 mol
cm - 3

1g 1g 1g
cðNaClÞ = ≈ = = 0:017mol
l - 1
1000 g 1l 58:4g
mol - 1
1l
c = 17
10 - 5 mol
cm - 3

So, the equivalent conductivities are:

Λ0 ðNaClÞ = λ0,Naþ þ λ0,Cl - = ð50:1 þ 76:4ÞS
cm2
mol - 1

Λ0 = 126:5 S
cm2
mol - 1

Λ0 ðCaCl2 Þ = λ0,Ca2þ þ λ0,Cl - = ð59:0 þ 76:4ÞS
cm2
mol - 1

Λ0 = 135:4 S
cm2
mol - 1

Now, we use Kohlrausch’s square root law:

p
Λ = Λ0 - a c

with aðNaClÞ ≈ aðCaCl2 Þ = 80 S
cm2
mol2
l - 2

1 1

(continued)
412 9 Electrical Properties

p
ΛðNaClÞ = 126:5 - 80
0:017 S
cm2
mol - 1 = 116:1 S
cm2
mol - 1
p
ΛðCaCl2 Þ = 135:4 - 80
0:027 S
cm2
mol - 1 = 122:3 S
cm2

mol - 1

At last, we calculate the specific electric conductivity by use of Eq. (9.11)

κ = Λ
ν
c = ðΛ
ν
cÞNaCl þ ðΛ
ν
cÞCaCl2
κ = 116:1
1:7
10 - 5 þ 122:3
2
2:7
10 - 5 S
cm - 1
κ = 858 mS
m - 1

To get a rough estimation, we could leave out Kohlrausch’s law by

assuming a = 0; d.h. Then, we get κ = 946 mS
m-1 producing an error of
approximately 10%.

9.1.2.1 Frequency Dependency

Dissociated ions moving freely about in an aqueous electrolyte solution are always
surrounded by other nearby ions of opposite charge. Consider a solution of potas-
sium chloride (KCl). Positively charged potassium ions (K+) will be surrounded by
negatively charged chloride ions (Cl-). When placed within an electric field, the
positively charged potassium ion will experience force acting upon it in an opposite
direction to the force acting upon all the negatively charged chloride ions
surrounding it.
In the absence of an electric field, each positively charged ion together with its
surrounding negatively charged ions forms a sphere-like grouping within the elec-
trolyte solution. When placed within an electric field, the positively charged ions are
pulled in one direction, and the negatively charged ions are pulled in the opposite
direction. This causes a deformation in the shape of the initial spherical grouping; see
Fig. 9.4. The deformation causes a drag force on the central ion. This drag force
effect after Debye–Hückel–Onsager is called the cataphoretic effect.

Fig. 9.4 Cataphoretic effect

(I initial state, II response to
electric field)

I II
9.1 Electrical Conductivity 413

Fig. 9.5 Ions with

surrounding water
2–
2–
molecules—hydrated ions
2–

2–
2–
2– –
2– 2 2–
2– 2– 2– 2–
2– 2–
2–

If an electrolyte solution is exposed to an alternating electric field, the deforma-

tion of the ion cloud changes in the rhythm of the electric field. With increasing
frequency, the retarding effect of the electrophoretic and cataphoretic effect on ion
migration becomes lower; i.e., the specific electrical conductivity of the electrolyte
solution increases. The effect of an increasing conductivity of electrolyte solutions
with increasing frequency is called the Debye–Falkenhagen effect. We will see
below that this is the behavior of a capacitive resistor.
In addition to the ions, themselves, within an ion grouping, each grouping also
has a surrounding cloud of solvent molecules (mostly water), as illustrated in
Fig. 9.5. When an ion now starts to travel in response to forces established by an
electric field, this cloud consisting of solvent molecules causes an additional drag
force. This is called the electrophoretic effect. This effect is used in molecular
analysis with laboratory procedures called electrophoresis. Molecules with different
drag forces due to the electrophoretic effect can be separated by their different
traveling velocities in an electric field. A common application is the investigation
of proteins with different sizes, different side groups, or different stearic states. To
enhance the effect of different electrophoretic mobility, the molecules of interest are
brought into a gel, which is exposed then to an electric field (gel electrophoresis). If
an electrophoretic analysis is sensitive to molecular size only, we have to make the
analysis “blind” to the fact that molecules carry an electric charge and are charged
differently. This can be done by adding an ionic surfactant. Often, sodium dodecyl
sulfate (SDS) is used for this purpose. It links itself to the charged sites, and thus
compensates for ionic effects. When SDS is used on polyacrylamide gel, this
technique is called SDS PAGE (SDS polyacrylamide gel electrophoresis).

9.1.2.2 Temperature Dependency of Equivalent Conductivity

Both the electrophoretic and the cataphoretic effects depend on the viscosity of the
electrolyte solution. We learned earlier that the viscosity of many liquid foods is
temperature dependent (see Chap. 4) and will usually decrease with increasing
temperature. This decrease in viscosity will lower drag forces, and thus decrease
414 9 Electrical Properties

both effects with increasing temperature. Therefore, the equivalent conductivity of

an aqueous electrolyte solution will increase with increasing temperature, just as in
the case with electrical conductivity. In order to describe this behavior, we need to
introduce a temperature coefficient for the equivalent conductivity:
1 dΛ
k=
ð9:15Þ
Λ dT

k temperature coefficient in K-1

Λ equivalent conductivity in S
m-2
mol-1
T temperature in K

Example
Change in conductivity of water on raising the temperature from 20 °C to 21 °C

ΔΛ
= k
ΔT
Λ
so

ΔΛ
= 0:058 K - 1
2 K = 0:116 = 11:6%
Λ

This example illustrates the significant effect that temperature has on the equiva-
lent electrical conductivity. Therefore, it is imperative that laboratory measurements
of electrical conductivity be carried out under conditions of precise temperature
control, with the controlled temperature clearly and prominently recorded and
reported with the results of the measurement. Measurements made without control
or reporting of the temperature are of little value.

9.1.3 Conductivity of Solid Foods

When considering the electrical conductivity of biological materials such as food

and agricultural products, it is useful to distinguish between non-structured and cell-
like structured materials. Unprocessed foods of plant origin consist predominantly of
plant cells. Their cell walls consist largely of non-mobile molecules, so that the cell
walls act roughly simplified like electrical insulators. Although such foods contain
water and electrolytes, the electrical conductivity is low because the electrolytes in
aqueous solution are extracellularly and intracellularly enclosed between
non-conductive cell walls. In foods of animal origin such as meat and fish, intact
cells cause analogously a low electrical conductivity of the material. Once the cell
walls are damaged, the conductivity of these materials increases.
9.1 Electrical Conductivity 415

Fig. 9.6 Temperature

dependence of the electrical
A
conductivity of bananas,
schematic [2]. A—puree, B
B—chunks N

- / qC

Fig. 9.7 Frequency

dependence of the impedance
Z of potato. Intact tissue (III),
with cell damage (II),
maximum crushed tissue
(I) [6]

For this reason, crushed plant and animal foods have a much higher electrical
conductivity than their non-crushed original materials. Maximum milled products
behave like non-structured materials. For this reason, conductivity measurements or
impedance measurements can be used to track the structural degradation of food.
The so-called cell disintegration index (CDI) is based on an impedance measurement
[4, 5]. Structural degradation in meat and fish can occur by post-mortem autolytic
processes and in plant foods by enzymatic activity or by mechanical crushing
processes. Figure 9.6 shows the difference of intact and crushed material using
banana as example. Thermal processes such as freezing or boiling and steaming
also lead to structural degradation. On the other hand, a structure can be generated,
e.g., by dispersing processes or the thermal denaturation of proteins. Structured
foods can show an anisotropic electrical conductivity.
Biological materials with a cellular structure have not only electrical conductivity
but also an electrical capacitance. This leads to a frequency dependence of the
conductivity and the electrical resistance. Figure 9.7 shows that the electrical
capacitance of intact potato tissue and destroyed potato tissue differs greatly at a
frequency of 1 kHz, but no difference can be seen at 10 MHz. To understand this, let
us look at AC resistors now.
416 9 Electrical Properties

9.2 Capacitance and Inductance

Foods have an electrical capacitance that depends on its composition. The electrical
capacitance of a material is a consequence of the electrical polarizability of its
components. In electrical engineering, this property of a material is indicated by
the so-called permittivity ε. The permittivity of the vacuum is 1, that of air is ≈1, that
of water at room temperature is about 80, and that of ice at -20 °C is 16. The
polarizable components of a material mainly are dipolar molecules and dissolved
ions with their counter-ion clouds. However, in an electric field also nonpolar
molecules can be polarized temporarily. We will see below that the frequency of
the alternating electric field plays a crucial role in the polarization behavior of the
material.
The capacitance of a simple plate capacitor is

A
C = ε
ε0
ð9:16Þ
d

C capacitance in F
ε permittivity
ε0 electric field constant (8.854
10-12 C
V-1
m-1)
A area in m2
d distance in m

From Eq. (9.16) it can be seen that a capacitor filled with a material with the
permittivity ε has a capacitance ε-times higher than the vacuum-filled or air-filled
capacitor. This property is used in numerous sensors, such as sensors for distance,
level, and humidity; see Chap. 16.
In an electric field, the polarizable components of a material align themselves
according to the electric attraction and repulsive forces. In an alternating field, this
alignment takes place periodically in the rhythm of the frequency f. Then electrical
charges in the material move back and forth with the frequency of the alternating
field; i.e., an alternating electric current is flowing in the material. This alternating
current increases with increasing frequency; i.e., the resistance of the material seems
to decrease. This AC resistance is called the capacitive reactance XC of a material. It
can be calculated by

1
XC = ð9:17Þ
ω
C
or

1
XC = ð9:18Þ
2π
f
C
9.2 Capacitance and Inductance 417

XC capacitive reactance in Ω
XL inductive reactance in Ω
ω angular frequency in s-1
f frequency in Hz
C capacitance in F (Farad)
L inductance in H (Henry)

Example
What is the capacitive reactance of a capacitor with a capacitance of 3 μF at
50 Hz and at 10 kHz?
Solution:with

1 1
XC = =
ω
L 2π
f
C
we get at 50 Hz

1 1
XC = =
2π
50 Hz
3 μF 2π
50 s - 1
3
10 - 6 A
s
V - 1
V
XC = = 1061 Ω
2π
1:5
10 - 4 A
and at 10 kHz

1 1
XC = =
2π
10 kHz
3 μF 2π
10 s - 1
3
10 - 3 A
s
V - 1

V 102 V
XC = = = 5:4 Ω
2π
3
10 -2
A 2π
3 A

The inductance of a material leads to an opposite behavior of the AC resistance:

The inductive reactance increases with increasing frequency XL. It is

XL = ω
L ð9:19Þ

or

X L = 2π
f
L ð9:20Þ
418 9 Electrical Properties

Example
What is the inductive reactance of an object with an inductance of 3 mH at
50 Hz and at 10 kHz?
Solution:with

X L = ω
L = 2π
f
L

at 50 Hz

X L = 2π
50 Hz
3 mH

X L = 2π
50 s - 1
3
10 - 3 kg
m2
s - 2
A - 2
X L = 2π
0:150 kg
m2
s - 3
A - 2 = 1Ω

and at 10 kHz

X L = 2π
10 kHz
3 mH

X L = 2π
10
10 - 3 s - 1
3
10 - 3 kg
m2
s - 2
A - 2
X L = 2π
30 kg
m2
s - 3
A - 2 = 189Ω

Conversion of units:
1 kg
m2
s-3
A-2 = 1 N
m
s-1
A-2 = 1 J
s-1
A-2 = 1 W
A-
2
= 1 V
A
A-2 = 1 V
A-1 = 1 Ω

The cause of inductive reactance is the storage of energy by a magnetic field in the
material. In electrical engineering, this effect plays a major role; however, the
inductance of biological objects such as agricultural products and food is very
low, so their inductive resistance is considered negligible (XL ≈ 0).

9.3 Impedance and Admittance

The resistance of a material occurring in an alternating electric field is called

impedance. The impedance is composed of the ohmic resistance and the reactance.
impedance = ohmic resistance þ reactance ð9:21Þ

or in short
impedance = resistance þ reactance ð9:22Þ

The reactance is composed of two parts: capacitive reactance and inductive

reactance.
9.3 Impedance and Admittance 419

For easier mathematical handling, complex quantities are used for AC resistors.
Then the impedance is the complex resistance Z, which is composed of its real part
XR (resistance) and its imaginary part XB (reactance).
Z = XR þ i
XB ð9:23Þ

The reactance can be calculated by

1
XB = ω
L - ð9:24Þ
ω
C
so we can then write for impendance
1
Z  = X R þ i
ωL - ð9:25Þ
ωC
Here, i is the imaginary unit (see Appendix), which we have already used in
Chap. 4 (Rheology) and Chap. 8 (Thermal Properties).
From the ratio of imaginary part and real part, the phase shift φ between voltage
and current in the AC circuit can be calculated. It is
XB
tan φ = ð9:26Þ
XR
The absolute value of a complex quantity is the geometric sum of the real part and
the imaginary part, so the absolute value of the complex impedance is the geometric
sum of XR (resistance) and XB (reactance).

jZ  j = X 2R þ X 2B ð9:27Þ

In DC circuits, the capacitive and the inductive reactance are zero. Therefore,
impedance and ohmic resistance are identical in DC case. When conductivity
measurements on food are carried out with AC voltages (e.g., at 60 kHz) these
experimentally measured properties are the impedance of the sample.

Definition
The geometric sum is the square root of the sum of the squares of the
individual summands. In trigonometry, it is called Pythagorean sum.

n
c= ð9:28Þ
n
x2j
j=1

The reciprocal size of the impedance is called admittance. Admittance is thus the
complex conductance with unit S (Siemens). Admittance is composed by conduc-
tance + susceptance. The terms are listed in Table 9.1.
420 9 Electrical Properties

9.3.1 Frequency Dependance of Impedance

Let us imagine a biological material in an alternating electric field. The material has
an intact cellular structure. A simple model for the cell that we use here is an
electrically non-conductive cell wall that separates the intracellular electrolyte fluid
from the extracellular electrolyte liquid. In an alternating field, charged particles,
permanent and temporary dipoles, ions and ion clouds, etc. follow the alternating
electric field and thus contribute to the AC conductivity. With an increase in
frequency, it is observed that gradually the polarizability and thus the permittivity
decrease. This frequency dependence is called the dispersion of permittivity. A
closer analysis of this dispersion reveals that the various components of the
biological material (cell walls, ions, macromolecules, etc.) can no longer periodically
follow the alternating field above certain frequencies. With a further increase in
frequency, their contribution to polarizability and thus to alternating current conduc-
tivity tends to zero. This drop in impedance at 0.1 kHz is called α-dispersion, at
0.1 MHz β-dispersion and in the GHz range γ-dispersion [7]. The α-transition is
attributed to the lateral movement of ions and the β-transition to the electrical
charging and discharging of cell membranes [7]. Frequencies below β-dispersion
therefore can be used to detect intact cells by means of impedance measurement. The
principle is to compare the impedances at low frequency, e.g., at 5 kHz (detects the
intact cells) and at high frequency, e.g., 5 MHz (AC resistance without the contribu-
tion of the intact cells). In the case of an electrolyte solution without cells and a
homogenate with destroyed cells, these two values hardly differ. In a solution with
intact cells, however, they differ strongly (see Fig. 9.7). In this way, the cell growth
of yeast cells can be monitored by impedance measurements [8] or bacteria can be
detected [9]. The loss of quality of meat as a result of cell damage can be detected
with an impedance-based index [10] and the efficacy of electric cell perforation
methods can be quantified [6, 11].

→ Attention
The term dispersion is also used in another context, namely as a fine distribu-
tion of a phase in another phase, so-called disperse systems, cf. Chap. 3.

9.4 Measurement of Electrical Conductivity and Impedance

The determination of electrical conductivity is called conductometry. Immersion

cells are used for the conductometric testing of solutions. A distinction is made
between inductive sensors and conductive sensors. In the inductive sensor, there are
two coils (transmitter and receiver). The transmitter coil generates a permanent
alternating electromagnetic field. The receiver coil receives this field and generates
an induced voltage signal which depends on the electrical properties of the solution.
The advantage of the inductive sensors is that both coils are embedded in a
non-conductive plastic housing and thus have no contact with the solution. Inductive
9.4 Measurement of Electrical Conductivity and Impedance 421

sensors are robust, but do not have the same sensitivity as conductive sensors. A
conductive measuring cell normally consists of two electrodes which, depending on
the application, are plate-like, cylindrical, or similar in shape and are operated with
an alternating voltage of low frequency (<100 Hz to 100 kHz). The measurement
itself is a measurement of voltage and current (AC resistance and impedance). A
temperature sensor is integrated in the measuring cell.
To calculate the electrical conductivity κ from the measured AC resistance R, the
cell geometry is required, which is contained in the cell constant k.

1 1 A
κ= =
ð9:29Þ
ρ R l
with

l
k= ð9:30Þ
A
it is

1 1
κ=
ð9:31Þ
R k
The cell constant k is determined by calibration. For this purpose, the measuring
cell is immersed in standard liquids of different electrical conductivities and the cell
constant is taken from the slope of the κ - R-curve. Salt solutions such as KCl
solutions can be used as calibration substances (refer Table 9.7).
As mentioned previously, during all these measurements, temperature must be
precisely measured, controlled, recorded, and reported. Most laboratory commercial
instruments for the measurement of electrical conductivity are equipped with a
means for automatically adjusting or correcting for different temperatures as a
normal part of the instrument hardware. An example of this is the internal use of
the temperature coefficient in Eq. (9.14). The capability for such automatic correc-
tion for temperature is a very useful feature. For example, a measurement of
electrical conductivity taken at a temperature of 21 °C can be corrected to report
what the conductivity would be at 25 °C. Of course, the instrument must be
programmed with the correct value of temperature coefficient to use. Most
instruments allow the operator to enter this value as an input for operation of the
instrument. For most dissolved salt solutions, a temperature coefficient of
k = 0.022 K-1 is used (refer Table 9.8). Before the measurement, it should be
checked that this coefficient is valid for the system to be examined.

Table 9.7 Conductivities κ (S
m-1)

of KCl solutions
c (mol
dm-3) 18 °C 20 °C 25 °C
0.1 1.119 1.167 1.288
0.01 0.1225 0.1278 0.1413
422 9 Electrical Properties

Table 9.8 Examples of k (K-1)

temperature coefficients for
Strong acids 0.016
equivalent electric conduc-
tivity at room temperature Strong bases (chemical base) 0.019
[3] Salts 0.022
Water 0.058

Table 9.9 Values of elec- κ ‐ range

tric conductivities
Pure water 0.1 μ S
cm-1
Demineralised water 0.1 – 10 μ S
cm-1
Tap water (drinking water) 100 – 1000 μ S
cm-1
Wastewater 1–10 mS
cm-1
Seawater 1–100 mS
cm-1
Aqueous salt solutions 10–500 mS
cm-1
Concentrated acids and bases 100–1000 mS
cm-1

In order to minimize the undesirable interference caused by the cataphoretic and

electrophoretic effects that develop in a static electric field, conductivity
measurements are usually carried out in the presence of a dynamic oscillating
field, in which the voltage potential reverses periodically at a regular frequency.
This also helps prolong the lifetime of the electrodes. The response in an electric
circuit to such a periodic reversal in voltage is an alternating current that changes
direction periodically in response to the voltage. Likewise, the ions in an electrolyte
solution will attempt to move in such an oscillating way. The resistance to this type
of oscillating movement in the presence of a periodic alternating electric field is
called impedance Z. In order to keep the impedance of the measuring cell itself low,
it should have a low inductive and a high capacitive reactance. While the inductance
of two plates is negligible, a high capacitance can be achieved by increasing the
surface of the electrodes by electrochemical deposition of platinum. One drawback is
that these types of electrodes are sensitive to mechanical abrasion.
Table 9.9 gives an overview of the magnitude of the electrical conductivity of
water and aqueous solutions under different conditions.
The tracking of electrical conductivity is one method often used for indirect
determination of the oxidation stability index (OSI) of fats and oils. In order to
shorten the test duration, the oxidation stability is determined at elevated tempera-
ture, e.g., at 100 °C or 130 °C. Air is allowed to bubble through the fat at this
temperature and the escaping air is directed into demineralized water. Volatile
oxidation products such as short-chain fatty acids, which are discharged from the
fat with the air flow, dissolve in the water, dissociate there, and cause an increase in
the electrical conductivity of the water. The time after which a characteristic increase
in conductivity is observed is called induction time. Such studies can be used, for
example, to assess the effectiveness of antioxidants [12].
9.5 Zeta Potential 423

9.5 Zeta Potential

Particles in disperse systems are often electrically charged. Solid particles in

suspensions and liquid particles in emulsions often carry a negative electrical charge
at their interface. Systems with a positive charge occur less frequently [13]. In an
aqueous electrolyte solution, oppositely charged hydrated ions are attached to the
particle interface. According to Stern’s theory of the electric double layer [14], the
first layer with tightly bound molecules and thickness δ is called the Stern layer
[15]. In the second so-called diffuse layer, further hydrated ions are bound less
strongly.
In Fig. 9.8, the ions of the electric bilayer with their hydrate shells are shown
schematically. Due to the attachment of counter-charged ions, the electrical potential
φS of the surface decreases with increasing distance from the interface. Within the
Stern layer, the drop in potential is linear. This is followed by an exponentially
sloping curve (s. Fig. 9.8), which begins at Stern’s potential φS and goes asymptotic
in the direction of the electrical potential φL of the solution.
The length axis ζ (zeta, cf. Greek characters in the appendix) in Fig. 9.8 is a
transformed z-axis: ζ = z - δ, so the exponential part of the potential curve is
ζ
φðζ Þ = φS
e - β ð9:32Þ

The thickness of the diffuse layer β is defined by where the electric potential φ(ζ)
has fallen to the e-th part of its initial value φS. This thickness is identical to the
Debye length from Debye–Hückel’s theory [16]. The potential at this point is also
referred to as the Helmholtz potential φH [15].

Fig. 9.8 Electrical double

layer. Top: Thickness of the
Stern layer and the diffuse
layer. Bottom: Between
Stern’s potential φS and the
potential inside the solution
φL, the potential drops
exponentially
424 9 Electrical Properties

Definition
An electrical double layer consists of the Stern layer and diffuse layer. Its
thickness is determined by calculating the length by which the electrical
potential has fallen to the e-th part of the initial value.

The potential difference between the solution φL and the Helmholtz potential φH
is a measure of the agglomeration tendency of the particles. High values lead to
strong repulsion and thus to stable dispersions, while low values indicate
instabilities. Measurement of the zeta potential is carried out indirectly, e.g., via
the electrophoretic mobility of the particle, which depends on the thickness of the
diffuse layer. In an electric field E, the charged particles migrate at the speed v. The
quotient of v and E is called electrophoretic mobility and can be determined optically
or acoustically. It is assumed that at the movement of the charged particle, part of the
ion cloud is no longer taken along, but is sheared off. The place where this happens is
called the shear zone (slipping plane) or also called the outer Helmholtz layer. The
shear radius determined in this way provides the thickness of the diffuse layer β and
thus enables the calculation of the electrical potential φ(ζ) at this point, which means
φH = φ(ζ = β).
For the assessment of the stability of dispersions against flocculation and agglom-
eration, the potential difference between the shear zone φH and the solution φL is
used. The potential difference between the shear zone φH and the solution φL is used
for assessing the stability of dispersions against flocculation and agglomeration. This
value is referred to as “the zeta potential ZP of the dispersion” (Table 9.10):
This potential is:
ZP = φH - φL ð9:33Þ

ζ, z distance in m
β thickness of diffuse layer in m
δ thickness of the Stern layer in m
φH electrical potential at distance β in V
φS electrical potential at distance δ in V
φL electrical potential of solution in V
ZP zeta potential in V (=φH - φL)

Table 9.10 Dispersion ZP in mV

stability and zeta potential
0–5 Instable: agglomeration, flocculation
ZP, according to [17]
10–30 Beginning instability
30–40 Moderate stability
40–60 Good stability
>60 Very good stability
9.6 Plasma 425

Bottom line
Above a zeta potential ZP of about 30 mV, dispersions are stable against
flocculation and agglomeration.

→ Attention
A dispersion does not have a single zeta potential. Rather, there is a value of
the electrical potential φ(ζ) dependent on the distance ζ from the particle. The
difference of the potential in the shear zone φH and in the solution φL can be
used as a parameter for the stability of a dispersion and is often referred to as
“the zeta potential” in this context.

From a metrological point of view, methods of electrophoresis, electroosmosis,

dynamic laser diffraction, or sedimentation are suitable for determining the electrical
potential in the shear zone φH [18]. In any case, the calculation of the electrical
potential is based on the indirectly measured particle diameter (including the double
layer) using the Henry function and the Smoluchowski approximation [13]. For this
calculation, values of local viscosity and permittivity are required [15]. If these are
known only at approximate values, then the calculated value of the potential φH and
thus ZP is also an approximation.

Remark
The ion layer, which is strongly bound at an interface, was first described by
Helmholtz and since has been referred to as the Helmholtz layer. In the Gouy–
Chapman theory, the terms “inner Helmholtz layer” (for the Stern layer) and
“outer Helmholtz layer” (for the diffuse layer) are then used. Stern’s theory of
the electric double layer combines both layers. Since then, the firmly bound
ion layer has also been referred to as the Stern layer.

9.6 Plasma

A plasma is a mixture of ions, free electrons, radicals, and usually also neutral atoms
or molecules. It is an electrically conductive gaseous state and sometimes referred to
as a fourth state of matter: all that lies between the celestial bodies in space, as well as
within the Sun and other stars, which are in plasma state. There are plasmas in bolts
of lightning, flames, and gas discharge from fluorescence tube lamps. A plasma can
be characterized by the degree of ionization, as well as concentration and energy of
the (different) charge carriers.
426 9 Electrical Properties

A cold plasma is a gas with a low degree of ionization which can be produced at
atmospheric pressure and ambient temperature. If this cold plasma comes into
contact with a surface, reactions occur instantly in which reactive species such as
hydroxyl radicals, ozone, and NO are formed and biological structures such as
biofilms, bacteria, spores, and viruses inactivated. Due to the high reactivity of the
plasma, its penetration depth is low. In technical applications, therefore, focus is on
the treatment of surfaces, e.g., disinfection of surfaces of food packaging by cold
plasmas [19–21]. In biomedical engineering, cold plasmas are used for local disin-
fection and wound healing [22–24]. The surface treatment of food and applications
like non-thermal processes are the subject of numerous investigations [19, 20, 25–
32]; also, the inactivation of viruses in ambient air is discussed [33, 34].

Definition
In physics, a plasma is a gaseous, electrically conductive mixture of ions,
electrons, radicals, atoms, and molecules. Another definition of plasma that is
unrelated to physics or states of matter is used in medicine, and is defined as
the transparent, slightly yellowish liquid that makes up about 55% of the
blood.

9.7 Applications

An application in food technology based on the electrical properties of food

materials is conductive heating of food, also known as ohmic heating. Here, an
electric current is passed through the food, where heat is generated due to the ohmic
resistance. Important material properties here are the AC resistance and the permit-
tivity of the food or the individual components of a food. There is no need to use hot
surfaces for heat transfer, which eliminates heat transfer problems. Heating metal
surfaces by electromagnetic induction (induction cooker) is another way of
simplifying heat transfer. Another area is the electro-perforation of cells. For this
purpose, high-voltage pulses are passed through the food, which causes perforation
of cell walls [35]. This can have positive effects for preservation and subsequent
processes such as crushing or extraction. Plasma treatment of surfaces is used as a
hygienic measure.

Further Reading

Standard for measuring the electrical conductivity of water [36]

Conductive heating of food [37–45]
Electrical properties for Ohmic heating of potatoes [46]
Cell perforation by pulsed electric fields [47–49]
Potato: application with pulsed electric fields [50]

(continued)
References 427

Determination of cell perforation [51]

Agricultural products: electrical and rheological investigation of freeze damage [52]
Cereals: conductometric determination of water content [12]
Honey: water content by means of conductivity measurement [53]
Milk: modelling of conductivity [54]
Microbiological growth: monitoring via impedance [55, 56]
Zeta potential of Saccharomyces cerevisiae during fermentation [57]
Zeta potential of algae [58, 59]
Pimento: disinfection by cold plasma [28]
Plasma effect on the emulsifying properties of proteins [27]
Virus inactivation by means of cold plasma [33]
Plasma application in food packaging [29]
Fat oxidation by plasma application [26]
Plasma application in the dairy industry [31]
Plasma application in the meat industry [32]
Zeta potential changing nanoemulsions [60]

Summary
The electrical conductivity of food, its terminology, and causes are presented.
The main differences between DC and AC fields are demonstrated. The
resistances of AC and their mathematical handling by complex properties
are explained in simple terms. The measurement of electrical conductivity is
presented as well as the basics of the zeta potential and cold plasma. At the end
of the chapter, application examples are listed, which can be used for further
studies and as suggestions for your own scientific work.

References
1. Kohlrausch F (1996) Praktische Physik Bd.3. Teubner, Stuttgart. https://doi.org/10.1002/piuz.
19870180109
2. Reitler W (1990) Konduktive Erwärmung von Lebensmitteln. Dissertation, Technical Univer-
sity Munich
3. Atkins PW, De Paula J (2011) Physical chemistry for the life sciences, 2nd edn. W.H. Freeman,
Oxford
4. Angersbach A, Heinz V, Knorr D (1997) Elektrische Leitfähigkeit als Mass des
Zellaufschlussgrades von zellulären Materialien durch Verarbeitungsprozesse. LVT 42:195
5. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, München
6. Angersbach A, Heinz V, Knorr D (1999) Electrophysiological model of intact and processed
plant tissues: cell disintegration criteria. Biotechnol Prog 15(4):753. https://doi.org/10.1021/
bp990079f
428 9 Electrical Properties

7. Pliquett U (2010) Bioimpedance: a review for food processing. Food Eng Rev 2(2):74. https://
doi.org/10.1007/s12393-010-9019-z
8. Soley A, Lecina M, Gámez X, Cairó JJ, Riu P, Rosell X, Bragós R, Gòdia F (2005) On-line
monitoring of yeast cell growth by impedance spectroscopy. J Biotechnol 118(4):398. https://
doi.org/10.1016/j.jbiotec.2005.05.022
9. Furst AL, Francis MB (2019) Impedance-based detection of bacteria. Chem Rev 119(1):700.
https://doi.org/10.1021/acs.chemrev.8b00381
10. Pliquett U, Altmann M, Pliquett F, Schˆberlein L (2003) Py–a parameter for meat quality. Meat
Sci 65(4):1429. https://doi.org/10.1016/S0309-1740(03)00066-4
11. Toepfl S, Siemer C, Saldaña-Navarro G, Heinz V (2014) In: Sun D-W (ed) Overview of pulsed
electric fields processing for food. Academic Press, San Diego, pp 93–114. https://doi.org/10.
1016/b978-0-12-411479-1.00006-1
12. Nielsen S (2017) Food analysis. Springer, New York
13. Cano-Sarmiento C, Téllez-Medina DI, Viveros-Contreras R, Cornejo-Mazón M, Figueroa-
Hernández CY, García-Armenta E, Alamilla-Beltrán L, García HS, Gutiérrez-López GF
(2018) Zeta potential of food matrices. Food Eng Rev 10(3):113. https://doi.org/10.1007/
s12393-018-9176-z
14. Stern O (1924) Zur Theorie der elektrolytischen Doppelschicht. Zeitschrift für Elektrochemie
30(21–22):508
15. Brezesinski G, Mögel HJ (1993) Grenzflächen und Kolloide. Spektrum Akademischer Verlag,
Berlin
16. Walstra P (2002) Physical chemistry of foods. Marcel Dekker, New York
17. O’Brien RW, Midmore BR, Lamb A, Hunter RJ (1990) Electroacoustic studies of moderately
concentrated colloidal suspensions. Faraday Disc the Chem Soc 90:301. https://doi.org/10.
1039/DC9909000301
18. Mahl D (2011) Synthese, Löslichkeit und Stabilität von Gold- Nanopartikeln in biologischen
Medien. Universität Duisburg-Essen
19. Mandal R, Singh A, Pratap Singh A (2018) Recent developments in cold plasma decontamina-
tion technology in the food industry. 80. https://doi.org/10.1016/j.tifs.2018.07.014
20. Pankaj SK, Keener KM (2017) Cold plasma applications in food packaging. Elsevier. https://
doi.org/10.1016/b978-0-08-100596-5.21417-0
21. Scholtz V, Pazlarova J, Souskova H, Khun J, Julak J (2015) Nonthermal plasma – a tool for
decontamination and disinfection., vol 33. https://doi.org/10.1016/j.biotechadv.2015.01.002
22. Ojha S, Fröhling A, Durek J, Ehlbeck J, Tiwari BK, Schlüter OK, Bußler S (2021) Principles
and application of cold plasma in food processing. In: Knoerzer K, Muthukumarappan K (eds)
Innovative food processing technologies. Elsevier, Oxford, pp 519–540. https://doi.org/10.
1016/B978-0-08-100596-5.23033-3
23. Busco G, Robert E, Chettouh-Hammas N, Pouvesle J-M, Grillon C (2020) The emerging
potential of cold atmospheric plasma in skin biology. Free Radical Biol Med 161:290. https://
doi.org/10.1016/j.freeradbiomed.2020.10.004
24. Kang W-S, Hong Y-C, Hong Y-B, Kim J-H, Uhm HS (2010) Atmospheric-pressure cold
plasma jet for medical applications. Surf Coating Technol 205:S418. https://doi.org/10.1016/
j.surfcoat.2010.08.138
25. Bermudez-Aguirre D (2020) Chapter 2 – Advances in the inactivation of microorganisms and
viruses in food and model systems using cold plasma. In: Bermudez-Aguirre D (ed) Advances
in cold plasma applications for food safety and preservation. Academic Press, pp 49–91. https://
doi.org/10.1016/B978-0-12-814921-8.00002-5
26. Gavahian M, Chu YH, Khaneghah AM, Barba FJ, Misra NN (2018) A critical analysis of the
cold plasma induced lipid oxidation in foods. Trends Food Sci Technol 77:32. https://doi.org/
10.1016/j.tifs.2018.04.009
27. Jiang Y-H, Cheng J-H, Sun D-W (2020) Effects of plasma chemistry on the interfacial
performance of protein and polysaccharide in emulsion. Trends Food Sci Technol 98:129.
https://doi.org/10.1016/j.tifs.2020.02.009
References 429

28. Medvecká V, Mošovská S, Mikulajová A, Valík Ľ, Zahoranová A (2020) Cold atmospheric

pressure plasma decontamination of allspice berries and effect on qualitative characteristics. Eur
Food Res Technol 246(11):2215. https://doi.org/10.1007/s00217-020-03566-0
29. Misra NN, Yepez X, Xu L, Keener K (2019) In-package cold plasma technologies. J Food Eng
244:21. https://doi.org/10.1016/j.jfoodeng.2018.09.019
30. Hati S, Patel M, Yadav D (2018) Food bioprocessing by non-thermal plasma technology. Curr
Opin Food Sci 19:85. https://doi.org/10.1016/j.cofs.2018.03.011
31. Coutinho NM, Silveira MR, Rocha RS, Moraes J, Ferreira MVS, Pimentel TC, Freitas MQ,
Silva MC, Raices RSL, Ranadheera CS, Borges FO, Mathias SP, Fernandes FAN, Rodrigues S,
Cruz AG (2018) Cold plasma processing of milk and dairy products. Trends Food Sci Technol
74:56. https://doi.org/10.1016/j.tifs.2018.02.008
32. Misra NN, Jo C (2017) Applications of cold plasma technology for microbiological safety in
meat industry. Trends Food Sci Technol 64:74. https://doi.org/10.1016/j.tifs.2017.04.005
33. Filipić A, Gutierrez-Aguirre I, Primc G, Mozetič M, Dobnik D (2020) Cold plasma, a new hope
in the field of virus inactivation. Trends Biotechnol 38(11):1278. https://doi.org/10.1016/j.
tibtech.2020.04.003
34. Daeschlein G (2016) Antimikrobielle Wirksamkeit von Plasma. In: Metelmann H-R, von
Woedtke T, Weltmann K-D (eds) Plasmamedizin: Kaltplasma in der medizinischen
Anwendung. Springer, Berlin, pp 51–62. https://doi.org/10.1007/978-3-662-52645-3_4
35. Raso J, Heinz V, Alvarez I, Toepfl S (eds) (2022) Pulsed electric fields technology for the food
industry. Springer, Cham. https://doi.org/10.1007/978-3-030-70586-2
36. DIN EN 27888 (1993) Wasserbeschaffenheit; Bestimmung der elektrischen Leitfähigkeit.
Beuth, Berlin. https://doi.org/10.31030/3123472
37. Martín-Belloso O, Morales-de la Peña M (2018) Fruit preservation by ohmic heating and pulsed
electric fields. In: Rosenthal A, Deliza R, Welti-Chanes J, Barbosa-Cánovas GV (eds) Fruit
preservation: novel and conventional technologies. Springer, New York, NY, pp 441–456.
https://doi.org/10.1007/978-1-4939-3311-2_16
38. Ramaswamy H, Marcotte M, Sastry S, Abdelrahim K (eds) (2014) Ohmic heating in food
processing. CRC Press, Boca Raton. https://doi.org/10.1201/b16605
39. Fellows PJ (2000) Food processing technology – principles and practice. CRC Press, Boca
Raton
40. DFG (2015) Opinion on the use of ohmic heating for the treatment of foods. Senate Commis-
sion on Food Safety, Bonn
41. Knirsch MC, dos Santos CA, Vicente AAMDS, Pennaa TCV (2010) Ohmic heating – a review.
Trends Food Sci Technol 21(9):436. https://doi.org/10.1016/j.tifs.2010.06.003
42. Muhammad A, Tadda MA, Shitu A (2019) Ohmic heating as alternative preservation
technique – a review. Arid Zone J Eng Technol Environ 15:268
43. Sarkis JR, Mercali GD, Tessaro IC, Marczak LDF (2013) Evaluation of key parameters during
construction and operation of an ohmic heating apparatus. Innovative Food Sci Emerg Technol
18:145. https://doi.org/10.1016/j.ifset.2013.02.001
44. Fadavi A, Salari S (2019) Ohmic heating of lemon and grapefruit juices under vacuum
pressure – comparison of electrical conductivity and heating rate. J Food Sci 84(10):2868.
https://doi.org/10.1111/1750-3841.14802
45. Salari S, Jafari SM (2020) The influence of ohmic heating on degradation of food bioactive
ingredients. Food Eng Rev 12(2):191. https://doi.org/10.1007/s12393-020-09217-0
46. Cho W-I, Song S-H (2019) Electrical properties for the optimal ohmic heating of vegetable
foodstuffs. Food Eng Prog 23(3):155. https://doi.org/10.13050/foodengprog.2019.23.3.155
47. Heinz V, Raso JV (2006) Pulsed electric fields technology for the food industry. Springer,
New York, NY. https://doi.org/10.1007/978-0-387-31122-7
48. Nowosad K, Sujka M, Pankiewicz U, Kowalski R (2020) The application of PEF technology in
food processing and human nutrition. J Food Sci Technol. https://doi.org/10.1007/s13197-020-
04512-4
430 9 Electrical Properties

49. Dellarosa N, Laghi L, Ragni L, Dalla Rosa M, Galante A, Ranieri B, Florio TM, Alecci M
(2018) Pulsed electric fields processing of apple tissue: spatial distribution of electroporation by
means of magnetic resonance imaging and computer vision system. Innovative Food Sci Emerg
Technol 47:120. https://doi.org/10.1016/j.ifset.2018.02.010
50. Fauster T, Ostermeier R, Scheibelberger R, Jäger H (2021) Pulsed Electric Field (PEF)
application in the potato industry. In: Knoerzer K, Muthukumarappan K (eds) Innovative
food processing technologies. Elsevier, Oxford, pp 253–270. https://doi.org/10.1016/B978-0-
12-815781-7.00013-5
51. Vorobiev E, Lebovka N (2020) Techniques to detect electroporation. In: Vorobiev E, Lebovka
N (eds) Processing of foods and biomass feedstocks by pulsed electric energy. Springer
International, Cham, pp 51–84. https://doi.org/10.1007/978-3-030-40917-3_3
52. Ohnishi S, Fujii T, Miyawaki O (2002) Electrical and rheological analysis of freezing injury of
agricultural products. Int J Food Prop 5(2):317. https://doi.org/10.1081/JFP-120005788
53. BVL L 40.00-5 (2003) Untersuchung von Lebensmitteln - Untersuchung von Honig -
Bestimmung der elektrischen Leitfähigkeit. Bundesamt für Verbraucherschutz und
Lebensmittelsicherheit, Berlin
54. Therdthai N, Zhou W (2002) Hybrid neural modeling oftthe electrical conductivity property of
recombined milk. Int J Food Prop 5(1):49. https://doi.org/10.1081/JFP-120015593
55. Bolton FJ (1991) Conductance and impedance methods for detecting pathogens. In: Vaheri A,
Tilton RC, Balows A (eds) Rapid methods and automation in microbiology and immunology.
Springer, Heidelberg. https://doi.org/10.1007/978-3-642-76603-9_21
56. Yang L, Bashir R (2008) Electrical/electrochemical impedance for rapid detection of foodborne
pathogenic bacteria. Biotechnol Adv 26(2):135. https://doi.org/10.1016/j.biotechadv.2007.
10.003
57. Lavaisse LM, Hollmann A, Nazareno MA, Disalvo EA (2019) Zeta potential changes of
Saccharomyces cerevisiae during fermentative and respiratory cycles. Colloids Surf B
Biointerfaces 174:63. https://doi.org/10.1016/j.colsurfb.2018.11.001
58. Pandey A, Pathak VV, Kothari R, Black PN, Tyagi VV (2019) Experimental studies on zeta
potential of flocculants for harvesting of algae. J Environ Manag 231:562. https://doi.org/10.
1016/j.jenvman.2018.09.096
59. Ahmad S, Kothari R, Singh HM, Tyagi VV, Singh B, Sari A (2021) Experimental investigation
of microalgal harvesting with low cost bottom ash: Influence of temperature and pH with zeta
potential and thermodynamic function. Environ Technol Innov:101376. https://doi.org/10.
1016/j.eti.2021.101376
60. Sharifi F, Jahangiri M, Nazir I, Asim MH, Ebrahimnejad P, Hupfauf A, Gust R, Bernkop-
Schnürch A (2021) Zeta potential changing nanoemulsions based on a simple zwitterion. J
Colloid Interf Sci 585:126. https://doi.org/10.1016/j.jcis.2020.11.054
Magnetic Properties
10

Magnetic properties play a role in many food and packaging materials in the field of
food technology, as well as in sensors and nuclear resonance spectroscopy. Which
substances can be detected with metal detectors and which materials are suitable for
induction cookers depends on their magnetic and electrical properties. Magnetic
fields also have effects on biological materials. In this chapter, the effects of
magnetic fields on free radicals, ions, hydrogen bonds, function of cell membranes,
and activity of enzymes are discussed [1–3]. In the literature, however, there are
contradictory findings on the effect of magnetic fields on components of biological
materials. Often necessary details are missing with regard to the field strength of the
applied fields or field homogeneity, duration of exposure, and frequency of the
periodic or pulse-like applications [4].
We begin our studies in this chapter with a static, homogeneous magnetic field: If
a material is brought into such a magnetic field, it is magnetically polarized;
colloquially it is said to be magnetized. What does this mean? In the simple model
of elementary magnets, we imagine that submicroscopic components of the material,
so-called elementary magnets, align themselves through the external magnetic field.
The different ability of the substances to show such a magnetic polarization is called
magnetic permeability. Substances with high magnetic permeability show strong
magnetization in a given magnetic field, while substances with low permeability
show low magnetization. If we look on the causes of magnetic polarization, a
distinction can be made between materials that contain permanent magnetic dipoles
at the atomic level (paramagnetic materials, paramagnetism) and materials that do
not contain these permanent magnetic dipoles (diamagnetic materials, diamagne-
tism). Paramagnetism arises from the atoms of a material having magnetic moments
(orbital moments and/or spin moments). Often the influence of unpaired electron
spins predominates. The atoms of diamagnetic materials, on the other hand, have an
electron configuration that does not result in a magnetic moment, e.g., fully paired
electron spins. The macroscopic magnetic moment of a solid results from the
vectorial sum of all atomic magnetic moments and is called magnetization.

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 431
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_10
432 10 Magnetic Properties

10.1 Magnetic Material Properties

10.1.1 Paramagnetism

An example of a paramagnetic material is aluminum. The electron configuration of

aluminum is 1s22s22p63s23p1 and is shown in Fig. 10.1. The aluminum atom has an
unpaired electron, which gives it permanent magnetic momentum. This type of
paramagetism is called Langevin paramagnetism and is most often the reason for
materials being paramagnetic. Other contributions to the paramagnetic behavior can
be given by excited conduction electrons in metals (Pauli paramagnetism) or by
atoms out of their normal energy level (van-Vleck paramagnetism).
The addition of different contributions to atomic angular momentum can affect
the paramagnetic behavior of a material. For simplification we can consider matter
like this, consisting of tiny elemental magnets which can be oriented in response to
an external magnetic field (see Fig. 10.2).
Recall that when a paramagnetic material is brought into a magnetic field, the
polarized atoms will orient themselves to become aligned with the polarity of the
magnetic field. In so doing, the strength of the magnetic field (flux) is amplified by this
aligned orientation. This is called magnetic field amplification by a paramagnetic material.
In paramagnetic materials, the ability of the atoms to orient themselves in
response to an external magnetic field diminishes with increasing temperature,
until a temperature is reached at which the atoms no longer respond, and paramag-
netism is lost. This is known as the Curie effect, and the critical temperature at which
magnetism ceases is known as the Curie temperature. Applications of the Curie
effect and to thermal analysis were discussed in Chap. 8.

1s 2s 2p 3s 3p

K- L- M-shell

Fig. 10.1 Electron configuration of aluminum, schematic

Fig. 10.2 Aluminum without outer magnetic field (left). Aluminum in an outer magnetic field is
magnetically polarized (right)
10.1 Magnetic Material Properties 433

10.1.2 Ferromagnetism

Ferrous materials like iron (Lat. ferrum) contain microscopic regions in which the
elemental polarized atoms are already aligned in a fixed orientation. These are called
Weiss regions. However, each Weiss region may have all its atoms oriented in a
direction that is different from that in other regions, such that the direction of
orientation is randomly distributed among all the Weiss regions throughout the
ferrous material. When such a material is brought into a magnetic field, those
Weiss regions with atoms already oriented in the correct direction in alignment
with the polarity of the magnetic field will grow, while the others will diminish and
ultimately disappear.
For this reason, ferrous materials develop a very strong magnetic polarization in
response to an external magnetic field, which is much stronger than paramagnetic
materials. This behavior is called ferromagnetism.
However, just as with paramagnetic materials, this ferromagnetic ability to
develop strong magnetic polarity in response to an external magnetic field
diminishes with increasing temperature until a critical temperature is reached at
which the magnetism is lost. Thus, ferromagnetic materials experience the same
Curie effect and Curie temperature discussed earlier with paramagnetism.
Iron, ferritic steels, cobalt, and nickel are examples of ferromagnetic materials.
Austenitic steels and thus many stainless steels used in food processing are not
ferromagnetic, but paramagnetic, although iron is their main component. This shows
that not only the composition but the lattice structure of a material is decisive for its
magnetic properties. We have already become acquainted with a laboratory applica-
tion of the temperature dependence of permanent magnetism in the section on
thermogravimetry (Chap. 8). Here, the Curie transition is used to simulate a weight
force in the temperature calibration of a thermal balance.

10.1.3 Diamagnetism

Recall that diamagnetic materials are made up of atoms with paired electron spins
and have no magnetic momentum. These types of materials will develop no mag-
netic polarization when brought into a magnetic field and are totally unresponsive to
an external magnetic field. In addition to being unresponsive to a magnetic field,
these materials have also been found to have a slight adverse effect on the magnetic
field by weakening the field near the outer regions of the diamagnetic material. This
effect can be explained by induction. When such materials are brought into the
presence of a magnetic field, atomic currents are induced near the surface of the
material. According to Lenz’s law, these induced atomic currents will flow in a
direction which causes a weakening of the intruding field.
This is what we call the diamagnetic effect. However, this effect is not limited to
only diamagnetic materials. All materials possess this diamagnetic effect, including
ferromagnetic and paramagnetic materials. It is not observed in these other materials
434 10 Magnetic Properties

Table 10.1 Comparison of diamagnetic, paramagnetic, and ferromagnetic materials

Diamagnetic Paramagnetic Ferromagnetic
Permanent dipoles No Yes Yes, and readily
present ordered
Electron spins are Paired Unpaired Unpaired
Materials amplify field No Yes Yes, strongly
Materials have Yes Yes Yes
diamagnetic effect
Examples Lead, copper, water, Aluminum, Iron, cobalt,
nitrogen, hydrogen platinum, oxygen nickel

because it is a very slight effect that is overshadowed by the stronger magnetic

polarization of these other materials.
We can summarize up to this point by defining diamagnetic materials as those
which possess no paramagnetism or ferromagnetism (Table 10.1). We can also
conclude that the diametric effect occurs at the subatomic level in a material and
not at the molecular level. Therefore, it is immune to Brownian molecular motion
and does not depend on temperature.
Normally, the magnetic moments of non-paired electron spins are parallel to each
other. But some materials show anti-parallel orientation. When that happens, these
opposite magnetic moments compensate for each other and cancel each other out so
that no magnetic polarization develops. In this case, the material is apparently
diamagnetic. This is called anti-ferromagnetism. Mostly oxides of iron, manganese,
and chromium show this behavior.
Again, there exists a critical temperature, like the Curie temperature, at which this
anti-ferromagnetic behavior disappears and changes to paramagnetic behavior. This
transition temperature is called the Néel temperature.
When we have materials that show anti-parallelism in their electron spins, but the
electron spins are not of equal strength, then they do not completely compensate for
each other, and some residual magnetism is evident. This behavior is called ferri-
magnetism and the materials which exhibit it are called ferrites. Ferrites from
sintered oxides of iron, manganese, nickel, zinc, and cadmium have very low
electrical conductivity in contrast to metallic ferromagnetic substances.

10.1.3.1 Magnetic Refrigeration

A material that is magnetized in an external magnetic field has a lower entropy than
the same material without magnetization. If the external magnetic field is switched
off, the material returns to the non-magnetized state under energy absorption.
Hereby the entropy of the material increases. According to the second law of
thermodynamics (see Chap. 8), with adiabatic entropy increase the temperature
decreases. This means that with adiabatic demagnetization, the temperature of a
material decreases, and with adiabatic magnetization, the temperature rises.
This so-called magnetocaloric effect can be used to achieve low temperatures,
i.e., temperatures below 1 K. Copper or platinum can serve as a paramagnetic
cooling core. Pre-cooling is carried out with liquid helium, and coils made of
10.2 Magnetization 435

superconducting materials are used for magnetic field generation. In order to exclude
remanence in the vicinity of the cooling core, paramagnetic salts are used, e.g.,
FeNH4(SO4)2
12H2O.
Atomic nuclei can also have magnetic moments, which, however, are 1000 times
lower than the magnetic moments resulting from the momentum of unpaired
electrons. In adiabatic demagnetization of atomic nuclei, exactly these properties
of atomic nuclei are used (cf. Sect. 10.3 nuclear resonance) and thus temperatures in
the μK range can be reached [5]. By low-temperature physics, material properties
and effects can be studied at temperatures below the lowest temperatures occurring
in space.

10.2 Magnetization

Just like describing an electrical field, we need two quantities by which to describe a
→ →
magnetic field: the field strength H and the flux density B . Flux density is also
→
sometimes called magnetic induction B . The relationship between these is:
→ →
B = μ
μ0
H ð10:1Þ

B magnetic flux density in V
s
m-2

H magnetic field in A
m-1
μ0 magnetic field constant (μ0 = 4π
10-7V
s
A-1
m-1)
μ magnetic permeability
J magnetization in V
s
m-2

The flux density of a magnetic field within a given space will depend on the
nature of the material matter occupying the space and will be a minimum when no
matter is present (i.e., in a vacuum). When we measure the flux density in a vacuum
B0, and then with a sample material B, we can determine the magnetic permeability μ
of the sample material by taking the ratio of flux density with sample over flux
density in vacuum. When matter is present, Eq. (10.1) is

B = μ
μ0
H ð10:2Þ

without matter

B = μ0
H ð10:3Þ

Comparing gives:

B = μ
B0 ð10:4Þ
436 10 Magnetic Properties

and

B
μ= ð10:5Þ
B0
Therefore, magnetic permeability is a measure of field amplification ability of
materials. For a vacuum, magnetic permeability is 1.
The difference in flux density between a material and a vacuum B - B0 is called
the magnetization M of the material:
M = ΔB = B - B0 ð10:6Þ

with
M = μ
μ 0
H - μ0
H ð10:7Þ

the magnetization is

M = ðμ - 1Þμ0
H ð10:8Þ

with the magnetic susceptibility χ and

χ=μ-1 ð10:9Þ

the magnetization simply is

M = χ
μ0
H ð10:10Þ

or

M=χ
B ð10:11Þ

and

dM
χ= ð10:12Þ
dB
and analogous

dM
μ= ð10:13Þ
dH
The property χ is called magnetic susceptibility and is also a measure of the
magnetization of a material and its ability for field amplification. Table 10.2 shows
the range of values for magnetic permeability and susceptibility of materials with
different magnetisms. Accordingly, materials are diamagnetic if its χ is negative and
paramagnetic if its χ is positive. In Tables 10.3 and 10.4, some magnetic material
properties are shown by way of example.
10.2 Magnetization 437

Table 10.2 Magnetic permeability and susceptibility of materials with different magnetism
Diamagnetic material Paramagnetic material Ferromagnetic material
μ<1 μ>1 μ> >1
χ<0 χ>0 χ> >0

Table 10.3 Magnetic per- Material μ

meability of various
Vacuum 1
materials, order of magni-
tude [6, 7] Air ≈1
Water, 20 °C 0.999991
Common steel 1000
Ferritic steel 500–3000
Austenitic steel 1–1.6

Table 10.4 Magnetic sus- Material χ

ceptibility of various
Copper -9.65∙10-6
materials (at room tempera-
ture) [7] Water -9.03∙10-6
Nitrogen (g) -8.60∙10-9
Oxygen (g) 1.86∙10-6
Oxygen (l) 3.62∙10-3
Aluminum 2.08∙10-5
Cast iron 50–500

10.2.1 Hysteresis in Magnetization

When a ferromagnetic material is placed within a magnetic field, its level of

magnetization will depend on the strength of the external magnetic field. As the
field strength H is increased the magnetization will also increase, but at a retarding
rate as it approaches an upper saturation limit (Fig. 10.3). When the field strength is
decreased, the level of magnetization will follow a different path and remain at
higher levels for a given field strength in response to decreasing field strength than
when responding to increasing field strength. This type of behavior is known as

Fig. 10.3 Hysteresis

behavior in a ferromagnetic
material, schematic. When the
field strength H is reduced to
zero, the magnetization
M does not return to zero
438 10 Magnetic Properties

hysteresis. The amount of magnetization retained upon release of the magnetic field
is known as remanent magnetization.
Because of this behavior, the magnetic permeability of a material is not constant
but will depend on the magnetic history of the material. Figure 10.3 shows the
hysteresis curve of a ferromagnetic material. The point at which the curve intersects
the magnetic field strength axis (H-axis) gives the coercive field strength. This
coercive field strength is the field strength required to bring the remaining magneti-
zation in the material (remanent magnetization) back to zero.
Materials with high levels of remanent magnetization tend to retain their magne-
tism. These types of materials are most useful in such commercial products as
magnetic memory tapes, magnetic strips, and permanent magnets in electric motors
and generators.
On the other hand, materials with low levels of remanent magnetization tend to
change their magnetism. They are good, e.g., for write-read devices for magnetic
data storage units. Classic computer hard disk drives (HDDs) are magnetic devices.
They consist of rotating disks coated with a magnetic material. A read/write head
writes bits by magnetizing the surface in two states called 1 or 0. One bit on the
magnetic material has a dimension of about 10 nm. Reading of data takes place in
reverse; the read head scans the magnetization of the surface without contact.
Magnetic hard drives are increasingly being replaced by solid-state drives
(SSDs). Like a field-effect transistor (FET), the SSD flash semiconductor memory
can have two different states: electrically charged and uncharged.

Definition
Hysteresis is a change that occurs with a delay to the cause. As a result,
materials can have properties that result from a previous treatment of the
material. With hard magnetic materials in magnetic strips or computer hard
drives, these properties are used to store information.

The source used for a magnetic field can be a permanent magnetic material
(magnet), or it can be an electromagnet, acting like a magnet only when energized
by an electric current, which can be switched on and off (Fig. 10.4). Our planet Earth
also has a magnetic field, but its strength is very small when compared to the strength
of magnetic fields from magnets used in commercial industry. The magnetic field
strength H has the SI unit A
m‐1; the magnetic flux density B is given in the unit
(Tesla) 1 T = 1 V
s
m-2.
The Earth’s magnetic field has a horizontal magnetic flux density of about 40 μT
near the ground, clinical MRI examinations are carried out with about 3 T, and the
magnitude of strong magnetic fields that can be produced in the laboratory is 10 T, in
short pulses up to 100 T.
10.2 Magnetization 439

Fig. 10.4 Magnetic field in the vicinity of an electric current: wire (right), wire loop (left),
schematic

10.2.2 Metal Separation

The most obvious property of the magnetic field is the force effect of a magnet on
other bodies. In material recycling, such as in food packaging, metals can be
separated and sorted in this way. Sorting is carried out according to susceptibility.
While ferromagnetic and paramagnetic metals are attracted to magnets, diamagnetic
metals experience a repulsive force in alternating magnetic fields. This results from
the magnetic field of the eddy currents induced in the diamagnetic metal. Strong
alternating magnetic fields can be easily generated by rotating permanent magnets.
In order to separate even small paramagnetic metal parts from the product stream,
e.g., a powder, strong magnets are required at a short distance from the product.
Packaged products can be tested for metallic contaminants with inductive metal
detectors consisting of a transmitter coil and a receiver coil. Products being inspected
are passed through the alternating electromagnetic field of the transmitter coil. The
field change by each individual product is monitored with the receiver coil. A
material with electrical conductivity experiences an electrical induction voltage in
the variable magnetic field, which leads to eddy currents in the material. The
magnetic field of these eddy currents weakens the causative field and can be detected
by the receiver coil. As a result of induction, a foreign body with electrical conduc-
tivity can be detected in the product. The higher the magnetic susceptibility of the
foreign body, the stronger the signal; i.e., ferromagnetic materials are more easily
detected than paramagnetic and diamagnetic materials. The penetration depth of the
alternating electromagnetic field depends on the frequency of the field. Detection
limits are often given as the diameter of a sphere made of the material in question,
which can barely be detected. Since foods themselves have a certain electrical
conductivity (see Chap. 9), they themselves cause a signal in the metal detector,
the so-called product effect. For foods that have a large product effect, the detection
of small metal parts becomes more difficult. Inductive metal detectors are often
tunnel-shaped so that bulk materials, liquids, or conveyor belts with packaged food
can be passed through them. In Fig. 10.5, an arrangement of transmitter and receiver
coils for a tunnel-shaped metal detector is shown schematically.
440 10 Magnetic Properties

Fig. 10.5 Transmitter (I) and

receiver (II) of a three-coil
arrangement, schematic

10.2.3 Lorentz Force

When electrically charged particles are moving at a velocity v in a magnetic field

with field strength B, they will be subjected to a force acting on them called the
Lorentz force FL. The Lorentz force will increase with increasing velocity, charge,
and field strength and will act in the direction that is mutually perpendicular to both
the velocity of the moving particle and the magnetic field. The Lorentz force acting
on a charged particle moving in a magnetic field can be expressed as follows:
→ →
jF L j = Q
v
B
sin∢ v ; B ð10:14Þ

FL Lorentz’s force in N
Q electric charge in C
v velocity in m
s-1
B magnetic flux density in V
s
m-2

The Lorentz force is the reason why charged particles cannot travel in a straight
line when passing through a magnetic field. The Lorentz forces will “push” them in a
direction perpendicular to that of their initial travel direction, so that charged
particles will move along a curved path instead of a straight path.
In symmetrical cases, this leads to circular currents called eddy currents. Eddy
currents cause heating of a material. In this way, the temperature of electrically
conductive materials can be increased by induction without contact. This is the
principle of so-called induction cooking.

10.2.4 Induction Cooker

Under a glass-ceramic plate an electrically operated coil is placed that generates an

alternating electromagnetic field of 20–100 kHz. In the bottom of a cooking pot
made of an electrically conductive material, eddy currents are generated, which
10.2 Magnetization 441

Fig. 10.6 Example of the construction of induction cookware, schematic: 1. ferromagnetic stain-
less steel on the outside, 2. aluminum or copper in the core, 3. standard stainless steel on the inside

cause heating of the bottom of the pot. An electrically non-conductive material on

the glass-ceramic plate shows no effect. The advantage of inductive heating is that
the heat is generated within the bottom of the pot and does not have to get there by
heat conduction (hot plate), heat transfer (hot flame), or heat radiation (infrared
radiator). For this reason, the heating rate in inductive cooking is high. The electrical
resistance of the material from which the bottom of the pot is made must not be so
low to get heat generated by eddy currents become high. Here, stainless steel with a
higher specific resistance is more advantageous than copper or aluminum which
have a higher electrical conductivity. Due to the skin effect, eddy currents of high-
frequency fields are predominantly located in shallow material depths. In order to
effectively transport the heat generated there into the volume of the pot bottom,
materials with high thermal conductivity can be encapsulated in the stainless-steel
pot bottom. While inductive heating works in principle with all electrically conduc-
tive materials, a magnetic stainless steel (ferritic or martensitic stainless steel) is
advantageous for induction pots: Below the Curie temperature, the periodic demag-
netization of the Weiss districts in the alternating electromagnetic field of the
induction cooker causes additional heat generation. It is significantly larger than in
paramagnetic or diamagnetic materials where no Weiss districts exists. The high
permeability of magnetic stainless steel further concentrates the electromagnetic
field of the cooker in the pot, which means radiation losses are decreased. Some
induction cookers even cannot be started if the magnetic permeability of the cook-
ware used is detected as too low (Fig. 10.6).

Bottom line
Cookware for induction cookers is often made of several materials to obtain a
combination of their properties. The bottom of a pot or pan from the outside to
the inside: Ferromagnetic stainless steel on the outside ensures optimal use of
the magnetic field. This is followed by a layer of aluminum or copper,
sometimes both, as these metals have a higher thermal conductivity than
stainless steel. On the inside there is again a layer of stainless steel, because
aluminum or copper is not sufficiently corrosion-resistant against food. Cast
iron can be used in place of stainless steel. It is ferromagnetic and has a slightly
higher thermal conductivity than stainless steel. As an option, the inside of the
cookware can be provided with a non-stick coating for ease of cleaning.
442 10 Magnetic Properties

Fig. 10.7 Generation of an

Hall voltage UH, schematic

A Hall probe consists of a semiconductor platelet through which an electric

current flows. A magnetic field perpendicular to the current direction leads to a
Lorentz force on the charge carriers. In Fig. 10.7 the principle is illustrated: An
electric current I flows through the plate perpendicular to the B field. The Lorentz
force FL directs positive charge carriers in the semiconductor upward and negative
charge carriers downward; this leads to a voltage UH, the Hall voltage, perpendicular
to the current direction. The Hall voltage is proportional to the magnetic flux density
B, so Hall probes can be used to detect magnetic fields and measure the field
strength. Since electric currents are the cause of magnetic fields, currents can be
measured without contacting any surface because of the Hall effect. By using a Hall
probe to detect magnetic field changes, it can, for example, act as a distance sensor or
as a speed sensor. Hall sensors are used in processing plants for pig detection. A
piston-shaped cleaning device, which is used in pipe systems for cleaning or
inspection, is colloquially referred to as a pig [8].
In magnetic inductive flowmeters (MID, also called magnetars), the Lorentz force
generates an electrical voltage proportional to the flow velocity, analogous to the
Hall effect. This makes it possible to measure the volume flow in a pipe. The
principle of magnetic inductive flow sensors is explained in Chap. 16.
The Lorentz force can also be used to direct electron beams. In this way, in
electron microscopes, rays can be focused with so-called magnetic lenses. The
deflection of charged particles in mass spectrometers, synchrotrons, and cyclotrons
is also caused by the Lorentz force.

10.3 Magnetic Resonance

Spectroscopic methods, in which the transitions between different precession states

of magnetic moments are exploited, are called magnetic resonance methods. A
distinction is made between nuclear magnetic resonance (NMR), electron spin
resonance (ESR), cyclotron resonance, ferromagnetic resonance (FMR), nuclear
quadrupole resonance (NQR), and muon spin resonance (μSR). Let us first look at
the nuclear spin resonance.
10.3 Magnetic Resonance 443

Fig. 10.8 Rotating sphere N

with a magnetic moment μ P
(magnetic dipole with north
pole N and south pole S) as a
simple model of an atomic
nucleus with a magnetic
moment
S

Table 10.5 Atomic nuclei with and without magnetic moments

Number of protons Mass number Named Magnetic moment Example
19
Odd Odd Odd-odd nucleus Yes F
13
Even Odd Even-odd nucleus Yes C
1
Odd Even Odd-even nucleus Yes H
12
Even Even Even-even nucleus No C

Fig. 10.9 Adjustment

options of the nuclear spin in
magnetic field B. The two P
states differ by the energy
difference ΔE

B P B

Atomic nuclei with an odd number of protons (atomic number) or an odd atomic
mass number (the sum of neutrons and protons) have a magnetic moment. The
magnetic moment of an atomic nucleus is very small; it is about a factor of 1000 less
than the magnetic moment of the atomic shell. We can imagine such an atomic
nucleus with the help of a simple model where a rotating sphere within which the
nuclear charge has a magnetic north pole and south pole (Fig. 10.8).
Atomic nuclei of the isotopes 1H, 13C, 19F, or 31P therefore have a magnetic
moment. Not the atomic nucleus of 12C, which consists of six neutrons and six
protons. 12C cannot therefore be investigated by NMR. In Table 10.5, these
examples are listed.
Like atoms, atomic nuclei can also go to higher energy levels by excitation,
i.e., by energy absorption. For quantum mechanical reasons, there are only certain
discrete states that can be taken. There are two possible states for the nuclear
spin. Without an external magnetic field, these two states are practically the same
energetically, but due to the magnetic moment of the atomic nucleus, in a
→
magnetic field it is a difference whether μ is pointing in the direction of the
field or in the opposite direction. Figure 10.9 shows the two states; the vector
444 10 Magnetic Properties

Fig. 10.10 Energetic states

of atomic nucleus spin. In a
magnetic field (II), the energy
difference ΔE between energy
states is higher than without a
magnetic field (I)

→ →
μ performs a precession movement around vector B with a characteristic fre-
quency, the Larmor frequency. Figure 10.10 illustrates the splitting into two energy
levels, which differ by ΔE.

Definition
Precession is the change of direction of the axis of a rotating body when
external forces act on it. The part of the external force acting perpendicular to
the axis of rotation causes a torque that causes the axis of rotation to precess. If
the torque is constant, the axis of rotation circles around its original position
with the precession frequency.

This energy difference ΔE is used in magnetic resonance spectroscopy. By

studying which frequencies of an electromagnetic external field a material can
absorb, information about this excitation energy ΔE is obtained. As soon as an
excitation frequency matches the Larmor frequency of the material, resonance
occurs. In nuclear resonance, these frequencies are in the range of radio waves
(30 MHz to 300 GHz). The resonance frequency of the atomic nucleus depends
on its magnetic moment μ and the strength of the magnetic field B used. It
results from:
h
fL =2
μ
B ð10:15Þ

to

2
μ
B
fL = ð10:16Þ
h

h Planck’s constant in J
s

fL Lamor frequency in s-1
μ magnetic moment in C∙m
B magnetic flux density in V
s
m-2
10.3 Magnetic Resonance 445

Example
The Lamor frequency of a proton 1H results according to
2
μ
B
fL = = B
42:5 MHz
T - 1
h
In a magnetic field with 1 T
2
μ
B
fL = = 1 T
42:5 MHz
T - 1 = 42:5 MHz
h
and in a magnetic field with 24 T

2
μ
B
fL = = 24 T
42:5 MHz
T - 1 = 1:02 GHz
h

However, the exact resonance frequency depends somewhat on the physical and
chemical composition of the atomic space surrounding the nucleus. This influence is
known as chemical shift of the resonance frequency. The electron clouds of an atom,
as well as those of neighboring atoms and their atomic bonding, also further
influence the resonance frequency slightly. These combined influences cause a
measurable difference in resonant frequencies between small and large atoms and
their types of bonding to other atoms. For example, we can use detection of
resonance frequency to distinguish between the hydrogen nucleus in a C–H bonding
and an O–H bonding. Nuclear magnetic resonance (NMR) spectroscopy is based
upon the ability to make this type of distinction.
NMR techniques are used to measure the resonance frequencies (absorption
frequencies) of atomic nuclei. This gives us information, not only about the atom
but also about the state of chemical bonding in the neighborhood of the atom. In this
way, NMR spectra can help us to identify chemical groups and side groups, as well
as their chemical state. This is the principle of material investigation by magnetic
resonance.

Definition
The Zeeman effect is the splitting of an energy level into two different energy
levels by a magnetic field. The Zeeman effect was discovered during the
splitting of spectral lines as an effect of the magnetic field on the electrons
of an atom. Since certain atomic nuclei in the magnetic field also have a
splitting of nuclear spin energy levels, this is analogously referred to as the
Zeeman effect of the atomic nucleus.

Simplified, an NMR spectrometer consists of a magnetic field in which a sample

holder is located. A transmitter coil generates electromagnetic radio waves, and a
446 10 Magnetic Properties

Fig. 10.11 Schematic of an

NMR spectrometer. 1 Magnet,
2 sample holder, 3 transmitter
and receiver coil

Fig. 10.12 Relaxation

signal. The attenuation of the
signal is called FID, free
induction decay
0
I

detector coil records the behavior of the sample, i.e., checks for resonance.
Figure 10.11 shows an NMR spectrometer schematically.
A distinction is made between pulsed NMR and continuous wave methods (CW).
While in pulse operation the radio frequency signal is irradiated for a short moment,
only, the radio frequency in CW operation is permanent. To find the resonance
frequency in CW operation, we can either slowly vary the magnetic field strength
(field sweep) or vary the radio frequency (frequency sweep) with a constant mag-
netic field. In pulse operation, the atomic nuclei are excited in the constant magnetic
field with a radio pulse. The duration of the pulse is in the ms range. The direction of
the radio frequency field is inclined toward the direction of the B field by 90°. The
atomic nuclei absorb the “appropriate” frequencies and absorb energy. After the end
of the radio pulse, the release of this absorbed energy by the atomic nuclei begins.
This is called relaxation. The relaxation signals of an atomic nucleus appear at the
detector coil as a cosine oscillation, the amplitude of which decreases rapidly as a
result of damping. The temporal decrease of this signal is called free induction
decay, FID (Fig. 10.12). From the kinetics of damping, we get information about the
relaxing nuclei.
10.3 Magnetic Resonance 447

Definition
Radio waves are electromagnetic waves in the range of 30 MHz and 300 GHz.

Bottom line
Nuclear magnetic resonance spectroscopy (NMR) determines the frequencies
that are absorbed by the atomic nuclei of a material in the magnetic field. The
frequencies provide information about the gyromagnetic ratio of the nuclei,
which depends on the material environment of the nucleus. The intensities of
the single frequencies and the relaxation behavior provide further information.

In high-resolution nuclear resonance spectroscopy (HR-NMR) the relaxation

signal is analyzed to see what frequencies it contains. This is achieved by Fourier
transformation of the signal. The Fourier transformation provides the frequency
spectrum, i.e., the intensities of the occurring frequencies, see Fig. 10.13. If the
determined intensities are not applied as a function of the frequency but as a function
of the chemical shift compared to a reference value, this is called NMR spectrum in
the chemical laboratory language [9].
The intensities in the NMR frequency spectrum are a measure of the number of
atomic nuclei that show resonance at that frequency. The frequencies are character-
istic of the atomic nucleus, i.e., for the isotope under consideration and its electronic
physicochemical environment. For example, the frequencies of 13C and 12C isotope
differ, as well as the frequencies of proton (1H) and deuteron (2D). The isotope ratios
in food depend on their origin, e.g., composition of soil, rainwater, and animal feed.
This allows conclusions to be drawn about the growing region of a food such as the
origin with the help of the NMR spectrum. To authenticate food, the NMR spectrum
is compared with spectra stored in databases [9, 10]. The different metabolism in C3
and C4 plants also leads to characteristic 12C/13C isotope ratios, so that conclusions
can be drawn from the NMR spectrum in combination with mass spectroscopy about
which plants were used [11]. In this way, e.g., glucose sugar from sugar cane and
sugar beet can be distinguished from glucose sugar obtained from corn [12]. Similar
applications exist for honey [13], fats and oils [14], wine [15], and other foodstuffs
where fraud plays a role [16].
The spectra of NMR, in which the intensity is plotted over a frequency axis, are
one-dimensional spectra (1D NMR). If we add another coupling quantity, we get
spectra with two frequency axes on which the intensity axis is perpendicular,
so-called two-dimensional spectra (2D NMR). With the help of multidimensional
NMR spectroscopy (2D, 3D, etc.) the analytic specificity regarding the chemical
environment of the atomic nuclei under consideration can be increased and thus the
field of application of NMR can be expanded [10, 12, 17].
448 10 Magnetic Properties

Fig. 10.13 HR-NMR: By

Fourier transformation of the
free induction decay (upper
picture) we get the intensities
of the resonance frequencies
(picture in the middle) of a
sample. Often the frequency 0
shift related to a standard is I
used in such spectra (picture at
the bottom)

5 4 3 2 1 0
chemical shift/ ppm
10.3 Magnetic Resonance 449

Definition
Isotopes are atoms with the same atomic number but different masses. They
thus have the same number of protons but a different number of neutrons in the
atomic nucleus. Isotopes can be distinguished by NMR in combination with
mass spectroscopy. For example, the isotope ratio of 12C and 13C can be
determined in food.

In low-resolution nuclear resonance spectroscopy (LR-NMR), the FID signal is

not analyzed in regard to the frequencies contained, but to its shape, i.e., with regard
to its attenuation, which depends on the relaxation time of the atomic nuclei. The
FID signal results from a superposition of spin–lattice relaxation (relaxation time T1)
and spin–spin relaxation (T2) of the atomic nuclei involved [9].
Protons (1H atomic nuclei) relax at different rates depending on the physicochem-
ical environment of the atomic nuclei. For example, protons in solid phases relax
faster than protons in liquid phases and protons in water relax faster than protons in
oil. For these reasons, LR-NMR can be used to determine, e.g., fat contents, water
contents, and solid–liquid ratios in the sample. In a strict sense, this type of NMR
spectroscopy (so-called time domain NMR) does not belong to the spectroscopic
methods, since no spectra (i.e., intensity as a function of frequency or frequency
domain NMR) are obtained. Table 10.6 lists the terms.
In Fig. 10.14, the FID signal of a fat sample is shown schematically. The initial
loss of intensity is mainly caused by the solid-phase protons in the investigated
material. The longer delay part of the signal is determined by liquid-phase protons,
which have a longer relaxation time. By mathematically reconstructing these two

Table 10.6 Pulse NMR, differences between high-resolution and low-resolution NMR
Low resolution High resolution
Synonym: Time domain NMR Frequency domain NMR
Fourier transformation of FID: No Yes
Evaluation of: Attenuation of FID Frequency spectrum

Fig. 10.14 FID signal of a

fat sample after an NMR
pulse: solid (I) and liquid
phase (Il) show different
relaxation curves. The
measured FID (III) is the
superposition of both curves
450 10 Magnetic Properties

relaxation curves from the measurement curve, the solid–liquid ratio of the sample
(e.g. solid fat content, SFC) can be determined [18, 19].
For the 1H-low-resolution NMR various possibilities of signal evaluation are
possible, such as comparison of intensities, evaluation of the relaxation behavior,
and evaluation of the spin echo. From the intensity, we get information about the
number of protons. By appropriate sample preparation, it can then be achieved that
this number of protons correlates with the fat content or the water content in the
sample. Due to the different relaxation behavior of protons in a solid and liquid
phase, solid–liquid ratio can be determined. To enhance the different relaxation
behavior of solid and liquid substance, the spin–echo technique can be used. Here,
after a radio pulse, a short moment is waited until most of the solid-phase protons are
relaxed and then a second 90° rotated radio pulse is irradiated. The result is another
decay signal, the so-called spin echo, which is caused by the nuclei with the greater
relaxation time [12].
Imaging methods of nuclear resonance are referred to as MRI (magnetic reso-
nance imaging) methods. Like in biomedical engineering, it is possible to visualize
the structure of food or agricultural products. The investigation of food structure can
serve in conducting quality assurance [20–23] in food processing [24–26].
While high-resolution NMR devices need more space, low-resolution 1 H-NMR
(time domain NMR) devices are available as compact desktop devices [27]. Further
developments include portable NMR devices for rapid quality checks; see Chap. 16.

10.3.1 Electron Spin Resonance

While we have been studying effects from the magnetic moment of atomic nuclei so
far, we now look at the magnetic moment that comes from unpaired electrons in an
atom. Materials with atoms that have one or more unpaired electrons show the
Zeeman effect in the magnetic field. This means that these atoms absorb a character-
istic radio frequency in the magnetic field similar to NMR. The analysis method is
called electron spin resonance (ESR) or electron paramagnetic resonance (EPR).
Since chemical radicals are also compounds with an unpaired electron, they are
detectable by ESR. This is used to detect irradiation of food and packaging. High-
energy electromagnetic radiation such as gamma radiation leads to the formation of
radicals, which can be detected by ESR [28]. However, due to the high reactivity of
radicals their concentration is not constant over time, and this must be considered
when interpreting the ESR spectra. There are methods for irradiation checks of meat,
fruits, vegetables, spices, cereals, oilseeds, coffee, hazelnuts, sugars, cellulose, etc.
[29–35]. Since the oxidative spoilage of food is also associated with radicals, ESR
can be used to assess the antioxidant capacity of foods and additives [30]. Electron
spin resonance imaging (ESRI) is a method to visualize the spatial concentration
distribution of the radicals [28].
References 451

Further Reading

Freezing under the influence of magnetic fields [36–38]

Preservation process under the influence of magnetic fields [39]
Thermometry with the help of paramagnetic particles [40]
Honey, beer, spices: authentication by NMR [13, 41,
42]
Coffee: authentication by NMR [43]
Wine: authentication by NMR [15]
Apple: tissue characterization by MRI [20]
Automated classification of food products using 2D low-field NMR [44]
Oilseed residues—simultaneous determination of oil and water content by [45, 46]
pulsed NMR
Animal and vegetable fats and oils, determination of solid content of fat [19]
(SFC)
Cheese and dairy products: metabolites by NMR [47]
Meat structure using MRI [21, 22]
Whey protein gels, drying, MRI [48]
NMR for rheological characterization of fluids [49]
Hazelnut: irradiation detection by ESR [34]
Meat (beef, pork, chicken), irradiation detection by ESR [32]
Oxidation of food, examination by ESR [30]
Antiplastization of starch-sucrose blends—positron lifetime and NMR study [50]

Summary
The reasons for the different behavior of materials in the magnetic field are
explained in a fundamental way and illustrated based on applications such as
induction cooking and metal detection. The different forms of magnetic
resonance to study the composition, structure, and origin of food are explained
in simple words. Imaging techniques and electron spin resonance for the
detection of radicals in food are addressed. At the end of the chapter, applica-
tion examples are listed, which illustrate the potential of magnetic resonance
spectroscopy and may stimulate further studies.

References
1. Rodriguez AC, Sanchez-Benitez J, Sanz PD (2017) Simulation of the magnetic freezing process
applied to foods. Food Eng Rev 9(4):271. https://doi.org/10.1007/s12393-017-9169-3
2. Minano HLA, Silva ACD, Souto S, Costa EJX (2020) Magnetic fields in food processing
perspectives, applications and action models. Processes 8(7). https://doi.org/10.3390/
pr8070814
452 10 Magnetic Properties

3. Barbosa-Canovas GV, Schaffner DW, Pierson MD, Zhang QH (2000) Oscillating magnetic
fields. J Food Sci 65(s8):86. https://doi.org/10.1111/j.1750-3841.2000.tb00622.x
4. Rodríguez Plaza AC (2017) Characterization of electromagnetic freezing in food matrixes and
model food. University Polytechnika, Madrid
5. Pobell F (2007) Matter and methods at low temperatures. Springer, Heidelberg. https://doi.org/
10.1007/978-3-540-46360-3
6. Kohlrausch F (1996) Praktische Physik Bd.3. Teubner, Stuttgart. https://doi.org/10.1002/piuz.
19870180109
7. Czichos H, Hennecke M, e.V. AVH (eds) (2008) HÜTTE - Das Ingenieurwissen. Springer,
Heidelberg. https://doi.org/10.1007/978-3-540-71852-9
8. Mühlthaler W, Hiltscher G, Smits J (1999) Molchtechnik. Wiley-VCH, Weimheim. https://doi.
org/10.1002/9783527625789
9. Hatzakis E (2019) Nuclear magnetic resonance (NMR) spectroscopy in food science: a com-
prehensive review. Compr Rev Food Sci Food Saf 18(1):189. https://doi.org/10.1111/
1541-4337.12408
10. Cao R, Liu X, Liu Y, Zhai X, Cao T, Wang A, Qiu J (2020) Applications of nuclear magnetic
resonance spectroscopy to the evaluation of complex food constituents. Food Chem:128258.
https://doi.org/10.1016/j.foodchem.2020.128258
11. Zhou L, Zhang C, Qiu Z, He Y (2020) Information fusion of emerging non-destructive
analytical techniques for food quality authentication: a survey. TrAC Trends Anal Chem 127:
115901. https://doi.org/10.1016/j.trac.2020.115901
12. Marcone MF, Wang S, Albabish W, Nie S, Somnarain D, Hill A (2013) Diverse food-based
applications of nuclear magnetic resonance (NMR) technology. https://doi.org/10.1016/j.
foodres.2012.12.046
13. Kuballa T, Brunner TS, Thongpanchang T, Walch SG, Lachenmeier DW (2018) Application of
NMR for authentication of honey, beer and spices. Curr Opin Food Sci 19:57. https://doi.org/
10.1016/j.cofs.2018.01.007
14. Guyader S, Thomas F, Portaluri V, Jamin E, Akoka S, Silvestre V, Remaud G (2018)
Authentication of edible fats and oils by non-targeted 13C INEPT NMR spectroscopy. Food
Control 91:216. https://doi.org/10.1016/j.foodcont.2018.03.046
15. Mascellani A, Hoca G, Babisz M, Krska P, Kloucek P, Havlik J (2021) 1H NMR chemometric
models for classification of Czech wine type and variety. Food Chem 339:127852. https://doi.
org/10.1016/j.foodchem.2020.127852
16. Sobolev AP, Thomas F, Donarski J, Ingallina C, Circi S, Cesare Marincola F, Capitani D,
Mannina L (2019) Use of NMR applications to tackle future food fraud issues. Trends Food Sci
Technol 91:347. https://doi.org/10.1016/j.tifs.2019.07.035
17. Gouilleux B, Charrier B, Akoka S, Felpin F-X, Rodriguez-Zubiri M, Giraudeau P (2016)
Ultrafast 2D NMR on a benchtop spectrometer: applications and perspectives. TrAC Trends
Anal Chem 83:65. https://doi.org/10.1016/j.trac.2016.01.014
18. Linke C, Guthausen G, Floter E, Drusch S (2018) Solid fat content determination of dispersed
lipids by time-domain NMR. Eur J Lipid Sci Technol 120(4). https://doi.org/10.1002/ejlt.
201700132
19. DIN EN ISO 8292 (2010) Tierische und pflanzliche Fette und Öle - Bestimmung des Festanteils
von Fett durch das Verfahren mit gepulster magnetischer Kernresonanz. Beuth, Berlin. https://
doi.org/10.31030/1633109
20. Defraeye T, Lehmann V, Gross D, Holat C, Herremans E, Verboven P, Verlinden BE, Nicolai
BM (2013) Application of MRI for tissue characterisation of ‘Braeburn’ apple. Postharv Biol
Technol 75:96. https://doi.org/10.1016/j.postharvbio.2012.08.009
21. Clerjon S, Bonny J-M (2018) NMR imaging of meat. In: Webb GA (ed) Modern magnetic
resonance. Springer International, Cham, pp 1609–1628. https://doi.org/10.1007/978-3-319-
28388-3_131
22. Bajd F, Skrlep M, Candek-Potokar M, Sersa I (2017) MRI-aided texture analyses of compressed
meat products. J Food Eng 207:108. https://doi.org/10.1016/j.jfoodeng.2017.03.026
References 453

23. Van As H, van Duynhoven J (2013) MRI of plants and foods. J Magn Reson 229:25. https://doi.
org/10.1016/j.jmr.2012.12.019
24. Ezeanaka MC, Nsor-Atindana J, Zhang M (2019) Online low-field nuclear magnetic resonance
(LF-NMR) and magnetic resonance imaging (MRI) for food quality optimization in food
processing. Food Bioprocess Tech 12(9):1435. https://doi.org/10.1007/s11947-019-02296-w
25. Xu FF, Jin X, Zhang L, Chen XD (2017) Investigation on water status and distribution in
broccoli and the effects of drying on water status using NMR and MRI methods. Food Res Int
96:191. https://doi.org/10.1016/j.foodres.2017.03.041
26. Bows JR, Patrick ML, Nott KP, Hall LD (2001) Three-dimensional MRI mapping of minimum
temperatures achieved in microwave and conventional food processing. Int J Food Sci Techol
36(3):243. https://doi.org/10.1046/j.1365-2621.2001.00444.x
27. Blümich B (2016) Introduction to compact NMR: a review of methods. TrAC Trends Anal
Chem 83:2. https://doi.org/10.1016/j.trac.2015.12.012
28. Leveque PP, Godechal Q, Gallez B (2008) EPR spectroscopy and imaging of free radicals in
food. Isr J Chem 48(1):19. https://doi.org/10.1560/IJC.48.1.19
29. Yu LL, Cheng Z (2008) Application of electron spin resonance (ESR) spectrometry in nutra-
ceutical and food research. Mol Nutr Food Res 52(1):62. https://doi.org/10.1002/mnfr.
200700395
30. Andersen ML, Skibsted LH (2017) ESR spectroscopy for the study of oxidative processes in
food and beverages. In: Webb GA (ed) Modern magnetic resonance. Springer International,
Cham, pp 1–14. https://doi.org/10.1007/978-3-319-28275-6_25-1
31. Ahn J-J, Akram K, Kim H-K, Kwon J-H (2012) Electron spin resonance spectroscopy for the
identification of irradiated foods with complex ESR signals. Food Anal Method 6(1):301.
https://doi.org/10.1007/s12161-012-9440-4
32. Park YD, Kim DY, Jin CH, Yang HS, Choi DS, Yook H-S, Byun M-W, Jeong IY (2011)
Development of a method based on ESR spectroscopy for the identification of irradiated beef,
pork, and chicken meats. Food Sci Biotechnol 20(2):367. https://doi.org/10.1007/s10068-011-
0051-4
33. Marchioni E (2006) ESR as a technique for food irradiation detection. In: Webb GA
(ed) Modern magnetic resonance. Springer, Dordrecht, pp 1855–1860. https://doi.org/10.
1007/1-4020-3910-7_211
34. Chiappinelli A, Mangiacotti M, Tomaiuolo M, Trotta G, Marchesani G, Chiaravalle AE (2019)
Identification of X-ray-irradiated hazelnuts by electron spin resonance (ESR) spectroscopy. Eur
Food Res Technol 245(10):2323. https://doi.org/10.1007/s00217-019-03349-2
35. Barba FJ, Roohinejad S, Ishikawa K, Leong SY, El-Din A Bekhit A, Saraiva JA, Lebovka N
(2020) Electron spin resonance as a tool to monitor the influence of novel processing
technologies on food properties. Trends Food Sci Technol 100:77. https://doi.org/10.1016/j.
tifs.2020.03.032
36. Otero L, Rodríguez AC, Pérez-Mateos M, Sanz PD (2016) Effects of magnetic fields on
freezing: application to biological products. Compr Rev Food Sci Food Saf 15(3):646. https://
doi.org/10.1111/1541-4337.12202
37. Dalvi-Isfahan M, Hamdami N, Xanthakis E, Le-Bail A (2017) Review on the control of ice
nucleation by ultrasound waves, electric and magnetic fields. J Food Eng 195:222. https://doi.
org/10.1016/j.jfoodeng.2016.10.001
38. Woo MW, Mujumdar AS (2010) Effects of electric and magnetic field on freezing and possible
relevance in freeze drying. Dry Technol 28(4):433. https://doi.org/10.1080/
07373930903202077
39. Grigelmo N, Soliva-Fortuny R, Barbosa-Cánovas G, Martin-Belloso O (2011) Use of
oscillating magnetic fields in food preservation. In: Zhang HQ, Barbosa-Cánovas GV,
Balasubramaniam VM, Dunne CP, Farkas DF, Yuan JTC (eds) Nonthermal processing
technologies for food. Blackwell, pp 222–235. https://doi.org/10.1002/9780470958360.ch16
40. Ghiron K, Litchfield B (1997) Magnetic thermometry in the aseptic processing of foods
containing particulates. J Appl Phys 81(8):4321. https://doi.org/10.1063/1.364756
454 10 Magnetic Properties

41. Galvin-King P, Haughey SA, Elliott CT (2018) Herb and spice fraud; the drivers, challenges
and detection. Food Control 88:85. https://doi.org/10.1016/j.foodcont.2017.12.031
42. Siddiqui AJ, Musharraf SG, Choudhary MI, Rahman AU (2017) Application of analytical
methods in authentication and adulteration of honey. Food Chem 217:687. https://doi.org/10.
1016/j.foodchem.2016.09.001
43. Milani MI, Rossini EL, Catelani TA, Pezza L, Toci AT, Pezza HR (2020) Authentication of
roasted and ground coffee samples containing multiple adulterants using NMR and a
chemometric approach. Food Control 112:107104. https://doi.org/10.1016/j.foodcont.2020.
107104
44. Greer M, Chen C, Mandal S (2018) Automated classification of food products using 2D
low-field NMR. J Magn Resonan 294:44. https://doi.org/10.1016/j.jmr.2018.06.011
45. ISO 10632 (2000) Ölsaatenrückstände - Gleichzeitige Bestimmung des Öl- und Wassergehalts -
Spektralphotometrisches Pulse NMR-Verfahren. Beuth, Berlin
46. DIN EN ISO 10565 (1998) Ölsamen - Gleichzeitige Bestimmung des Öl- und Wassergehaltes -
Verfahren mit gepulster Kernresonanzspektroskopie. Beuth, Berlin. https://doi.org/10.31030/
7537439
47. Scano P, Cusano E, Caboni P, Consonni R (2019) NMR metabolite profiles of dairy: a review.
Int Dairy J 90:56. https://doi.org/10.1016/j.idairyj.2018.11.004
48. Fan LY, Yang JX, Casali RA, Jin X, Chen XD, Mercade-Prieto R (2017) Magnetic resonance
imaging (MRI) to quantify the swelling and drying of whey protein hydrogels. J Food Eng 214:
97. https://doi.org/10.1016/j.jfoodeng.2017.06.033
49. Callaghan PT (1999) Rheo-NMR: nuclear magnetic resonance and the rheology of complex
fluids. Rep Prog Phys 62(4):599. https://doi.org/10.1088/0034-4885/62/4/003
50. Martini F, Hughes DJ, Badolato Bönisch G, Zwick T, Schäfer C, Geppi M, Alam MA, Ubbink J
(2020) Antiplasticization and phase behavior in phase-separated modified starch-sucrose
blends: a positron lifetime and solid-state NMR study. Carbohydr Polym 250:116931. https://
doi.org/10.1016/j.carbpol.2020.116931
Electromagnetic Properties
11

Electromagnetic properties of foods are all those properties that are related to
interaction of the food with electromagnetic waves. Among these interactions is
the absorption of a certain wavelength, which can give us information about the food
material. This information can lead to the design and development of sensors for
quality testing, or energy transfer to accomplish heating, as well as other uses.
Applying different frequency ranges of electromagnetic radiation (Table 11.1) can
lead to the formation of such energy patterns as microwaves, radar, X-ray, and
ultraviolet radiation. In this chapter, we first focus on the reasons for the interaction
of matter with electromagnetic rays and then discuss microwaves and terahertz
waves. The application of lower frequency electromagnetic rays has already been
discussed in Chap. 9 “Electrical properties” and Chap. 10 “Magnetic properties.”
The range of higher frequencies (NIR, visible light, UV, and X-ray) is dealt with in
the following chapter (optical properties) and in Chap. 14 on gamma radiation.

11.1 Electromagnetic Waves

Electromagnetic waves consist of an electric field E and a magnetic field B, which

propagates simultaneously with a periodically changing field strength. In the propa-
gation of the wave, the distance between two wave peaks is the wavelength λ
(Fig. 11.1). The product of wavelength and frequency f is the propagation speed
c of the wave.
λ
f =c ð11:1Þ

c propagation speed in m
s-1

λ wavelength in m
f frequency in Hz

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 455
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_11
456 11 Electromagnetic Properties

Table 11.1 Electromagne- Frequency (Hz) Name

tic radiation
3∙1024
3∙1023
3∙1022 γ-Radiation
3∙1021
3∙1020 X-ray
3∙1019
3∙1018
3∙1017
3∙1016 Ultraviolet
3∙1015
3∙1014 Light infrared (heat)
3∙1013
3∙1012
3∙1011 Radio waves
3∙1010
3∙109
3∙108
3∙107
3∙106
3∙105
3∙104
3∙103 AC electricity
3∙102
3∙101

Fig. 11.1 Electromagnetic

wave with wavelength λ

Electromagnetic waves propagate at the speed of light. The speed of propagation

in air is close to the vacuum speed of light (c0 = 3
108 m
s-1). The propagation rate
in a material (dielectric) is lower; it depends on the interaction of the material with
the electromagnetic wave in question. Figure 11.1 shows that electric and magnetic
field strength oscillate at a 90° angle to each other. If the field strength oscillates in a
fixed plane as in Fig. 11.1, we have a linear polarized wave. Also in terms of electric
material properties, there is the concept of polarization or polarizability. While we
have already learned about magnetic polarizability in Chap. 10), we now want to
understand the causes of electric polarization.
11.2 Electric Polarization 457

11.2 Electric Polarization

As stated earlier, materials that possess polar molecules will respond to an electric
field such as microwave radiation because these polar molecules will become
polarized in the presence of such an electrical field. First, we must distinguish
between molecules that are permanent electric dipoles like water (H2O) from those
that are temporary dipoles like hydrocarbons. When permanent dipoles are brought
into the presence of an electric field, they will orient themselves to become aligned
with the direction of the electric field. This is called orientation polarization.
Temporary dipoles are not normally polar molecules, but they can be deformed
under the influence of an electric field. This causes them to acquire dipoles and
become temporarily polarized, so they will also orient themselves to become aligned
with the direction of the electric field. However, they will immediately lose their
polarization as soon as the electric field disappears when switched off.
Polarization of both permanent and temporary dipoles depends on the strength of
the external electrical field. Both together contribute to the overall polarization P of a
volume V of material by the following expression:

p
P= ð11:2Þ
V

P polarization in C
m-2
p dipole moment in C
m
V volume in m3

The degree of polarization exhibited by a material when in the presence of an

electric field can be quantified by comparing it with the degree of polarization
exhibited in an empty space (vacuum) in that same electric field. Let us consider
an electric displacement field D:
It is

D = ε
ε0
E ð11:3Þ

in a vacuum
D 0 = 1
ε0
E ð11:4Þ

so, the difference is:

D - D 0 = ε
ε 0
E - ε 0
E = ε 0
E ð ε - 1Þ = P ð11:5Þ

the polarization is
P = ðε - 1Þε0
E ð11:6Þ
458 11 Electromagnetic Properties

and with
χ = ðε - 1Þ ð11:7Þ

it is

P = χ
ε0
E ð11:8Þ

D electric displacement field in C
m-2

E electric field in V
m-1
ε0 electric field constant: 8.854
10-12 C
V-1
m-1
ε relative permittivity
χ electric susceptibility

The relative permittivity ε and the susceptibility χ, are measures for the polariza-
tion potential of a material.
The greater the polarization potential of a material the more energy it can absorb
from an electric field. When the direction of the electric field alternates periodically
at a given frequency, then the dipoles also alternate their orientation to follow the
alternating direction of the electric field at the same frequency. Thus, the direction of
polarization in the material alternates with the alternating electric field.
This continuously changing orientation is not completely elastic. A significant
part of the absorbed energy is lost in the form of heat that is generated when the
dipoles try to reverse their direction of orientation at high frequencies. This heat is
absorbed by the material and results in an increase in the temperature of the material.
This is why foods become heated when placed in a microwave oven.
Analogous to mechanics, in the lossless case we speak of an elastic behavior of
the dipoles. The heating of a material due to energy dissipation is thus a consequence
of in-elastic polarization. When distinguishing the two cases, the elastic case is
characterized with χ ′ and ε′ and the inelastic case with χ ″ and ε″, the so-called loss
factor.
How well a material can be heated by electromagnetic absorption is determined
00
by the ratio εε0 and χχ 0 .
00

These material properties play a major role in heating of food by absorption of

electromagnetic radiation, the so-called dielectric heating or microwave heating. The
values of these material sizes depend on the frequency of the electromagnetic
radiation. To understand this let us look at the causes of electromagnetic absorption.
The polarization potential of a material αtotal can be calculated as the sum of the
polarization potential of the temporary (induced) and permanent dipoles αind and
αpermanent, respectively.

αtotal = αind þ αpermanent ð11:9Þ

11.2 Electric Polarization 459

So, we can say the polarization is:

P = N
αtotal
E ð11:10Þ

The orientation of permanent dipoles, such as water (H2O) molecules, when placed
into an external electric field, is impeded by the thermal movement of the molecules.
The increasing temperature causes the permanent polarization to decrease. This
decrease in permanent polarization potential as a function of temperature can be
estimated as follows:

μ2
αpermanent = ð11:11Þ
3kT
With this it is:

μ2
P = N
αind þ
E ð11:12Þ
3k B T

P polarization in C
m-2
N particle concentration in m-3
αind induced polarization potential in C2
m2
J-1
μ molecular dipole moment in C
m
k Boltzmann’s constant: kB = 1.38064852
10-23 m2
kg
s-2
K-1
T temperature in K
E electric field in V
m-1
ε0 electric field constant: 8.854
10-12 C
V-1
m-1

Data for the induced polarization potential αind of different molecules are listed in
published literature. Often instead of αind the so-called polarization volume α′ is
listed. It is:
α
α0 = ð11:13Þ
4π
ε0
Also, values for the permanent dipole moment μ for different molecules are
available from referenced literature. A few examples are given in Table 11.2.

Table 11.2 Dipole moments and polarization potential of some simple molecules, examples from
[1]
μ (C∙m) μ (D) αind (C2
m2
J-1) α0ind (cm3)
H2 0 0 9.11
10-41 8.19
10-25
N2 0 0 1.97
10-41 1.77
10-24
CO2 0 0 2.93
10-40 2.63
10-24
H2O 6.17
10-30 1.85 1.65
10-40 1.48
10-24
C2H5OH 5.64
10-30 1.69 – –
CH3OH 5.70
10-30 1.71 3.59
10-40 3.23
10-24
460 11 Electromagnetic Properties

Sometimes the dipole moment μ is expressed in units of Debyes (D), instead of the
SI units of C∙m. A Debye is D = 3.3
10-30 C∙m.
The Clausius–Mossotti–Debye Eq. (11.14) describes the relationship between
relative permittivity ε and polarization potential αind. It can be used to express the
polarization potential of a material [1]:

ε-1 N μ2
= αind þ ð11:14Þ
ε þ 2 3
ε0 3kB T

This expression has no unit, as we can see by analysis of the right side of the
V
m C2
m2 2
equation:C
m 3 J J
m2
C = 1. We can (Eq. 11.14) also write as
= V
C
C
m

ε-1 3 1 μ2

=N
αind þ ð11:15Þ
ε þ 2 4π 4πε0 3k B T

by using Eq. (11.9) we get:

μ2
αtotal = αind þ αpermanent = αind þ ð11:16Þ
3kB T
With Eq. (11.15) we get
ε-1 3 1

=N

α ð11:17Þ
ε þ 2 4π 4πε0 total
with Eq. (11.13) we can abbreviate
φN = N
α0total ð11:18Þ

The quantity φN is dimensionless and represents the polarization potential caused

by a particle concentration N.
The molar polarization potential can be calculated by:
With Eq. (11.14) φN is identical to

N μ2
φN =
αind þ ð11:19Þ
4π
ε0 3kB T

with Eq. (11.16) we get the relation between φN and the relative permittivity
ε-1 3
φN =
ð11:20Þ
ε þ 2 4π
The mass-specific polarization potential φm can be calculated by dividing φN by
the mass density ρ of the material. It has the SI unit m3
kg-1.
11.2 Electric Polarization 461

φN ε-1 3
φm = =
ð11:21Þ
ρ ε þ 2 4π
ρ
In physical chemistry, instead of mass-specific polarization potential φm the
molar polarization potential φn is often used instead of mass-specific quantities.
We get φn by multiplication of Eq. (11.22) with the molar mass M of the material:
φN ε-1 3
M
φn =
M=
ð11:22Þ
ρ ε þ 2 4π
ρ
Because of

N N
NA n
NA n
NA
M m
NA ρ
NA
N= = = = = = ð11:23Þ
V V
NA V V
M V
M M
we see that multiplication by Mρ is identical to a division by the particle density N and
parallel multiplication by Avogadro’s constant:

M 1
=
NA ð11:24Þ
ρ N

φN volume-based polarization potential in m3
m-3

φn mole-based polarization potential in m3
mol-1
φm mass-based polarization potential in m3
kg-1
M molar mass in kg
mol-1
ρ density in kg
m-3
NA Avogadro’s constant: 6.022
1023 mol-1
N particle density in m-3
N number of particles
V volume in m3
n amount of substance in mol

11.2.1 Temperature Dependency

From Eqs. (11.20) and (11.22) we can derive the following expression which
indicates the temperature dependency of the polarization potential and the permittiv-
ity of a material:

φN N 1 μ2
φm = =
αind þ ð11:25Þ
ρ ρ 4πε0 3k B T

When we measure the relative permittivity ε of a material at different

temperatures, we can calculate the mass-based polarization potential φm according
to Eq. (11.22) and plot it versus 1/T. Then we are able to obtain the permanent dipole
moment μ and the induced polarization potential αind of the material. As shown in
462 11 Electromagnetic Properties

Fig. 11.2 Temperature

dependency of mass-based
polarization potential by
graphical evaluation
b

Mm

1
T

Fig. 11.2 the intercept, a, and the slope, b, provide us with the following expressions
from Eq. (11.25):
αind
a= ð11:26Þ
4π
ε0
ρ
and

μ2
b= ð11:27Þ
12π
ε0
ρ
k B
For this type of characterization of the electrical polarizability of biological
materials such as food, only knowledge of the density is necessary. For questions
in which the molecular weight or at least an average molecular weight M of the
material in question is known, the graphical application of the molar polarizability
volume φn can also be used for evaluation. Then we have
M 3 ε - 1 NA 3 ε - 1
φn =

=

ð11:28Þ
ρ 4π ε þ 2 N 4π ε þ 2
it is

NA N μ2
φn =
αind þ ð11:29Þ
N 4π
ε0 3k B T

or

NA μ2
φn = αind þ ð11:30Þ
4π
ε0 3kB T
11.2 Electric Polarization 463

then the intercept, a, and slope, b, in the φn versus T1 plot are

N A
αind N A
μ2
a= ; b=
4π
ε0 12π
ε0
kB
We can see that the polarization potential and the permittivity of a material
decrease with rising temperature. The reason for this is that with increasing temper-
ature, the random thermal movement of the molecules increases and counteracts the
orientation of the permanent dipoles by the external electric field. The induced
polarization potential, however, is not affected by the movement of the molecules.
As a result, materials consisting of polar molecules show less polarizability at high
temperatures than at low temperatures.
The relative permittivity number of liquid water in a static, electric field (fre-
quency = 0) at 100 °C is ε = 55.7 and at -35 °C it is ε = 107.7 [2]. In dielectric
heating of food [3, 4] knowledge of polarizability and its temperature dependence is
necessary.

Remark
Water that is still liquid at temperatures below 0 °C is called supercooled
water. Very strong supercooling can only be achieved under laboratory
conditions.

11.2.2 Frequency Dependency

The energy absorbed by a molecule leads to translation, oscillation, or rotation of the

molecule. Let us assume the rotation of a small molecule takes about 10-12 s. If we
now increase the frequency of an electric field to over 1012 Hz, only less than 10-12 s
are available per change of direction. As a result, permanent dipoles such as water
molecules can no longer follow an alternating electric field from about this frequency
onward. The contribution of orientation polarization to the overall polarization
therefore decreases significantly when this frequency is reached. The contribution
of displacement polarization, on the other hand, does not change. Figure 11.3
illustrates the course of polarizability with increasing frequency.
Orientation polarization therefore plays practically no role at optical frequencies.
If the frequency of the irradiated alternating field is increased even further up to the
UV range, the displacement polarization is also finally eliminated (Fig. 11.3).
Alternating electromagnetic fields with frequencies in the UV or X-ray range only
interact with the electrons of a molecule and serve as the electronic contribution to
polarization in Fig. 11.3. This means that at these and higher frequencies, the model
of aligned Debye dipoles is no longer applicable. When calculating polarizability,
the contribution of the permanent dipoles can be neglected in this frequency range,
i.e.,
464 11 Electromagnetic Properties

Fig. 11.3 Frequency

dependency of electric
polarization potential [1]
permanent dipoles

Dtotal
temporary dipoles

electronic

radio microwave IR vis UV

f

NA μ2 NA
φn = αind þ ffi
α ð11:31Þ
4π
ε0 3kB T 4π
ε0 ind

The measurement of polarizability with frequencies in the IR range or with

frequencies of visible light provides only the proportion of polarizability based on
displacement polarization. Maxwell’s equation

ε = n2 ð11:32Þ

expresses the electrical polarization at frequencies in the optical range, which is

determined by the refractive index n (s. Chap. 12, optical properties):

NA M 3 ε-1 M 3 n2 - 1
φn = αind =

==

2 ð11:33Þ
4πε0 ρ 4π ε þ 2 ρ 4π n þ 2
What we find here is a strong frequency dependence of the polarizability of the
permittivity and refractive index of the material in question. The dependence of a
physical quantity on the frequency of a wave is called dispersion.

→ Attention
The Latin word dispergere is used with different meanings. In materials
science, dispersion is a multiphase system (cf. Chap. 3, Disperse systems).
Light is scattered by multiphase (“disperse”) systems. In physics, however, the
term is for the frequency dependence of quantities.
11.2 Electric Polarization 465

11.2.3 Complex Permittivity

Materials can absorb electromagnetic radiation elastically or inelastically. Elastic

absorption is an energy storage as it occurs in elastic elements in mechanics. Inelastic
absorption, on the other hand, leads to energy dissipation, i.e., to an increase in
entropy, which is usually associated with heating of the material. Ideally elastically
absorbed radiation can be emitted without loss like the elastic energy of an ideally
mechanical spring. Inelastically absorbed radiation, on the other hand, is not emitted
again, but is converted into heat that remains in the material. This inelastic part is
also called the loss part.
Complex permittivity ε is used to mathematically express the material behavior.
It consists of a real part ε′ for the elastic behavior and an imaginary part ε″ for the
inelastic behavior.
It is

ε = ε0 þ i
ε00 ð11:34Þ

ε complex permittivity
ε′ real permittivity
ε″ imaginary permittivity
p
i imaginary unit -1

Complex quantities have the advantage that two properties can be expressed with
one size. In polar coordinates, the real part and the imaginary part are orthogonal
vectors whose Pythagorean sum indicates the amount of the complex quantity:

ε = ε0 2 þ ε00 2 ð11:35Þ

We have already worked with complex quantities to describe elastic and inelastic
parts of a physical property in previous chapters (Chap. 4 “Rheological properties,”
Chap. 8 “Thermal properties,” and Chap. 9 “Electrical properties”).
An ideal oscillator, i.e., a completely elastic and frictionless oscillator, would
oscillate indefinitely long not losing energy to its environment. Real oscillators, on
the other hand, lose energy to the environment, so the amplitude of the oscillation
decreases over time. This is called the damping of the oscillation. The degree of
damping is determined by the inelastic part ε″ (so-called loss factor). The ratio of the
real part and the imaginary part is the tangent of the angle δ between both vectors. In
this context, the angle δ is called dielectric loss angle and the tangent is called loss
tangent.

ε00
= tan δ ð11:36Þ
ε0
For a material that behaves ideally elastic, i.e., lossless, the angle is δ = 0°. In the
case of a material that behaves completely inelastic, the angle is δ = 90°. For this
466 11 Electromagnetic Properties

Fig. 11.4 Frequency

dependence of the permittivity
of water. Real part ε′ (top) and
imaginary part ε″ (bottom),
according to [5]

reason, the real part is also called “in-phase component” and the imaginary part “out-
of-phase component.”
Let us consider liquid water as our model material: Water can absorb electromag-
netic radiation. Due to the dispersion of permittivity, the absorption is strongly
dependent on the frequency of electromagnetic radiation. At a frequency of
100 MHz, elastic absorption predominates. At a frequency of a few GHz, inelastic
absorption predominates. This leads to heating of the water and is the basis of
microwave heating [5, 6]. In Fig. 11.4, the frequency dependence of the permittivity
is shown.
The substance data of water shown in Fig. 11.4 were determined for high-purity
water with a very low electrical conductivity. If an aqueous solution has a significant
electrical conductivity, the value of the imaginary permittivity increases by the
so-called conductivity term σ, which in turn is frequency-dependent [2]:

σ
ε = ε0 - i
ε00 þ ð11:37Þ
ε0
ω

with
ω = 2π
f ð11:38Þ
11.3 Microwaves 467

11.3 Microwaves

Electromagnetic waves in the frequency range between 300 MHz and 300 GHz are
called microwaves. Since microwaves are used for navigation and
telecommunications purposes (air traffic, satellites, mobile phones, radar, etc.),
their use is regulated by law. In most industrialized nations, the frequencies
915 ± 13 MHz and 2450 ± 50 MHz are reserved for industrial, scientific, and
medical use (regulation by ITU, International Telecommunication Union). With the
relationship for the propagation speed of electromagnetic waves
c=λ
f ð11:1Þ

frequencies and wavelengths can be calculated.

c propagation speed in m
s-1

λ wavelength in m
f frequency in Hz

Example
What is the wavelength of microwave radiation of the frequency 915 MHz or
the frequency 2.45 GHz?
Solution:with c = λ
f it is

c
λ=
f
assuming the speed of light in a vacuum as propagation speed, it is

2:99
108 m
s - 1
λ= = 0:328 m
915
106 s - 1
or

2:99
108 m
s - 1
λ= = 0:122 m
2:45
109 s - 1

Microwaves are basically subject to the same laws as electromagnetic waves of

other frequencies, such as IR or light. Microwaves can be reflected, absorbed, and
transmitted. With the help of metal reflectors, microwaves can be redirected,
focused, or scattered. Materials that strongly absorb microwaves have a high degree
of absorption and thus a low degree of transmission. Materials such as glass,
ceramics, and plastics show little or no microwave absorption. Due to their specific
electrical polarizability (low dissipation factor ε″), they have a low degree of
468 11 Electromagnetic Properties

absorption in this frequency range, but this is dependent on temperature and

frequency.
Figure 11.4 shows the high capability of liquid water for absorption of electro-
magnetic waves in the GHz range. A high value for the loss factor ε″ means a high
absorption coefficient. Therefore, it is not surprising that telecommunication
transmissions in this frequency range are influenced by water in the atmosphere.
For example, heavy rain leads to the weakening of mobile communications
(1–4 GHz), Wi-Fi (2–6 GHz), satellite signals (10–12 GHz), and other similar
services based on microwaves. The attenuation of the telecommunications fields
by atmospheric water is an undesired effect. However, due to the dispersion of the
absorption coefficient of water, there are frequency ranges that are less attenuated by
liquid water, such as radio waves or visible light.

11.3.1 Conversion of Microwaves into Heat

The inelastic absorption of microwaves leads to the heating of a material. While the
conversion of field energy into heat in electrical engineering is considered an
“energy loss” (loss factor ε″), this effect is the desired effect in microwave heating.
The electrically polarizable molecules of a material absorb energy from the irradiated
field if this energy “fits.” In the case of water and similarly constructed molecules,
the frequencies of molecular rotation are in the range of microwave radiation.
Because the absorbent molecules release energy to neighboring molecules, it does
not come to (elastic) reflection of the microwaves. Instead, there is (inelastic)
dissipation of the radiation energy, which leads to a temperature increase in the
material. This does not work with the frequency of radio waves, because at
frequencies of a few MHz, water has a very low absorption coefficient. The
proportion of absorbed radiant energy that is converted into heat depends linearly
on the imaginary part of the electrical permittivity ε″ (loss factor). The power PD
dissipated per unit volume is proportional to the frequency and depends on the
square of the electric field strength E. It is [7]

PD = 2π
ε0
f
ε00
jEj2 ð11:39Þ

or as a numerical value equation

2
PD W
m - 3 = 55:61
10 - 12 jE j V
m - 1
f ðHzÞ
ε00 ð11:40Þ

PD dissipated power in W
E electric field strength in V
m-1
f frequency in Hz
ε″ loss factor
11.3 Microwaves 469

Example
What is the volume-based heating power in a microwave oven at 2 GHz, when
water is heated within it and the field strength is 100 V/m?
Solution:
Asssuming ε″ = 10 with

PD = 2π
ε0
f
ε00
jEj2

we get

PD = 2π
8:854
10 - 12 C
V - 1
m - 1
2
109 Hz
10
ð100 V
m - 1 Þ

2

PD = 55:6
10 - 12 C
V - 1
m - 1
2
1014 s - 1
V2
m - 2
PD = 11, 126:3 V
A
s
s - 1
m - 3 = 11, 126:3 W
m - 3
PD = 11:1 kW
m - 3

or in short:
2
PD ðW
m - 3 Þ = 55:61
10 - 12 jEj ðV
m - 1 Þ
f ðHzÞ
ε00

PD ðW
m - 3 Þ = 55:61
10 - 12 ð100Þ2
2
109
10 = 55:61
10 - 12

2
1014 = 11, 122

The field strength |E| used in this example is the field strength within the
absorbent material. Since this is hardly known in most cases, the performance of a
microwave oven is most often determined calorimetrically. For this purpose, the
temperature change ΔT of a known material is measured, which occurs during a
microwave treatment with a defined duration Δt. Then it is

ρ
cp
ΔT
PD = ð11:41Þ
Δt

PD dissipated power in W
cp specific heat capacity in J
kg-1
K-1
ρ density in kg
m-3
ΔT temperature increase in K
Δt duration in s
470 11 Electromagnetic Properties

Example
Water with a density of 1000 kg
m-3 and a specific heat capacity of
4.2 kJ
kg-1
K-1 shows a change in temperature from 10 °C to 40 °C within
1 min in a microwave oven.
The dissipated power is

ρ
cp
ΔT
PD =
Δt
1000 kg
m - 3
4:2 kJ
kg - 1
K - 1
30 K
PD = = 2100 kJ
s - 1
m - 3
60 s
= 2100 kW
m - 3

i.e.

PD = 2:1 kW per liter

In conventional heating processes, heat first must be transported to the outside of

a food. The transport takes place by means of heat transfer through a wall material
(e.g., cooking pot) or convective heat transfer from a heat transfer medium (e.g., hot
air). From the outside of the food, the heat enters the interior of the food through heat
conduction or convection (see Chap. 8, Thermal Properties). With microwave
heating, the heat transport to the outside of the food is omitted. In addition, the
heat is generated within the internal mass of the food and does not enter it from the
outer surface only. However, it is a mistake to believe that microwave heating begins
at the core of the food. The microwaves also penetrate the food from the outside and
have defined penetration depths depending on the material.

11.3.2 Penetration Depth of Microwaves

When microwaves penetrate a material and its polarizable dipoles absorb energy, the
energy of the microwave increasingly decreases as it propagates into that material.
The higher the absorption coefficient, the stronger the attenuation of the penetrating
microwave. In highly absorbent materials such as liquid water, microwaves lose a
large part of their energy after just a few centimeters.
The distance d, after which the microwave power has decreased by absorption to 1e
of its original power, is called the penetration depth z of the microwave [8]. The
microwave power decreases exponentially [9].

P = P0
e - d
z
ð11:42Þ
11.3 Microwaves 471

The penetration depth is

1
2

c 2
z=
ð11:43Þ
2
π
f ε00 2
ε0 1þ ε0 -1

z penetration depth in m
P power in W
E electric field strength in V
m-1
f frequency in Hz
ε″ imaginary part of permittivity (loss factor)
ε′ real part of permittivity
c propagation speed in m
s-1

Example
Consider a location inside the food at a distance d from the surface. In this
example, the distance d should have exactly the calculated value of z. Then the
microwave power at place d is

P = P0
e - d
z

P = P0
e - 1

or
P 1
= = 0:368
P0 e
The microwave power at location d is only 36.8% of its original power. The
other 63.2% have already been absorbed by the previously irradiated material.

An exponential decrease in microwave intensity along the absorption pathway

means that the inner areas of a food receive a lower intensity than exterior areas.
During microwave heating, the rate of temperatures increase depends on the loca-
tion. An inhomogeneous temperature distribution leads to spatially inhomogeneous
physical properties (e.g., permittivity), which makes it difficult to model microwave
heating [10].
472 11 Electromagnetic Properties

→ Attention
High microwave absorption results in a low penetration depth, while lower
absorption results in a high penetration depth. This seemingly paradoxical
connection is clarified when we recall the definition of the depth of penetra-
tion: It is the depth where the power has fallen to 36.8%.

Example
What is the penetration depth of 2450 MHz microwaves in a raw potato?
The permittivity data of the potato are ε′ = 64 and ε″ = 15.
Solution:
1
2

c 2
z=

2
π
f ε00 2
ε0 1þ ε0 -1

1
2

3
108 m
s‐1 2
z=
= 0:019 m
1:0739
2
π
2450
106 s‐1 15 2
64 1þ 64 -1

= 0:021 m

z = 2:1 cm

While the penetration depth in a raw potato is 2 cm (see example), it is several

meters for materials with a low microwave absorption such as glass or porcelain.
This means that the microwaves show no noticeable loss of performance during the
passage of a plate made of glass or porcelain. Also, solid water (ice) has low
microwave absorption, i.e., a large penetration depth.

→ Attention
Sometimes, instead of the physical penetration depth where the power has
fallen to 36.8%, the penetration depth is given as the distance where the power
has fallen to 50%, the so-called half-power depth. The conversion is
z50% = 0.347
z36.8%
11.3 Microwaves 473

11.3.3 Microwave Heating

The main differences between conventional food heating and microwave heating of
food [10, 11] have already been listed in Sect. 11.3.1. The advantage of heating with
microwaves is the high heating speed. Because there is no external heat transfer
resistance, microwaves can be used for very fast heating processes. The microwave
power can be controlled without delay. Foods with high moisture content have a
high loss factor and, as a rule, are well suited for microwave heating. While liquid
water shows high values for loss factor ε″ (Fig. 11.4), solid water (ice) shows very
low values, and thus, low microwave absorption (Table 11.3).
The composition of food is often inhomogeneous; think of pizza, quiche, chunky
stews, and ready meals from several foods. This means that there is a complicated
spatial distribution of electrical permittivity, which leads to an inhomogeneous
temperature distribution throughout the food during microwave heating. The inho-
mogeneity of the material leads to a spatially different heat gain. The corresponding
inhomogeneous temperature distribution in the material can further be changed over
time by different thermal conductivities and heat capacities of the food components.
Because of the limited penetration depth of microwaves, foods with large
dimensions have a lower core temperature than small particles of the same material
[12]. It is often necessary to downregulate the microwave power or to work
intermittently (“on/off”) to avoid selective overheating of the food. Then a tempera-
ture equalization can be reached within the food by heat conduction.
The electrolyte content of an aqueous solution affects the value of the loss factor
ε″. Therefore, the electrolyte concentration is a factor that, in addition to the water
content, affects microwave absorption. Materials that have a high porosity, on the
other hand, can act like insulators in thermal processes. Gas-filled pores have a very
low thermal conductivity (see Chap. 8, Thermal Properties) and can therefore also
impede heat exchange between hot areas and less hot areas in the microwaved food.
Glass, ceramics, and porcelain do not show microwave absorption due to their low
loss factors ε″. The same applies to microwave-grade packaging. However, it is
possible to equip packaging with materials that show high absorption in the micro-
wave field to reach higher temperatures. These materials are called susceptors and
have the purpose of bringing the food into contact with a hot surface during
microwave treatment to support a thawing process or to cause browning reactions.
Instead of metallic susceptors, there are also ways to make susceptor materials from
biological materials [12, 14–16].

Table 11.3 Dielectric properties of solid water [13]

ε′ ε″ tan δ
Ice 3.2 0.0029 0.0009
474 11 Electromagnetic Properties

11.4 Terahertz Waves

Electromagnetic waves with frequencies in the range between 100 GHz and 10 THz
are referred to as terahertz waves. The wavelength range is between 3 mm and
10 μm. It thus covers the area of the far infrared (FIR) and extends into the area in the
mid-infrared (MIR), cf. Table 11.4. For this reason, terahertz radiation is sometimes
referred to as far-infrared radiation (FIR).
Wavelengths in the range of λ = 0.78...1000 μm are called infrared radiation (IR).
In IR spectroscopy, in addition to the wavelength, the reciprocal size or so-called
wavenumber is often used. Although the SI unit of wave number is m-1, the unit
cm-1 is often used. Table 11.4 illustrates the subdivision of the IR range based on
wavelength, wave number, and frequency.
It is

1
ν= ð11:44Þ
λ

λ wavelength in m
ν wave number in m-1

Definition
Terahertz waves are electromagnetic waves in the frequency range between
microwave and infrared radiation.

Since terahertz radiation can excite molecular rotations, terahertz radiation can be
used for molecular spectroscopy. The absorption of terahertz radiation is strongly
substance specific. The absorption coefficient of polar substances such as water is
high; less polar materials such as paper, cardboard, wood, plastics, and textiles show
lower absorption but can be distinguished. Metals reflect terahertz radiation. For this
reason, terahertz radiation is suitable for the inspection of packaged goods. The
energy of terahertz radiation is much lower than, for example, that of X-rays (see
Chap. 13). So, terahertz inspectors can also be used for personal control at airports as
an example. The penetration depth of terahertz radiation in textiles and plastics is
very high, while they hardly penetrate metals and water. Since terahertz absorption
depends on the chemical bonds of the inspected materials, high-contrast images of
packaged or hidden materials can be generated and it is possible to chemically

Table 11.4 Subdivision of IR wavelength ranges according to ISO 20473 [17]

Wavelength Wave number Frequency
Designation λ (μm) ν (cm-1) f (Hz)
Near infrared NIR 0.78–3 3333–12,820 1∙1014–3.8∙1014
Mid-infrared MIR 3–50 200–3333 6∙1012–1∙1014
Far infrared FIR 50–1000 10–200 3∙1011–6∙1012
11.5 NIR 475

identify organic solids in parallel. [18]. Terahertz imaging offers numerous

possibilities for non-invasive inspection with simultaneous material identification.
Compared to X-ray inspection, the distinguishability of different organic materials in
terahertz inspection is high. Since terahertz waves are reflected by metals, the
applications only apply to cases where there is no metallic shielding. Due to the
high absorptiveness of liquid water, terahertz waves are absorbed by aqueous
biological materials practically at the surface [19, 20]. In this way, the water
distribution of biological surfaces can be imaged or water in thin samples of dried
materials can be determined. Pulsed terahertz waves can be used to determine the
properties of thin dielectric films in a very short time. As an example, it is possible to
build sensors with semiconductor chips that are protein-sensitive or gene-sensitive,
whose signal is derived from the terahertz absorptiveness of the chip [21].

11.5 NIR

Electromagnetic waves with frequencies in the NIR range (Table 11.4) can be
absorbed by molecular bonds and are therefore suitable for spectroscopy of
materials. For this purpose, the absorption spectrum is measured, which is the
graphical application of the absorption over the frequency or the wavelength or the
wave number. The absorption spectrum provides information on which molecular
bonds occur frequently and less frequently in the material and thus allows
conclusions to be drawn about the composition of the material. In addition to
absorption spectra, transmission spectra are also common.
Let us consider the water molecule as an example of an infrared-active molecule
(Table 11.5). The stretch oscillations and the scissoring oscillation of the O–H bonds
in the H2O-molecule can be excited by specific IR frequencies. If these frequencies
or wave numbers occur in the absorption spectrum of a mixture of substances, this is
an indication of the presence of water in the sample.

Table 11.5 Examples of fundamental oscillations of the H2O molecule

Designation ν (cm-1) (absorption)
Symmetric stretching 3652

Asymmetric stretching 3756

Scissoring 1596
476 11 Electromagnetic Properties

Table 11.6 Fundamental Designation Frequency Wavelength

oscillations and harmonics
Fundamental oscillation f λ
1. Harmonic 2
f λ
2
2. Harmonic 3
f λ
3
etc.

The wave numbers given in Table 11.5 belong to the so-called fundamental
oscillations of the H2O molecule. In addition to these fundamental oscillations,
molecules can also perform oscillations with the double, triple, etc. frequency,
so-called harmonics. Table 11.6 shows the systematics.

Definition
Harmonics have an integer multiple frequency of a fundamental oscillation. In
analogy to acoustics, harmonics are sometimes referred to as overtones.

We have looked at H2O molecules up to this point and see that despite the simple
molecular structure, several absorption frequencies exist. Larger molecules have
much more absorption frequencies. The coupling of adjacent oscillations leads to
additional so-called combination vibrations. Although NIR spectra have many peaks
that are hardly distinguishable peaks, there are absorption frequencies that are
characteristic for groups of substances; see Table 11.7. Figure 11.5 illustrates the
formation of typical bands in a spectrum. Typical absorption ranges of food
ingredients are listed in Table 11.8.

Table 11.7 Absorption bands (middle MIR) of molecular groups [1]

Oscillation ν (cm-1) (absorption)
Alkanes – CH stretching and scissoring 2800–3000
– CH2 stretching and scissoring 1420–1470
– CH3 stretching and scissoring 1340–1380
Alkenes – CH olefin stretching 3000–3100
Alkynes – CH acetylene stretching 3300
Aromatic compounds – CH aromatic stretching 3000–3100
– C=C – stretching 1600
– OH stretching 3200–3600
– OH scissoring 1300–1500
C–O stretching 1000–1200
Ethers C–O asymmetric stretching 1000–1220
Amines – NH primary and secondary amine stretching 3300–3500
Aldehydes, ketones – C=O stretching 1700–1735
– OH (doublet) 2700–2850
Carboxylic acids – C=O stretching 1720–1740
Amides – C=O stretching 1640–1670
– NH stretching 3100–3500
– NH scissoring 1550–1640
11.5 NIR 477

Fig. 11.5 IR spectrum of 1.O

cellulose (bottom) caused the 2.O }
} C
}
absorption of chemical bonds
(top), C: fundamental H2O H2O H2O
oscillation, 1.O, 2.O: first and
ROH ROH
second harmonics, from [22] RCOOH
RNOH RNH2

CH CH CH
CH2 CH2 CH2
CH3 CH3 CH3

1100 1500 1900 2300

O / nm
0.6
0.5
0.4
0.3
1
log 0.2
R
0.1
0
-0.1
1100 1500 1900 2300
O / nm

Table 11.8 Absorption ranges (NIR) of food ingredients [1]

Component Oscillation λ (nm)
Water – OH stretching/combination 1920–1950
– OH stretching 1400–1450
Proteins, peptides – NH deformation 1560–1670
2080–2220
Fats – CH stretching 2300–2350
– CH2 and – CH3 stretching 1680–1760
Carbohydrates C-O and O-H stretching/combination 2060–2150

Definition
NIRS stands for NIR spectroscopy, i.e., near-infrared spectroscopy.

The great advantage of NIR spectroscopy is that materials can be examined

without time-consuming sample preparation; even measurements in the running
478 11 Electromagnetic Properties

process are possible. For powders, however, physical parameters such as particle
size, crystallinity, and bulk density slightly influence the spectrum. This must be
considered during calibration. Conversely, this offers the possibility of including
these physical quantities in NIR quality control [23].

Definition
Hyperspectral imaging (HSI) is the integration of spectroscopy and imaging
techniques. While in a conventional image a single property, such as color, is
assigned to each pixel, in HSI there is a spectrum for each pixel [24]. This can
be an NIR spectrum for each image pixel.

Further Reading

Dielectric polarization
Moisture determination in agricultural materials using dielectric properties [2, 5, 25]
Meat: rapid moisture determination using dielectric measurement [26]
Dielectric properties of pharmaceutical powders [27]
Honey: water content measurement using dielectric properties [28]

Microwaves
Potato chips produced by microwave multiflash drying [29]
Edible microwave susceptor made by alginate and salt [14]
Microwave sterilization: comparison to conventional sterilization [3]
Monitoring of food drying and freezing microwaves [30]
More homogeneous temperature distribution using microwave guides [31]
Computational Modeling of heat transport during microwave treatment [32]
Microwave-assisted freeze drying [33]
Relationship between dielectric response and water activity [34]
Radiofrequency thawing of frozen minced fish [35]
Rice model food system: dielectric properties relevant to microwave sterilization [36]

Terahertz waves
Tomatoes: pesticide residue by THz-spectroscopy [37]
Antibiotic residues by THz-spectroscopy [38]
Chocolate: foreign body detection pulsed THz spectroscopy [39]
Sausage: detection of foreign materials by THz spectroscopic imaging [40]
Cocaine detection by THz-spectroscopy [18]
Terahertz spectroscopy: agricultural applications [41, 42]

(continued)
References 479

Hyperspectral imaging on meat [43]

Terahertz detection of transgenic food [44]
Terahertz sensing for food and water security [45]
Identification of food by measurement of complex permittivity [46]
Terahertz phase imaging and biomedical applications [47]

NIR
Quality of fruit and vegetable by NIR spectroscopy [48, 49]
Linking flavour sensory and NIRS [50]
Quality of oilseeds and edible oils by NIRS [51]
Quality of cocoa bean and cocoa bean products by NIRS [52]
Beer: vis/NIR spectroscopy during fermentation [53]
NIR-fluid bed process monitoring [54]
Pharmaceutical technology support by NIR [23, 55]
Dairy: NIR process control [56]
Nuts: characterization by IR-spectroscopy [57]
Mushrooms: authentication and quality analysis by NIRS [58]
Maize flour: secondary biochemical components by NIR [59]
Hyperspectral imaging of oil, milk, yogurt [60]

Summary
The electrical polarizability of molecules is the cause of the ability to absorb
electromagnetic radiation of certain wavelengths. Polarizability is explained
based on the model of the Debye dipole in simple words and described with
the complex permittivity. Applications such as heating with the help of
microwaves and material characterization with terahertz radiation and NIR
spectroscopy illustrate the interaction between electromagnetic radiation and
biological materials such as food. At the end of the chapter, examples from the
literature are listed, which can be used for further studies and as suggestions
for one’s own scientific work.

References
1. Atkins PW, De Paula J (2011) Physical chemistry for the life sciences, 2nd edn. W.H. Freeman,
Oxford
2. Udo K, Christof H (2010) Electromagnetic techniques for moisture content determination of
materials. Meas Sci Technol 21(8):82001. https://doi.org/10.1088/0957-0233/21/8/082001
480 11 Electromagnetic Properties

3. Soni A, Smith J, Thompson A, Brightwell G (2020) Microwave-induced thermal sterilization –

a review on history, technical progress, advantages and challenges as compared to the conven-
tional methods. Trends Food Sci Technol 97:433. https://doi.org/10.1016/j.tifs.2020.01.030
4. Meda V, Orsat V, Raghavan V (2017) Microwave heating and the dielectric properties of foods.
Woodhead Publishing, pp 23–43. https://doi.org/10.1016/b978-0-08-100528-6.00002-4
5. Andryieuski A, Kuznetsova SM, Zhukovsky SV, Kivshar YS, Lavrinenko AV (2015) Water:
promising opportunities for tunable all-dielectric electromagnetic metamaterials. Sci Rep 5:
13535. https://doi.org/10.1038/srep13535
6. Kaatze U (1989) Complex permittivity of water as a function of frequency and temperature. J
Chem Eng Data 34(4):371. https://doi.org/10.1021/je00058a001
7. Tiwari G, Wang S, Tang J, Birla SL (2011) Analysis of radio frequency (RF) power distribution
in dry food materials. J Food Eng 104(4):548. https://doi.org/10.1016/j.jfoodeng.2011.01.015
8. Schubert H, Regie M (2006) Novel and traditional microwave applications in the food
industry. In: Willert-Porada M (ed) Advances in microwave and radio frequency processing.
Springer, Berlin, pp 259–270
9. Hippel vAR (1995) Dielectric and waves. Artech House, Boston
10. Chandrasekaran S, Ramanathan S, Basak T (2013) Microwave food processing – a review.
Food Res Int 52(1):243. https://doi.org/10.1016/j.foodres.2013.02.033
11. Guo QS, Sun DW, Cheng JH, Han Z (2017) Microwave processing techniques and their recent
applications in the food industry. Trends Food Sci Technol 67:236. https://doi.org/10.1016/j.
tifs.2017.07.007
12. Raaholt BW (2020) Influence of food geometry and dielectric properties on heating
performance. In: Erle U, Pesheck P, Lorence M (eds) Development of packaging and products
for use in microwave ovens. Woodhead Publishing, pp 73–93. https://doi.org/10.1016/B978-0-
08-102713-4.00002-5
13. Matsuoka T, Fujita S, Mae S (1996) Effect of temperature on dielectric properties of ice in the
range 5–39 GHz. J Appl Phys 80(10):5884. https://doi.org/10.1063/1.363582
14. Albert A, Salvador A, Fiszman SM (2012) A film of alginate plus salt as an edible susceptor in
microwaveable food. Food Hydrocoll 27(2):421. https://doi.org/10.1016/j.foodhyd.2011.
11.005
15. Risman PO (2020) Measurements of dielectric properties of foods and associated materials. In:
Erle U, Pesheck P, Lorence M (eds) Development of packaging and products for use in
microwave ovens. Woodhead Publishing, pp 201–223. https://doi.org/10.1016/B978-0-08-
102713-4.00005-0
16. Chen FY, Warning AD, Datta AK, Chen X (2017) Susceptors in microwave cavity heating:
modeling and experimentation with a frozen pie. J Food Eng 195:191. https://doi.org/10.1016/j.
jfoodeng.2016.09.018
17. ISO 20473 (2007) Optik und Photonik - Wellenlängenbereiche. Beuth, Berlin. https://doi.org/
10.31030/3123472
18. Dobroiu A, Otani C, Kawase K (2006) Terahertz-wave sources and imaging applications. Meas
Sci Technol 17(11):R161. https://doi.org/10.1088/0957-0233/17/11/r01
19. Wang C, Qin J, Xu W, Chen M, Xie L, Ying Y (2018) Terahertz imaging applications in
agriculture and food engineering: a review. Trans ASABE 61(2):411. https://doi.org/10.13031/
trans.12201
20. Ok G, Shin HJ, Lim M-C, Choi S-W (2019) Large-scan-area sub-terahertz imaging system for
nondestructive food quality inspection. Food Control 96:383. https://doi.org/10.1016/j.
foodcont.2018.09.035
21. Mickan S, Lee K-S, Lu T-M, Munch J, Abbott D, Zhang XC (2013) Double modulated
differential THz-TDS for thin film dielectric characterization. Microelectron J 33:1033.
https://doi.org/10.1016/S0026-2692(02)00108-8
22. Berntsson O (2001) Characterization and application of near infrared reflection spectroscopy for
quantitative process analysis of powder mixtures. Doctoral thesis, KTH, Stockholm
References 481

23. Jamrógiewicz M (2012) Application of the near-infrared spectroscopy in the pharmaceutical

technology. J Pharm Biomed Anal 66:1. https://doi.org/10.1016/j.jpba.2012.03.009
24. Park B, Lu R (2015) Hyperspectral imaging technology in food and agriculture. https://doi.org/
10.1007/978-1-4939-2836-1
25. Kozhevnikov A (2010) Wideband radio-frequency device for measurements of dielectric
properties of small volumes of liquids. Meas Sci Technol 21(4):043001. https://doi.org/10.
1088/0957-0233/21/4/043001
26. Wang J, Fan L, Zhou Q, Li J, Zhao P, Wang Z, Zhang H, Yan S, Huang L (2018) Rapid
determination of meat moisture content using radio-frequency dielectric measurement. IEEE
Access 6:51384. https://doi.org/10.1109/access.2018.2864308
27. Magee TRA, McMinn WAM, Farrell G, Topley L, Al-Degs YS, Walker GM, Khraisheh M
(2013) Moisture and temperature dependence of the dielectric properties of pharmaceutical
powders. J Therm Anal Calorim 111(3):2157. https://doi.org/10.1007/s10973-012-2739-y
28. Li Z, Haigh A, Soutis C, Gibson A, Sloan R (2017) Evaluation of water content in honey using
microwave transmission line technique. J Food Eng 215(Suppl C):113. https://doi.org/10.1016/
j.jfoodeng.2017.07.009
29. Barreto IMA, Tribuzi G, Marsaioli Junior A, Carciofi BAM, Laurindo JB (2019) Oil–free
potato chips produced by microwave multiflash drying. J Food Eng 261:133. https://doi.org/10.
1016/j.jfoodeng.2019.05.033
30. Pandey G, Vandermeiren W, Dimiccoli L, Stiens J (2018) Contactless monitoring of food
drying and freezing processes with millimeter waves. J Food Eng 226:1. https://doi.org/10.
1016/j.jfoodeng.2018.01.003
31. Zhu HK, Li D, Li SJ, Wang SJ (2017) A novel method to improve heating uniformity in
mid-high moisture potato starch with radio frequency assisted treatment. J Food Eng 206:23.
https://doi.org/10.1016/j.jfoodeng.2017.03.001
32. Rakesh V, Seo Y, Datta AK, McCarthy KL, McCarthy MJ (2010) Heat transfer during
microwave combination heating: computational modeling and MRI experiments. AICHE J
56(9):2468. https://doi.org/10.1002/aic.12162
33. Duan X, Zhang M, Mujumdar AS, Wang R (2010) Trends in microwave-assisted freeze drying
of foods. Dry Technol 28(4):444. https://doi.org/10.1080/07373931003609666
34. Renshaw RC, Dimitrakis GA, Robinson JP, Kingman SW (2019) The relationship of dielectric
response and water activity in food. J Food Eng 244:80. https://doi.org/10.1016/j.jfoodeng.
2018.08.037
35. Yang H, Chen Q, Cao H, Fan D, Huang J, Zhao J, Yan B, Zhou W, Zhang W, Zhang H (2019)
Radiofrequency thawing of frozen minced fish based on the dielectric response mechanism.
Innovative Food Sci Emerg Technol 52:80. https://doi.org/10.1016/j.ifset.2018.10.013
36. Auksornsri T, Tang J, Tang Z, Lin H, Songsermpong S (2018) Dielectric properties of rice
model food systems relevant to microwave sterilization process. Innovative Food Sci Emerg
Technol 45:98. https://doi.org/10.1016/j.ifset.2017.09.002
37. Suzuki T, Ogawa Y, Kondo N (2011) Characterization of pesticide residue, cis-permethrin by
terahertz spectroscopy. Eng Agric Environ Food 4(4):90. https://doi.org/10.1016/S1881-8366
(11)80007-8
38. Redo-Sanchez A, Salvatella G, Galceran R, Roldós E, García-Reguero J-A, Castellari M,
Tejada J (2011) Assessment of terahertz spectroscopy to detect antibiotic residues in food and
feed matrices. Analyst 136(8):1733. https://doi.org/10.1039/C0AN01016B
39. Jördens C, Koch M (2008) Detection of foreign bodies in chocolate with pulsed terahertz
spectroscopy. Opt Eng 47. https://doi.org/10.1117/1.2896597
40. Wang C, Zhou R, Huang Y, Xie L, Ying Y (2019) Terahertz spectroscopic imaging with
discriminant analysis for detecting foreign materials among sausages. Food Control 97:100.
https://doi.org/10.1016/j.foodcont.2018.10.024
41. Afsah-Hejri L, Akbari E, Toudeshki A, Homayouni T, Alizadeh A, Ehsani R (2020) Terahertz
spectroscopy and imaging: a review on agricultural applications. Comput Electron Agric 177:
105628. https://doi.org/10.1016/j.compag.2020.105628
482 11 Electromagnetic Properties

42. Li B, Zhang D, Shen Y (2020) Study on terahertz spectrum analysis and recognition modeling
of common agricultural diseases. Spectrochim Acta A Mol Biomol Spectrosc 243:118820.
https://doi.org/10.1016/j.saa.2020.118820
43. Feng CH, Makino Y, Oshita S, Martin JFG (2018) Hyperspectral imaging and multispectral
imaging as the novel techniques for detecting defects in raw and processed meat products:
current state-of-the-art research advances. Food Control 84:165. https://doi.org/10.1016/j.
foodcont.2017.07.013
44. Liu J, Li Z (2014) The terahertz spectrum detection of transgenic food. Optik 125(23):6867.
https://doi.org/10.1016/j.ijleo.2014.08.114
45. Ren A, Zahid A, Fan D, Yang X, Imran MA, Alomainy A, Abbasi QH (2019) State-of-the-art in
terahertz sensing for food and water security – a comprehensive review. Trends Food Sci
Technol 85:241. https://doi.org/10.1016/j.tifs.2019.01.019
46. Shin HJ, Oh SJ, Lim M-C, Choi S-W, Ok G (2018) Dielectric traces of food materials in the
terahertz region. Infrared Phys Technol 92:128. https://doi.org/10.1016/j.infrared.2018.05.022
47. Wan M, Healy JJ, Sheridan JT (2020) Terahertz phase imaging and biomedical applications.
Opt Laser Technol 122:105859. https://doi.org/10.1016/j.optlastec.2019.105859
48. Nicolaï BM, Beullens K, Bobelyn E, Peirs A, Saeys W, Theron KI, Lammertyn J (2007)
Nondestructive measurement of fruit and vegetable quality by means of NIR spectroscopy: a
review. Postharvest Biol Technol 46(2):99. https://doi.org/10.1016/j.postharvbio.2007.06.024
49. Walsh KB, McGlone VA, Han DH (2020) The uses of near infra-red spectroscopy in posthar-
vest decision support: a review. Postharvest Biol Technol 163:111139. https://doi.org/10.1016/
j.postharvbio.2020.111139
50. Chapman J, Elbourne A, Truong VK, Newman L, Gangadoo S, Rajapaksha Pathirannahalage P,
Cheeseman S, Cozzolino D (2019) Sensomics – from conventional to functional NIR
spectroscopy – Shining light over the aroma and taste of foods. Trends Food Sci Technol 91:
274. https://doi.org/10.1016/j.tifs.2019.07.013
51. Li X, Zhang L, Zhang Y, Wang D, Wang X, Yu L, Zhang W, Li P (2020) Review of NIR
spectroscopy methods for nondestructive quality analysis of oilseeds and edible oils. Trends
Food Sci Technol 101:172. https://doi.org/10.1016/j.tifs.2020.05.002
52. Teye E, Anyidoho E, Agbemafle R, Sam-Amoah LK, Elliott C (2020) Cocoa bean and cocoa
bean products quality evaluation by NIR spectroscopy and chemometrics: a review. Infrared
Phys Technol 104:103127. https://doi.org/10.1016/j.infrared.2019.103127
53. Giovenzana V, Beghi R, Guidetti R (2014) Rapid evaluation of craft beer quality during
fermentation process by vis/NIR spectroscopy. J Food Eng 142:80. https://doi.org/10.1016/j.
jfoodeng.2014.06.017
54. Liu R, Li L, Yin W, Xu D, Zang H (2017) Near-infrared spectroscopy monitoring and control of
the fluidized bed granulation and coating processes—a review. Int J Pharm 530(1):308. https://
doi.org/10.1016/j.ijpharm.2017.07.051
55. Chavan RB, Bhargavi N, Lodagekar A, Shastri NR (2017) Near infra red spectroscopy: a tool
for solid state characterization. Drug Discov Today 22(12):1835. https://doi.org/10.1016/j.
drudis.2017.09.002
56. Pu Y-Y, O’Donnell C, Tobin JT, O’Shea N (2020) Review of near-infrared spectroscopy as a
process analytical technology for real-time product monitoring in dairy processing. Int Dairy J
103:104623. https://doi.org/10.1016/j.idairyj.2019.104623
57. Teixeira AM, Sousa C (2019) A review on the application of vibrational spectroscopy to the
chemistry of nuts. Food Chem 277:713. https://doi.org/10.1016/j.foodchem.2018.11.030
58. Meenu M, Xu B (2019) Application of vibrational spectroscopy for classification, authentica-
tion and quality analysis of mushroom: a concise review. Food Chem 289:545. https://doi.org/
10.1016/j.foodchem.2019.03.091
59. Kahrıman F, Onaç İ, Öner F, Mert F, Egesel CÖ (2020) Analysis of secondary biochemical
components in maize flour samples by NIR (near infrared reflectance) spectroscopy. J Food
Meas Charact 14(4):2320. https://doi.org/10.1007/s11694-020-00479-0
60. Baiano A (2017) Applications of hyperspectral imaging for quality assessment of liquid based
and semi-liquid food products: a review. J Food Eng 214:10. https://doi.org/10.1016/j.jfoodeng.
2017.06.012
Optical Properties
12

Light is electromagnetic radiation that can be detected by the human eye. This part of
the electromagnetic spectrum with wavelengths between 780 and 400 nm is also known
as the optical range (Table 12.1). Although the optical properties of food are thus only a
part of the electromagnetic properties (Chap. 11), a separate chapter is devoted to the
optical properties. It is mainly about the color of food, which can be quantified with the
help of physical measurements. It is also about the polarization of light, diffraction, and
refraction, which are used in quality control. In Chap. 16 (on-line sensors), some optical
sensors for process analysis are presented. Applications of microwaves, terahertz
waves, infrared and near-infrared have already been dealt with in Chap. 11, ultraviolet
radiation follows in Chap. 13 (UV and X-ray radiation).
The product of wavelength and frequency provides the propagation speed of the
wave, in this case the speed of light:
c=λ
f ð12:1Þ

c propagation speed in m
s-1

λ wavelength in nm
f frequency in s-1

When electromagnetic radiation such as light hits a material, phenomena such as

refraction, polarization, reflection, absorption, transmission, scattering, diffraction
occur. We want to go through them one by one.

12.1 Refraction

Refraction is the change in direction of wave propagation when it hits (non-vertical)

an interface between media that have different propagation speeds for this wave.
According to Maxwell, the propagation speed of an electromagnetic wave is
given by

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 483
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_12
484 12 Optical Properties

Table 12.1 Electromagnetic radiation: visible light lies between UV and IR

Frequency Vacuum wavelength Wavelength
(Hz) Electromagnetic radiation (m) comparator
3∙1024 Cosmic radiation 10-16
3∙1023 10-15 Atomic nucleus
3∙1022 γ-Radiation 10-14
3∙1021 10-13
3∙1020 X-rays 10-12
3∙1019 10-11
3∙1018 10-10 H-atom
3∙1017 10-9
3∙1016 Ultraviolet radiation (UV) 10-8
3∙1015 10-7 Viruses
3∙1014 Light 10-6 Bacteria
3∙1013 Infrared radiation (IR) 10-5
3∙1012 10-4
3∙1011 Microwaves 10-3
3∙1010 High-frequency waves 10-2 Chip
3∙109 Very high frequency 10-1
(VHF)
3∙108 100
Metric waves
3∙107 Short waves 101 House
3∙106 Long waves 102 Skyscraper
3∙105 103
3∙104 104 Mount Everest
3∙103 Low frequency 105
3∙102 Alternating current 106
3∙101 Audio frequencies 107 Earth diameter
3∙100 108

1
c= ð12:2Þ
ε0
ε
μ 0
μ

in a vacuum it is

1
c0 = ð12:3Þ
ε0
μ 0

The ratio n of the light propagation speed c in a material with μ = 1 to the vacuum
speed of light c0 thus is

c ε0
ε
μ 0 p
n= = = ε ð12:4Þ
c0 ε0
μ0

that means
12.1 Refraction 485

Fig. 12.1 The angle of

incidence α of a beam and the
angle of the reflected beam are
the same α = α′. If the speed
of light in material 1 and
material 2 is different, then
β≠α

D D
n1
n2

p
n= ε ð12:5Þ

ε0 electric field constant: 8.854
10-12 C
V-1
m-1

ε relative permittivity
μ magnetic permeability
μ0 magnetic field constant: μ0 = 4π
10-7V
s
A-1
m-1

The change in direction of propagation of a light wave as a result of the change in

the propagation speed is referred to as refraction. The ratio n is called refraction
index (Fig. 12.1).
Let’s look at a beam of light (Fig. 12.2), which passes through the interface of
material 1 and material 2. Part of the light is reflected at the interface, the rest of the
light enters material 2.
According to the Huygens’ principle, all illuminated points at the interface are
starting points of spherical waves that propagate in all directions, including into
material 2. In material 1, the light beam has the propagation speed c1, i.e. in the time
span Δt it travels the distance BD. If the waves starting at these starting points in
material 2 have the propagation speed c2, then the light beam travels the path AC
during the same period Δt (Fig. 12.2). If the propagation speeds c1 and c2 are
different, then the distances BD and AC are of different lengths, the result is a
bend in the beam path. This bending of a light beam is called refraction. If the
propagation speed in material 2 is lower than in material 1 as in Fig. 12.2, then
refraction takes place toward the line mutually perpendicular to the interfaces, i.e. the
angle of refraction β is smaller than the angle of incidence α. With
486 12 Optical Properties

Fig. 12.2 Refraction as a

result of different propagation
speeds in materials 1 and 2

B
D

A D n1
n2

C
E

BD = Δt
c1
AC = Δt
c2

and

BD sin α
=
AC sin β
we have

sin α c1
=
sin β c2
Using the ratio of propagation speed in the material and propagation speed in
vacuum
cn
n= ð12:6Þ
c0
so

c1
n1 = ð12:7Þ
c0
and

c2
n2 = ð12:8Þ
c0
we get
12.1 Refraction 487

c0
sin α c1 n1 n2
= = == ð12:9Þ
sin β c2 c0
n2
n1

which is called Snellius’ law:

sin α n1
= ð12:10Þ
sin β n2
or

n2
sin β = sin α ð12:11Þ
n1
Snellius’ law describes the relation between refraction angle β and incidence
angle α.

c1 speed of light in material 1 in m
s-1

c2 speed of light in material 2 in m
s-1
c0 speed of light in vacuum in m
s-1
n1 refraction index material 1
n2 refraction index material 2
β angle of refraction
α angle of incidence

If the angle of incidence α is gradually increased, the angle of refraction β

increases according to Snellius’ law. When β reaches 90°, no more light enters
material 2. Now all incident light is reflected at the interface. This is called total
reflection. The angle of incidence at which this occurs is called critical angle of total
reflection.
For the sake of simplicity, we have talked about a beam of light up to this point.
On closer view, we see that the extent of the refraction also depends on the
wavelength or frequency of the radiation used. We had already learned about this
frequency dependence of physical quantities in the section on terahertz waves
(Chap. 11) as dispersion. An example of how the refractive index decreases with
increasing wavelength in the case of water is shown in Fig. 12.3. The cause of the
dispersion is that the wave propagation speed in any material is dependent on the
frequency.
The phenomenon of refraction is not limited to visible light. Even non-visible
electromagnetic waves such as infrared waves, microwaves, terahertz waves, etc. are
refracted at material interfaces if their propagation speeds in the materials differ on
both sides of the interface. This also applies to reflection and total reflection.
488 12 Optical Properties

Fig. 12.3 Dispersion of the

refractive index of water

Table 12.2 Values for the refractive indices of some materials [1]
Refractive index
Material (20 °C, 103 hPa, λ = 589nm)
Vacuum 1
Air 1.0003
CO2 1.0045
Ethanol 1.362
Water 1.333
20% (m/m) sucrose solution 1.364
KBr 1.46
Quartz glass 1.459
Boron-silicate glass 1.47
Polymethyl methacrylate (PMMA) 1.5
NaCl 1.89
AgBr 2.0
ZnSe 2.4
Diamond 2.414

12.1.1 Measurement of Refractive Index

In Table 12.2 examples of the refractive indices of different materials are listed.
Since the refractive index depends on the wavelength of the light used and the
temperature of the material, both parameters always must be specified. In the case of
gases, a dependence on pressure is added. From the table it can be seen that the
refractive indices differ from pure water and an aqueous sugar solution. This is
exploited by storing the concentration dependence of the refractive index of aqueous
sugar solutions in the form of tables or mathematical functions in order to obtain
information about the composition of the solution by measuring the refractive index.
A table for the sucrose-water system is listed in the appendix.
12.1 Refraction 489

Fig. 12.4 Refractive index

determination. 1 Glass
window, 2 sample, 3 cover,
4 incoming beam, 5 reflected 3
beam, 6 refracted beam,
7 total reflected beam, 6
8 detector

2
E 8
D 7

4
5

Such mathematical relationships can be integrated directly into measuring

instruments, so that many refractometers can display not only the refractive index
of a sample but also the concentration of ingredients in solutions such as sugar.
Refractive indices can be determined experimentally with the help of Snellius’
law. A metrological simplification consists in selecting 90° as the refraction angle
and only looking for the associated angle of incidence α, i.e. determining the critical
angle of the total reflection. In Fig. 12.4 a refractometer is shown schematically. The
liquid sample (2) lies on a glass window (1) which has a known refractive index n1,
the monochromatic light beam (4) penetrates the sample from below. The angle of
incidence α is varied until light escapes in the 90° direction, i.e. at the detector (8).
Then is:

sin α sin αG
= ð12:12Þ
sin β sin 90 °

sin αG n
= 2 ð12:13Þ
sin 90 ° n1

n2 = n1
sin αG ð12:14Þ

With the known refractive index n1 of the refractometer glass window, the
refractive index of the sample can be easily determined.

α angle of incidence
β angle of refraction
n2 refractive index of the sample
(continued)
490 12 Optical Properties

n1 refractive index of the refractometer glass

αG critical angle of refraction

Laboratory refractometers work with a few drops of sample fluid that are formed
into a thin film and deliver the refractive index in seconds. With continuous cuvettes
refractive indexes of the flowing liquids can be obtained on-line. In-line
refractometers can be used to control industrial processes.
By experimentally determining the refractive index, a material can be quickly
characterized. With the help of tabulated refractive index values, the sucrose con-
centration of foods such as beverages can be determined quickly and easily.

12.2 Polarization

Light waves emitted by a normal light source oscillate in all directions transversely
to the direction of wave propagation. With a polarization filter P1 in the beam path,
we can ensure that only light from a single vibration plane is transmitted. The light
now present is called polarized light. With a second polarization filter P2 in the beam
path, the polarization direction of the light can be detected: If the filter P2 is rotated
by 90° compared to P1, no light is transmitted at all. At smaller and larger angles,
some of the light passes through and when the filter P2 is in the polarization direction
of the light, maximum brightness occurs. The filter P2 is also called an analyzer and
P1 the polarizer.

Definition
Linearly polarized light refers to light whose electric field vector oscillates in
one spatial direction exclusively.

If we bring an optically transparent material into the beam between the polarizer
and the analyzer and find that the maximum brightness occurs at a different angle,
then we are dealing with the so-called optically active material. Optical activity
refers to materials that rotate the angle of polarized light by a certain angle. The
direction of this angular change (to the right or to the left) and the extent of the
change are characteristics of the substance, so optical rotation can be used in
analytics. The metrological determination of the optical rotation of a substance is
called polarimetry.
For this purpose, a solid sample or a cuvette with the aqueous solution of the
sample is placed in the beam path of a polarimeter and the angle of rotation of the
solution is determined (Fig. 12.5). Materials that change the angle of rotation to the
right (with a view into the light source) are referred to as positively rotating (+) or
right-turning materials, materials that change the angle of rotation to the left as
negatively rotating (-) or left-turning. The extent of the angular change is propor-
tional to the length of the radiated material and is specific to the substance under
12.2 Polarization 491

Fig. 12.5 Rotation of the polarization plane, schematic. In a linearly polarized wave that
propagates in the x direction, the electric vector oscillates in the z-direction (2), during the passage
of an optically active material (1), the direction of oscillation changes. The difference compared to
the initial direction is the angle of rotation α

observation. The material-specific rotation is called the specific rotation αS. In the
case of solutions, a proportionality to the concentration c of the solution is added.
The length of the cuvette filled with the solution then is the length l.
In the case of a solid sample:

α = αs
l ð12:15Þ

α rotation angle of the polarization plane in °

αs specific rotation in °
m-1
l thickness in m

In the case of a liquid sample, Biot’s law applies:

α = αs
l
c ð12:16Þ

There are different units for specific rotation, depending on which concentration
unit is used (%, mol
l-1, g
l-1, g
cm-3, kg
m-3, etc.). In SI units it is:

α angle of rotation of the polarization plane in rad

αs specific rotation in rad
m2
kg-1
l thickness (cuvette length) in m
c concentration in kg
m3

Specific rotation depends on the wavelength of the radiation used, this phenome-
non is called rotational dispersion. In addition there is a dependency on temperature
and, in the case of gaseous substances, also on pressure. Therefore, these parameters
must always be specified when measuring rotation angles. In tables, often values are
given for sodium light λ = 589 nm and ϑ = 20 ° C.
The optical activity of solids such as quartz is bound to the crystal structure, i.e. it
disappears when the material melts. Optical activity can also be caused by molecular
structure. For example, it occurs in carbon compounds containing at least one carbon
atom with four different substituents. Such a carbon atom is also called an
492 12 Optical Properties

Table 12.3 Stereoisomers (+)-Lactic acid (-)-Lactic acid

of lactic acid
D-lactic acid L-lactic acid
Dextrorotatory lactic acid Levorotatory lactic acid
Positive specific rotation Negative specific rotation

asymmetric carbon atom. In such a molecular structure, there are two different
possibilities that arrange four substituents around the carbon atom. This results in
two different molecules that behave like image and mirror image. These two
molecules are called enantiomers. Since they rotate polarized light in different
ways—one to the right and the other to the left—a distinction can be made between
D-enantiomer and L-enantiomer. D (Lat. dexter) means right, L (Lat. leavus) means
left. Table 12.3 lists the terms used for the enantiomers of lactic acid.

Example
We think both of our hands are identical. The number of fingers, number of
bones and joint is the same for every hand. This can be illustrated by imagining
you have both hands stretched out in front of you with palms facing down-
ward. The two hands are presumably identical to each other. But when you
place one hand on top of the other, you will see that the thumb and fingers do
not align with each other. The thumb of one hand is aligned with the small
finger of the other. You will realize that you cannot bring your two hands to
cover by no rotation. It is because they are structured like image and mirror
image. We have two equal hands, but concerning its structure they are
different. This shall help us to understand what is a molecule with an asym-
metric carbon atom. It also exists in two different structures like image and
mirror image. These two molecules are also called chiral molecules (the Greek
word χειρ means “hand”).

Bottom Line
Materials that rotate the oscillation plane of polarized light are optically active
substances. The measurement of the angle of rotation is used for identity and
purity testing as well as for concentration determinations.

A 1:1 mixture of the D-enantiomer and the L-enantiomer of a substance is called a

racemate. In a racemate, the rotations of the two components cancel each other out,
i.e. a racemate seems to be not optically active. By separating the enantiomers, two
optically active substances with opposing specific rotations can be obtained. The
value of the polarimetric angle of rotation of such a substance gives information
about the purity of the substance.
12.2 Polarization 493

In many substances, one of the two enantiomers is preferred in nature. The reason
for this may be, for example, that the chemical formation process runs via a natural
enzyme that is also chiral. Physiological processes are often adjusted to this,
i.e. often only one of the two enantiomers is effective, e.g. as a flavoring substance
or as a pharmacologically active ingredient, while the other enantiomer is ineffective
or even toxic. The L-amino acids asparagine, tryptophan, tyrosine, and isoleucine,
for example, have a bitter taste, while their D-forms are characterized as sweet
[2]. The racemates produced during the chemical synthesis of such substances
must therefore be separated at great expense or alternative manufacturing processes
with higher enantiomer purity must be found [3]. Polarimetry can be used as a
physical technique for determining purity. With the help of combinations of
polarimeters and CCD cameras, many samples can be examined simultaneously
(polarimetric imaging) [4].
Examples of food ingredients that have optically active isomers are carbohydrates
such as fructose, glucose, sucrose, galactose, mannose, lactose, maltose, dextrin,
carboxylic acids such as lactic acid, tartaric acid, malic acid, and amino acids such as
proline, alanine and flavorings such as pinene, limonene, and 3-3-methylthiobutanal
[2, 5–7].
We started the polarimetry section by describing linearly polarized light. In
linearly polarized light, the electric field vector of the monochromatic radiation
oscillates with constant frequency and amplitude unchanged in time in only one
plane. When the direction of the electric field vector changes in time and describes a
circular path around the direction of wave propagation, we speak of this as circularly
polarized light. In the general case, the amplitude can also be variable in time, with
the result that the electric field vector does not rotate circularly but elliptically around
the direction of wave propagation. We now have elliptically polarized light. Since
this effect is not limited to light, it is called elliptically polarized radiation.
One can imagine elliptically polarized radiation as the addition of two linearly
polarized waves perpendicular to each other, between which there is a phase
difference. At a phase difference of 90°, the ellipse becomes a circle, i.e. circularly
polarized radiation, at a phase difference of 0, the ellipse becomes a line, i.e. linearly
polarized radiation. There are two options for the direction of rotation of the electric
field vectors of elliptically polarized radiation: right-turning and left-turning.
With this understanding, the rotation of the polarization plane described above by
an enantiomer mixture in a cuvette is the splitting of the linearly polarized radiation
into two oppositely circularly polarized waves, which propagate with the same
frequency but different propagation velocities. The different propagation velocities
result from different interactions of the radiation with the two enantiomers. At the
end of the passage through the cuvette with the solution, both circularly polarized
waves add up again to form a linearly polarized wave, resulting in a changed angle of
rotation compared to the incident radiation [8].
494 12 Optical Properties

Definition
Ellipsometry is a non-destructive, noninvasive optical technique based on the
polarization of light, in which change in amplitude and direction of oscillation
of polarized light by a material is measured.

Concentrations of optically active substances thus can be determined

polarimetrically [9]. Polarimeters for sugar determination are also called
saccharimeters. Through the polarimetric observation of reactions or enzymatic
reactions, reaction kinetic data or enzyme activities can also be determined indirectly
from the change in sugar concentration [10, 11].
The appearance of a time-varying angle of rotation is called mutarotation. Here,
the angle of rotation of an aqueous solution changes until an enantiomer equilibrium
with an equilibrium rotation angle is set in the solution. The position of equilibrium
and the kinetics of equilibrium adjustment are dependent on temperature, concentra-
tion, and pH. The specific rotation itself can be dependent on concentration and even
the choice of solvent can affect the optical rotation.
If a layer of an optically inactive substance is placed between the polarizer and the
90° twisted analyzer, this arrangement is impermeable to light, it appears dark. When
the substance becomes optically active by applying an external electric field, light
can pass through the arrangement, it appears bright. In this way, we can switch a
pixel on and off in a display. Electro-optical displays of this type, which work with a
thin layer of a liquid crystalline layer, are called liquid crystal displays, LCD. There
is also a magneto-optical effect: By applying a magnetic field in the direction of
propagation direction of the linearly polarized electromagnetic wave, a dark
(optically inactive) arrangement becomes bright. The cause lies in a precession
motion of the oscillating charges around the magnetic field vector with the Larmor
frequency (cf. nuclear resonance in Chap. 10). It leads to a change in the propagation
velocities of right-circular and left-circular waves and thus to optical activity [8].
We learned chirality is the cause of the optical activity of molecules. As
examples, we looked at molecules with one chirality center. However, there are
also more complicated molecules. In addition to central chirality (molecules with
one chirality center), there is axial chirality (molecules with a chirality axis), planar
chirality (molecules with a chirality plane), and helicity (molecules in a helical
pattern such as DNA). Reactions in which chirality changes can be investigated
with femtosecond lasers [12].

12.3 Reflection, Absorption, and Transmission

When light hits an interface of two materials, normally a part of the light is reflected.
For directed reflection, the angle of reflection is equal to the angle of incidence. The
non-reflected part of the light enters the material behind the interface and is absorbed
or transmitted there.
12.3 Reflection, Absorption, and Transmission 495

If unpolarized light hits an interface of two transparent dielectric materials, then

for the directed reflection the angle of reflection is equal to the angle of incidence. If
there is an angle of 90° between the angle of incidence and the angle of reflection, the
reflected wave consists of linearly polarized light. This angle of incidence is called
the Brewster angle αB, it depends on the refractive indices of the materials:
In accordance with Snellius’ law (1.10) from the condition

αB þ β = 90 ° ð12:17Þ

follows
sin αB n
= 2 ð12:18Þ
sinðαB - 90Þ n 1

i.e.,
sin αB n
= tan αB = 2 ð12:19Þ
cos αB n1
or
n2
αB = arctan ð12:20Þ
n1
For this reason, reflected light is at least partially linearly polarized depending on
the angle of incidence. Reflections therefore can be filtered away with polarization
filters. On the other hand, incident linearly polarized light is completely, partially, or
not at all reflected, depending on the direction of polarization [8].
Reflectance is the ratio between incident and outgoing radiation. Smooth surfaces
lead to directed reflection, while rough surfaces cause diffuse reflection. The gloss of
smooth surfaces is a consequence of the reflections of directed reflection. The
smoother a surface is, the lower the proportion of diffuse reflection. Therefore, by
measuring the ratio of directed and diffuse reflection, the gloss of a body can be
quantified (glossmeter).
Emulsions are white for the viewer, although the main constituents, e.g. water and
oil, are not white. During the production of the disperse phase, in the form of small
oil droplets, the white color occurs. The cause again lies in the reflection. The (white)
light of our illumination is reflected by the oil droplets, so we see dispersed oil
droplets white. If, on the other hand, we look at the oil before making the emulsion,
we see a yellowish colored liquid. Here, the light penetrates the liquid oil phase, is
partially absorbed, and leaves the oil phase again in the direction of the eye of the
observer.
Absorption depends on the thickness of the radiated layer (Lambert-Beer law). In
photometers, the layer thickness of the sample is kept constant using cuvettes and the
Lambert-Beer law is used to determine the concentration by taking the absorption of
a liquid sample [8].
496 12 Optical Properties

12.4 Scattering and Diffraction

Scattering is the deflection of radiation from its original direction resulting from
interaction with a scattering center. For example, sunlight is scattered on the
molecules of the atmosphere. If, as is the case with the molecules of air, the
scattering centers are smaller than the wavelengths of the radiation, this is called
Rayleigh scattering. By the way, the Rayleigh scattering on the oxygen molecules is
also responsible for the blue of the sky and the diffuse scattered radiation ensures that
the sky appears bright during the day even when we do not look directly into the sun.
Particles in the atmosphere such as water droplets, ice and dust particles also scatter
part of the incident sunlight. If the particle dimensions are in the magnitude of the
wavelength of the radiation, Mie scattering occurs, with larger dimensions
Fraunhofer diffraction. Diffraction is the change in direction of light waves because
of interference on small objects. While we are usually not surprised when sound
waves come “around the corner,” we find it surprising when light not only moves
straight ahead, but like sound in many spatial directions at the same time. To
understand this, we use the model of a wave that moves from a point into a space
(spherical wave). When we shine light through a tiny circular aperture, we observe
colored, bright, and dark patterns on a screen behind the aperture, which we can only
explain by detaching ourselves from the model of the beam and using the wave
model. Using monochromatic light for this experiment, concentric bright and dark
circles appear on the screen. They are a consequence of the interference of waves
which, according to Huygens’ principle of elementary waves, are exiting from the
aperture. In spatial directions where a wave maximum meets a wave minimum
(negative interference) it is dark, in directions where wave maxima add up (positive
interference), it is bright. These brightness patterns are created by diffraction, which
is why they are also called diffraction patterns.
If the diameter of the aperture is reduced, an increase in the radius of the
concentric rings is observed. In this way it is possible to measure the size of the
diffraction pattern. This principle is used in particle size determination by means of
laser diffraction [13, 14] (Chap. 3). The laser light diffracted on tiny particles creates
bright circles as interference patterns, the diameter of which is used to calculate the
radius of the diffracting particles. If a sample consists of particles of different sizes,
the interference pattern produces circles with different diameters, the intensity of
which is a measure of the number of particles with this diameter. In this way, laser
diffraction devices can record complete particle size distributions in a second. The
measurement can also be performed on moving particles, which expands the
possibilities of industrial process analytics [15]. By means of laser diffraction,
dispersed particles of all kinds can be measured quickly and easily, e.g. in emulsions,
suspensions, aerosols, dusts [16]. While we can determine particle sizes with the
help of Mie scattering and Fraunhofer diffraction, we also use the scattered light of
illuminated materials, e.g. for spectroscopic quality assessment [17] and determina-
tion of color.
12.5 Colorimetry 497

12.5 Colorimetry

Colorimetry is the measurement and quantification of color. Colors are familiar to us

as visual sensations that arise from the impact of certain electromagnetic radiation on
the retina of the eye. The color sensed by a human viewer depends on physical
parameters, such as

• illuminance
• type of light
• illumination angle
• observation angle
• surface properties

but also, from the individual color vision ability of the viewer, his color perception,
and his linguistic description of the optical sensation. The goal of colorimetry is to
get an objective, quantitative description of color that is unique and can therefore be
used in technical systems.

12.5.1 Color as a Result of Selective Absorption

Visible light is electromagnetic radiation with wavelengths in the range of

400–800 nm. Larger wavelengths (IR) and smaller wavelengths (UV radiation) are
invisible to humans [18]. Within the visible area, further subdivisions can be made.
The visible spectrum can be split into several different colors (“colors of the
rainbow”). A rough classification is shown in Table 12.4.
By mixing light of different wavelengths, any color can be produced. This is
called additive color mixing. On the other hand, filtering out certain wavelengths
leads to a new color of the remaining light. This is called subtractive color mixing.
When we direct a beam of white light at a red grape berry, the anthocyanins in the
fruit absorb some of the light. Since the anthocyanin dyes prefer to absorb energy
around 500 nm (green), the reflected light lacks this green component and therefore
appears red to us. We recognize that the color of the grape berry is caused by
selective absorption of visible electromagnetic radiation. In this example we have
considered the reflection, we can look at the transmission analogously. To do this,
we shine white light on a cuvette with a sample of the liquid red grape juice and look
at the light coming out on the back of the cuvette. The transmitted light lacks the
green wavelengths because of the anthocyanin dyes and therefore it appears red. The
extent of selective absorption depends on the concentration of the absorbent
substances and the layer thickness of the sample. This behavior is described by

Table 12.4 Wavelength Red 700–770 nm

ranges of visible light
Yellow 570–590 nm
Blue 400–475 nm
498 12 Optical Properties

Lambert-Beer law. It says that the extinction E (the reciprocal transmittance) is

proportional to the concentration c and the layer thickness d.
I = I 0
e - μ
x ð12:21Þ

with fixed layer thickness (cuvette length) d

I = I 0
e - k
c
d ð12:22Þ

i.e.,

I0
ln =k
c
d ð12:23Þ
I
or
E=k
c
d ð12:24Þ

x penetration depth in m
d thickness in m
I intensity at x
I0 intensity at x=0
μ absorption coefficient
E extinction
k extinction coefficient
c concentration

In this way, with known extinction coefficient and defined layer thickness, the
concentration of dissolved substances can be determined conveniently and quickly
in this way. This is used in photometry and spectrophotometry. In addition to visible
wavelengths, other wavelengths are used, such as ultraviolet.
In colorimetry we distinguish between self-luminous bodies and non-self-lumi-
nous bodies. The grape berry mentioned above is an example of non-self-luminous
bodies whose color is created by illumination and selective absorption (subtractive
color mixing).

Example
Subtractive color mixing: The dye of a text highlighter absorbs a certain
wavelength from the incident light, so that the reflected light appears colored
to us. The dye of a yellow highlighter absorbs blue, so that the remaining
mixture of visible wavelengths appears yellow. We can say the dye subtracts
blue radiation from the light. If we mark over the same highlighted area with
several different highlighters, further wavelengths are subtracted, and the
overall picture becomes darker until it finally appears black.
12.5 Colorimetry 499

Self-luminous bodies emit visible electromagnetic radiation with a characteristic

spectrum. When the frequencies of self-luminous bodies are superimposed, colors
are created. This is additive color mixing.

Example
Additive color mixing: Let’s imagine a blue, a green, and a red LED that we
can turn on separately. We try different combinations. The color of the
resulting light is created by adding the individual wavelengths of the LEDs
we have switched on. When all three LEDs are lit at the same time, white light
is produced by the superimposition of blue, green, and red. This is the
principle of LED displays.

12.5.2 Physiology of Color Vision

There are two types of light-sensitive receptors on the retina of the human eye. The
so-called rods are light-dark-sensitive, the cones are sensitive to colors. Of the cones
there are three different types, the photopigments of these types have their absorption
wavelengths at 420 nm (blue), 535 nm (green), and 565 nm (red). Put simply, there
are cones for red, green, and blue wavelengths. Due to the interaction of these three
sensor types, we can perceive numerous colors. These colors are created by additive
color mixing in the human eye. The International Commission on Illumination
(Commission Internationale de l’Eclairage, CIE) developed a graphic representation
of all colors perceptible by humans in the form of a so-called color triangle
(Fig. 12.6).

Fig. 12.6 Chromaticity

diagram: Inside the
horseshoe-shaped triangle are
all colors which can be
observed. Colors with
maximum brilliance are on the
horseshoe curve. Point E in
the middle is zero brilliance
(white)
500 12 Optical Properties

The maximum saturated spectral colors are applied and placed in sequential order
according to their wavelength. The result is a horseshoe-shaped spectral curve, the
vertices of which are formed by violet, green, and red. The line drawn from violet to
red (the purple line) leads to the formation of the triangle. All colors which can be
perceived by a so-called normal observer are in this triangle [19], as well as colors
that do not occur in the spectrum of the sun (rainbow colors) such as purple. The
triangle is designed in such a way that the additive mixture of any two-color shades
creates a mixed color that lies on the connecting line in the color triangle. For
example, the mixture of red and green gives the color impression yellow or the
mixture of red and blue gives a purple hue. An additive mixture of colors leading to
point E provides the color impression white. Spectral colors, the mixture of which
yields white, are called complementary colors.

12.5.3 Terminology

In Terminology of the Commission Internationale de l’Eclairage, CIE [20] a point in

the chromaticity diagram (Fig. 12.6) is called hue or color stimulus specification.
Red, green, blue with defined monochromatic values for the wavelength are used as
reference color stimuli (Table 12.5).
A color with color stimulus specification E can be represented as a linear
combination of the reference color stimuli:
→ → → →
E =r
R þ g
G þ b
B ð12:25Þ

The reference color stimuli R, G, and B have the role of unit vectors, while r, b,
g represent the contributions with which the reference color stimuli are present in the
additive color mixture.

12.5.3.1 Color as a Vector

For technical purposes, we use the idea that color is a point in a three-dimensional
color space [22]. The vector pointing to this point from the coordinate origin is the
color vector. It has three components:

• hue
• brilliance
• brightness

Table 12.5 Wavelengths λ/nm

of reference color stimuli
R (red) 700.0
[21]
G (green) 546.1
B (blue) 435.8
12.5 Colorimetry 501

Fig. 12.7 Color vector in

polar coordinate system,
L brightness axis

D a b a

Fig. 12.8 Representation of

a color with a
two-dimensional vector. The
angle α indicates the hue, the
length of the vector d stands
for the brilliance

If you draw the vector in a three-dimensional coordinate system, denote the

vertical axis as brightness L, and mark the horizontal plane with the coordinates
a and b, you get a representation as in Fig. 12.7. The direction of the color vector
results from the values of its coordinates L, a, and b. The values of a and b indicate
the hue and brilliance (see below), while L indicates the brightness.
The graphical representation can be simplified by temporarily ignoring the L-axis.
Then you get an image as in Fig. 12.8 where the location of the color is now specified
by a two-dimensional vector with the coordinates a and b. The length of this vector
d is called color brilliance with
502 12 Optical Properties

Fig. 12.9 Representation of

a color in the L a b system

d= a 2 þ b2 ð12:26Þ

The direction of the vector d in the horizontal plane is determined by the ratio of
b over a, which is the tangent of the hue angle α

b
tan α = ð12:27Þ
a
Often a representation in which the colors are indicated in the horizontal plane is
used for easier orientation as shown in Fig. 12.9.
In all these graphical representations, the color vector is determined by these three
components: brightness, hue, and brilliance.
Let’s look again at the horseshoe-shaped curve in Fig. 12.6. If we mentally add a
brightness axis perpendicular to the color plane to this diagram and insert the
coordinates a and b instead of x and y, you get representations like in Figs. 12.8
and 12.9. The distance of a color point from the white point in Fig. 12.6 indicates the
brilliance of the color. Colors have maximum brilliance on the horseshoe-shaped
curve. As the color point moves closer to the white point, the color becomes paler,
i.e. less brilliant. When a color point is located directly in the white point, it appears
white, i.e. uncolored. The white point also has a brightness value and appears white,
light gray, dark gray, or black depending on the brightness. If the Lab system is used
to specify a color, the position of the color in the color plane can already be read from
the value range of components, a and b. Positive values of a stand for red, and
12.5 Colorimetry 503

Table 12.6 Lab color L a b Description

values, reading examples
0 a b Completely dark (black)
100 a b Completely bright
L -70 0 Green shades
L +70 0 Red shades
L 0 -70 Blue shades
L 0 +70 Yellow shades

negative values of a stand for green, while positive values of b stand for yellow and
negative values of b stand for blue. Table 12.6 gives some reading examples.
About 200 hues can be distinguished, about 20–25 brilliance levels and about
500 brightness levels. There are thus 2 million possibilities of combining hue,
brilliance, and brightness in color vision.

Bottom Line
A color can be clearly quantified by specifying its vector in color space with
the three independent color coordinates: brightness, hue, and brilliance.

Definition
The Lab system, or CIE Lab system, is the specification of a color as a vector
in the color space with the coordinates L, a, and b (CIE LAB Color Space).

In addition to the described color systems, there are numerous other systems for
the film and media industry, manufacturers of lamps, paints, coatings, printers, and
displays [23, 24].

Color Difference
An advantage of using vectors to describe the color is the calculation of the color
difference or the so-called color distance [25]. In the Lab system, color difference
ΔE of two colors is the Pythagorean sum of the differences in their L, a, b-values:

ΔE = ðL2 - L1 Þ2 þ ða2 - a1 Þ2 þ ðb2 - b1 Þ2 ð12:28Þ

i.e.,

ΔE = ΔL2 þ Δa2 þ Δb2 ð12:29Þ

504 12 Optical Properties

Example
Two products have their individual color values of L, a, and b
L1 = 80, a1 = 56, b1 = - 31 and L2 = 81, a2 = 54, b2 = - 29.
The color difference is

ΔE = ΔL2 þ Δa2 þ Δb2

p
ΔE = 12 þ 22 þ 22 = 9
ΔE = 3

Automatic sorting systems make use of on-line color measurements that are
programmed with the help of such color differences.

12.5.3.2 Visual Color Measurement Methods

In visual assessment, two objects are referred to as identical in color if no color
difference can be detected when viewed under the same lighting conditions with the
naked eye. Defined color standards are used to help specify the color of an object by
selecting the standard which best matches the color of the object (color match). Color
standards are used in numerous industry-specific color charts. For liquids the
American and European pharmacopeias recommend aqueous solutions of CaCl3
(pink), FeCl3 (yellow), and CuSO4 (blue) and their mixtures as color matching
fluids. The preparation and dilution of these color reference solutions is described
on a table in the Annex.

Applications
A common application of color measurement lies in automated quality assessment.
In this way, ripe and immature, uniformly and unevenly colored agricultural
products can be distinguished [26] or the quality of processed and fresh food
indicated [27, 28].

Further Reading

Refraction
Pharmaceutical solids: refractive index measurement [29]
Winemaking: temperature dependence of grape must refractive [30]
Peanut oil: refractive index and density measurements for determining oleic [31]
and linoleic acid
Glucose syrup: determination of dry matter content by refractive index [32]
Qualitative identification of food by terahertz complex refractive index [33]
mapping
Fishery products: Application of refraction index measurement [34]
Seawater: refractive index and salinity [35]

(continued)
12.5 Colorimetry 505

Refractive index of aqueous solutions of Na-acetate, Na-carbonate, Na-citrate [36]

with and without glycerol
Wine: Effect of ethanol, glycerol, glucose on refraction index [37]
Urine: Photonic crystal-based biosensor for the detection of glucose [38]
concentration
Fiber optic refractive index sensors [39, 40]

Polarimetry
Urine: noninvasive glucose detection by computer-based polarimeters [9]
Polarimetric examination of enzymatic reactions with L-DOPA (Parkinson’s [10]
disease drug)
Imaging polarimetry for enantiomer screening [4]
Ellipsometry-based biosensing of immunoglobulins [41]

Reflection, Absorption, Transmission

Fruits: characterization by time-resolved reflectance spectroscopy, TRS [42, 43]
Fruits and Vegetables: gloss measurement [44]
Brewster angle microscopy on monolayers [45, 46]
Spectral photometer for color measurement [47]

Scattering and Diffraction

Particle sizing by laser diffraction [13, 14, 48]
Particle sizing: comparing sedimentation and laser diffraction [49]
Particle sizing by dynamic light scattering (DLS) ISO 22412 [50]
Polarized light and laser scatter imaging for inspection of packagings [51]
Dynamic light scattering (DLS) for food quality evaluation [52]
Robotic optical inspection of fruits and vegetables [53, 54]

Colorimetry
Electronic eye for food sensory evaluation [55]
Beer: Color measurment [56, 57]
CCD technique for image analytical color measurement [58]
Hibiscus beverages: Colorimetry [59]
Banana: Color model for quality assessment [60]
Wheat sorting by color [61]
Simple method for analyzing the color of food by digital camera [62]
Tracking bioactive compounds by color [63]
Meat: Quality and color [64–66]
Powders: Colorimetric determination of mixture quality [67]
Carotenoids in corn: Color difference estimation [68]

(continued)
506 12 Optical Properties

Inactivation of microorganisms by blue LEDs [69]

LED illumination for agriculture [70]

Imaging
Chocolate: Microstructural Imaging [71]
Fermented dairy products: Imaging by confocal laser microscopy [72, 73]
Hyperspectral imaging on food [74–76]
Meat: Hyperspectral imaging for quality assessment [77, 78]
Apple: Dry matter content by NIR hyperspectral imaging [79]
Coffee: Quality sorting by hyperspectral imaging [80]
Imaging of double emulsions [81]
Food structure by atomic force microscopy (AFM) [82–84]
Microstructure of food by light microscopy [85]

Summary
The visual appearance of food depends on optical phenomena such as absorp-
tion, reflection, scattering, gloss, and color. In this chapter, physical causes of
refraction, diffraction, absorption, and transmission are explained in a concise
and comprehensive way, and how these phenomena stem from the interaction
between electromagnetic radiation and matter. They are illustrated by numer-
ous examples. Causes for optical activity in food ingredients are explained and
techniques such as polarimetry and ellipsometry are presented. Color is
represented as a vector in different color systems and compared with visual
sensation. At the end of the chapter, examples of applications are listed that
can be used for further studies.

References
1. Fraden J (2016) Handbook of modern sensors physics, designs, and applications. https://doi.
org/10.1007/978-3-319-19303-8
2. Zawirska-Wojtasiak R (2006) Chirality and the nature of food authenticity of Aroma. Acta Sci
Pol Technol Aliment 5:21
3. Maier NM, Franco P, Lindner W (2001) Separation of enantiomers: needs, challenges,
perspectives. J Chromatogr A 906(1):3. https://doi.org/10.1016/S0021-9673(00)00532-X
4. Gibbs PR, Uehara CS, Nguyen PT, Willson RC (2003) Imaging polarimetry for high throughput
chiral screening. Biotechnol Prog 19(4):1329. https://doi.org/10.1021/bp025729l
5. Mosandl A, Hener U, Fuchs S (2000) Natürliche Duft- und Aromastoffe — Echtheitsbewertung
mittels enantioselektiver Kapillar-GC und/oder Isotopenverhältnis-massenspektrometrie. In:
References 507

Günzler H (ed) Analytiker-Taschenbuch. Springer, Heidelberg. https://doi.org/10.1007/978-3-

642-57180-0_2
6. Fanali C, D’Orazio G, Gentili A, Fanali S (2019) Analysis of enantiomers in products of food
interest. Molecules 24(6). https://doi.org/10.3390/molecules24061119
7. D’Orazio G, Fanali C, Asensio-Ramos M, Fanali S (2017) Chiral separations in food analysis.
TrAC Trends Anal Chem 96:151. https://doi.org/10.1016/j.trac.2017.05.013
8. Meschede D (2006) Gerthsen Physik. Springer, Berlin. https://doi.org/10.1007/978-3-662-
45977-5
9. Sutrisno YA (2017) Noninvasive and painless urine glucose detection by using computer-based
polarimeter. Mater Sci Eng 202:012030. https://doi.org/10.1088/1757-899X/202/1/012030
10. Carroll A, Dick J, Selph L, Murphy M (2017) Using polarimetry to examine the kinetics of
enzymatic reactions with the Parkinson’s disease drug L-DOPA. In: Research & Creative
Activity Symposium, Montgomery, AL. Alabama State University. https://doi.org/10.1016/j.
ijbiomac.2013.01.031
11. Honig W, Moshudis E, Oette K (1981) Polarimetric determination of alpha-amylase activity
(author’s transl). J Clin Chem Clin Biochem / Zeitschrift fur klinische Chemie und klinische
Biochemie 19(10):1057
12. Baykusheva D, Zindel D, Svoboda V, Bommeli E, Ochsner M, Tehlar A, Wörner HJ (2019)
Real-time probing of chirality during a chemical reaction. Proc Natl Acad Sci 116(48):23923.
https://doi.org/10.1073/pnas.1907189116
13. Alexander M, Dalgleish DG (2006) Dynamic light scattering techniques and their applications
in food science. Food Biophys 1(1):2. https://doi.org/10.1007/s11483-005-9000-1
14. ISO 13320 (2020) Partikelgrößenanalyse - Partikelmessung durch Laserlichtbeugung. Beuth,
Berlin. https://doi.org/10.31030/2333155
15. Ruscitti O, Franke R, Hahn H, Babick F, Richter T, Stintz M (2008) Zum Einsatz der
Partikelmesstechnik in der Prozessintensivierung. Chem Ing Techn 80(1-2):191. https://doi.
org/10.1002/cite.200700168
16. Xu R (2015) Light scattering: a review of particle characterization applications. Particuology
18:11. https://doi.org/10.1016/j.partic.2014.05.002
17. Mollazade K, Omid M, Tab FA, Mohtasebi SS (2012) Principles and applications of light
backscattering imaging in quality evaluation of agro-food products: a review. Food Bioprocess
Technol 5(5):1465. https://doi.org/10.1007/s11947-012-0821-x
18. Udayakumar N (2014) Visible light imaging. In: Manickavasagan A, Jayasuriya H (eds)
Imaging with electromagnetic spectrum: applications in food and agriculture. Springer, Berlin,
pp 67–86. https://doi.org/10.1007/978-3-642-54888-8_5
19. DIN EN ISO/CIE 11664 (2020) Farbmetrik - Teil 1: CIE farbmetrische Normalbeobachter.
Beuth, Berlin. https://doi.org/10.31030/3091836
20. DIN EN ISO/CIE 11664 (2020) Farbmetrik - Teil 3: CIE-Farbwerte. Beuth, Berlin. https://doi.
org/10.31030/3092071
21. Kohlrausch F (1996) Praktische Physik Bd.1. Teubner, Stuttgart. https://doi.org/10.1007/978-3-
322-87205-0
22. DIN 5033 (2017) Farbmessung - Teil 1: Grundbegriffe der Farbmetrik. Beuth, Berlin. https://
doi.org/10.31030/2705354
23. Parkin A (2016) Colorimetry. In: Parkin A (ed) Digital imaging primer. Springer, Berlin, pp
269–281. https://doi.org/10.1007/978-3-540-85619-1_16
24. Lübbe E (2013) Farbempfindung, Farbbeschreibung und Farbmessung. Springer Vieweg,
Wiesbaden. https://doi.org/10.1007/978-3-8348-2228-4
25. ISO/CIE 11664-6 (2014) Farbmetrik - Teil 6: CIEDE2000 Formel für die Farbdifferenz. Beuth,
Berlin
26. Kang SP (2011) Color in food evaluation. In: Gliński J, Horabik J, Lipiec J (eds) Encyclopedia
of agrophysics. Springer, Dordrecht, pp 138–141. https://doi.org/10.1007/978-90-481-3585-1_
236
508 12 Optical Properties

27. Culver CA, Wrolstad RE (2008) Color quality of fresh and processed foods. American
Chemical Society; Distributed by Oxford University Press, Washington, DC. https://doi.org/
10.1021/bk-2008-0983
28. Wu D, Sun D-W (2013) Colour measurements by computer vision for food quality control – a
review. Trends Food Sci Technol 29(1):5. https://doi.org/10.1016/j.tifs.2012.08.004
29. Mohan S, Kato E, Drennen JK, Anderson CA (2019) Refractive index measurement of
pharmaceutical solids: a review of measurement methods and pharmaceutical applications. J
Pharm Sci 108(11):3478. https://doi.org/10.1016/j.xphs.2019.06.029
30. Jiménez-Márquez F, Vázquez J, Úbeda J, Sánchez-Rojas JL (2016) Temperature dependence of
grape must refractive index and its application to winemaking monitoring. Sens Actuators B
Chem 225:121. https://doi.org/10.1016/j.snb.2015.10.064
31. Davis JP, Sweigart DS, Price KM, Dean LL, Sanders TH (2013) Refractive index and density
measurements of peanut oil for determining oleic and linoleic acid contents. J Am Oil Chemists
Soc 90(2):199. https://doi.org/10.1007/s11746-012-2153-4
32. ISO 1743 (1982) Glucose syrup - determination of dry matter content- refractive index method.
Vernier, Switzerland
33. Shin HJ, Choi SW, Ok G (2018) Qualitative identification of food materials by complex
refractive index mapping in the terahertz range. Food Chem 245:282. https://doi.org/10.1016/
j.foodchem.2017.10.056. Epub 2017 Oct 13
34. Mathew S, Raman M, Kalarikkathara PM, Rajan DP (2019) Techniques used in fish and fishery
products analysis. In: Fish and fishery products analysis. Springer, Singapore. https://doi.org/
10.1007/978-981-32-9574-2_5
35. Tengesdal ØA (2012) Measurement of seawater refractive index and salinity by means of
optical refraction. University, Bergen
36. Bagheri M, Kiani F, Koohyar F, Khang NT, Zabihi F (2020) Measurement of refractive index
and viscosity for aqueous solution of sodium acetate, sodium carbonate, trisodium citrate,
(glycerol + sodium acetate), (glycerol + sodium carbonate), and (glycerol + trisodium citrate)
at T = 293.15 to 303.15 K and atmospheric pressure. J Mol Liq 309:113109. https://doi.org/10.
1016/j.molliq.2020.113109
37. Shehadeh A, Evangelou A, Kechagia D, Tataridis P, Chatzilazarou A, Shehadeh F (2020) Effect
of ethanol, glycerol, glucose/fructose and tartaric acid on the refractive index of model aqueous
solutions and wine samples. Food Chem 329:127085. https://doi.org/10.1016/j.foodchem.2020.
127085
38. Robinson S, Dhanlaksmi N (2016) Photonic crystal based biosensor for the detection of glucose
concentration in urine. Photonic Sens 7. https://doi.org/10.1007/s13320-016-0347-3
39. Wu Y, Liu B, Wang J, Wu J, Mao Y, Ren J, Zhao L, Sun T, Nan T, Han Y, Liu X (2021) A
novel temperature insensitive refractive index sensor based on dual photonic crystal fiber. Optik
226:165495. https://doi.org/10.1016/j.ijleo.2020.165495
40. Li J (2020) A review: Development of novel fiber-optic platforms for bulk and surface refractive
index sensing applications. Sens Actuators Rep 2(1):100018. https://doi.org/10.1016/j.snr.
2020.100018
41. Li K, Wang S, Wang L, Yu H, Jing N, Xue R, Wang Z (2017) Fast and sensitive ellipsometry-
based biosensing. Sensors (Basel, Switzerland) 18(1):15. https://doi.org/10.3390/s18010015
42. Torricelli A, Spinelli L, Contini D, Vanoli M, Rizzolo A, Eccher Zerbini P (2008) Time-
resolved reflectance spectroscopy for non-destructive assessment of food quality. Sens
Instrumen Food Qual 2(2):82. https://doi.org/10.1007/s11694-008-9036-2
43. Kim M, Lee K, Chao K, Lefcourt A, Jun W, Chan D (2008) Multispectral line-scan imaging
system for simultaneous fluorescence and reflectance measurements of apples: multitask apple
inspection system. Sens Instrumen Food Qual 2(2):123. https://doi.org/10.1007/s11694-008-
9045-1
44. Mizrach A, Lu RF, Rubino M (2009) Gloss evaluation of curved-surface fruits and vegetables.
Food Bioprocess Technol 2(3):300. https://doi.org/10.1007/s11947-008-0083-9
References 509

45. Pawluczyk R (1990) Modified Brewster angle technique for the measurement of the refractive
index of a DCG layer. Appl Opt 29:589. https://doi.org/10.1364/AO.29.000589
46. Daear W, Mahadeo M, Prenner EJ (2017) Applications of Brewster angle microscopy from
biological materials to biological systems. Biochim Biophys Acta Biomembr 1859(10):1749.
https://doi.org/10.1016/j.bbamem.2017.06.016
47. Fernandes GM, Silva WR, Barreto DN, Lamarca RS, Lima Gomes PCF, Petruci JFS, Batista
AD (2020) Novel approaches for colorimetric measurements in analytical chemistry – a review.
Anal Chim Acta 1135:187. https://doi.org/10.1016/j.aca.2020.07.030
48. Lee Black D, McQuay MQ, Bonin MP (1996) Laser-based techniques for particle-size mea-
surement: a review of sizing methods and their industrial applications. Prog Energy Combust
Sci 22(3):267. https://doi.org/10.1016/S0360-1285(96)00008-1
49. Ferro V, Mirabile S (2009) Comparing particle size distribution analysis by sedimentation and
laser diffraction method. J Agric Eng. https://doi.org/10.4081/jae.2009.2.35
50. ISO 22412 (2017) Particle Size Analysis – Dynamic Light Scattering (DLS). Beuth, Berlin
51. Barnes M, Dudbridge M, Duckett T (2012) Polarised light stress analysis and laser scatter
imaging for non-contact inspection of heat seals in food trays. J Food Eng 112(3):183. https://
doi.org/10.1016/j.jfoodeng.2012.02.040
52. Rao P, Yu Z, Han H, Xu Y, Ke L (2019) Dynamic light scattering for food quality
evaluation. In: Zhong J, Wang X (eds) Evaluation technologies for food quality. Woodhead
Publishing, pp 535–557. https://doi.org/10.1016/B978-0-12-814217-2.00020-2
53. Zhang B, Gu B, Tian G, Zhou J, Huang J, Xiong Y (2018) Challenges and solutions of optical-
based nondestructive quality inspection for robotic fruit and vegetable grading systems: a
technical review. Trends Food Sci Technol 81:213. https://doi.org/10.1016/j.tifs.2018.09.018
54. Li S, Luo H, Hu M, Zhang M, Feng J, Liu Y, Dong Q, Liu B (2019) Optical non-destructive
techniques for small berry fruits: a review. Artif Intell Agric 2:85. https://doi.org/10.1016/j.aiia.
2019.07.002
55. Xu C (2019) Electronic eye for food sensory evaluation. In: Zhong J, Wang X (eds) Evaluation
technologies for food quality. Woodhead Publishing, pp 37–59. https://doi.org/10.1016/B978-
0-12-814217-2.00004-4
56. Koren D, Hegyesné Vecseri B, Kun-Farkas G, Urbin Á, Nyitrai Á, Sipos L (2020) How to
objectively determine the color of beer? J Food Sci Technol 57(3):1183. https://doi.org/10.
1007/s13197-020-04237-4
57. Fengxia S, Yuwen C, Zhanming Z, Yifeng Y (2004) Determination of beer color using image
analysis. J Am Soc Brew Chem 62(4):163. https://doi.org/10.1094/ASBCJ-62-0163
58. Mendoza F, Dejmek P, Aguilera JM (2006) Calibrated color measurements of agricultural foods
using image analysis. Postharvest Biol Technol 41(3):285. https://doi.org/10.1016/j.
postharvbio.2006.04.004
59. Camelo-Mendez GA, Vanegas-Espinoza PE, Escudero-Gilete ML, Heredia FJ, Paredes-Lopez-
O, Del Villar-Martinez AA (2018) Colorimetric analysis of hibiscus beverages and their
potential antioxidant properties. Plant Food Hum Nutr 73(3):247
60. Nannyonga S, Bakalis S, Andrews J, Mugampoza E, Gkatzionis K (2016) Mathematical
modelling of color, texture kinetics and sensory attributes characterisation of ripening bananas
for waste critical point determination. J Food Eng 190:205. https://doi.org/10.1016/j.jfoodeng.
2016.06.006
61. Pearson T, Brabec D, Haley S (2008) Color image based sorter for separating red and white
wheat. Sens Instrumen Food Qual 2(4):280
62. Yam KL, Papadakis SE (2004) A simple digital imaging method for measuring and analyzing
color of food surfaces. J Food Eng 61(1):137. https://doi.org/10.1016/s0260-8774(03)00195-x
63. Sant’Anna V, Gurak PD, Ferreira Marczak LD, Tessaro IC (2013) Tracking bioactive
compounds with colour changes in foods – a review. Dyes Pigments 98(3):601. https://doi.
org/10.1016/j.dyepig.2013.04.011
510 12 Optical Properties

64. Ijaz M, Li X, Zhang D, Hussain Z, Ren C, Bai Y, Zheng X (2020) Association between meat
color of DFD beef and other quality attributes. Meat Sci 161:107954. https://doi.org/10.1016/j.
meatsci.2019.107954
65. Faustman C, Suman SP (2017) The eating quality of meat: I—Color. In: Toldra F (ed) Lawrie’s
meat science, 8th edn. Woodhead Publishing, pp 329–356. https://doi.org/10.1016/B978-0-08-
100694-8.00011-X
66. Warner R (2014) Measurement of meat quality | Measurements of water-holding capacity and
color: objective and subjective. In: Dikeman M, Devine C (eds) Encyclopedia of meat sciences.
Academic Press, Oxford, pp 164–171. https://doi.org/10.1016/B978-0-12-384731-7.00210-5
67. Shenoy P, Innings F, Lilliebjelke T, Jonsson C, Fitzpatrick J, Ahrné L (2014) Investigation of
the application of digital colour imaging to assess the mixture quality of binary food powder
mixes. J Food Eng 128:140. https://doi.org/10.1016/j.jfoodeng.2013.12.013
68. Nkhata SG (2020) Total color change (ΔE) is a poor estimator of total carotenoids lost during
post-harvest storage of biofortified maize grains. Heliyon 6(10):e05173. https://doi.org/10.
1016/j.heliyon.2020.e05173
69. Hyun J-E, Lee S-Y (2020) Blue light-emitting diodes as eco-friendly non-thermal technology in
food preservation. Trends Food Sci Technol 105:284. https://doi.org/10.1016/j.tifs.2020.09.008
70. Gupta SD (2017) Light emitting diodes for agriculture. Springer, Singapore. https://doi.org/10.
1007/978-981-10-5807-3
71. Rousseau D (2016) Microstructural imaging of chocolate confectionery. In: Sozer N
(ed) Imaging technologies and data processing for food engineers. Springer International,
Cham, pp 311–333. https://doi.org/10.1007/978-3-319-24735-9_10
72. Ercili-Cura D (2016) Imaging of fermented dairy products. In: Sozer N (ed) Imaging
technologies and data processing for food engineers. Springer International, Cham, pp
99–128. https://doi.org/10.1007/978-3-319-24735-9_4
73. Lorén N, Langton M, Hermansson AM (2007) Confocal fluorescence microscopy (CLSM) for
food structure characterisation. In: McClements DJ (ed) Understanding and controlling the
microstructure of complex foods. Woodhead Publishing, pp 232–260. https://doi.org/10.1533/
9781845693671.2.232
74. Liu YW, Pu HB, Sun DW (2017) Hyperspectral imaging technique for evaluating food quality
and safety during various processes: a review of recent applications. Trends Food Sci Technol
69:25. https://doi.org/10.1016/j.tifs.2017.08.013
75. Baiano A (2017) Applications of hyperspectral imaging for quality assessment of liquid based
and semi-liquid food products: a review. J Food Eng 214:10. https://doi.org/10.1016/j.jfoodeng.
2017.06.012
76. Park B, Lu R (2015) Hyperspectral imaging technology in food and agriculture. https://doi.org/
10.1007/978-1-4939-2836-1
77. Feng CH, Makino Y, Oshita S, Martin JFG (2018) Hyperspectral imaging and multispectral
imaging as the novel techniques for detecting defects in raw and processed meat products:
current state-of-the-art research advances. Food Control 84:165. https://doi.org/10.1016/j.
foodcont.2017.07.013
78. Antequera T, Caballero D, Grassi S, Uttaro B, Perez-Palacios T (2021) Evaluation of fresh meat
quality by Hyperspectral Imaging (HSI), Nuclear Magnetic Resonance (NMR) and Magnetic
Resonance Imaging (MRI): a review. Meat Sci 172:108340. https://doi.org/10.1016/j.meatsci.
2020.108340
79. Ma T, Li X, Inagaki T, Yang H, Tsuchikawa S (2018) Noncontact evaluation of soluble solids
content in apples by near-infrared hyperspectral imaging. J Food Eng. https://doi.org/10.1016/j.
jfoodeng.2017.12.028
80. Nansen C, Singh K, Mian A, Allison BJ, Simmons CW (2016) Using hyperspectral imaging to
characterize consistency of coffee brands and their respective roasting classes. J Food Eng 190:
34. https://doi.org/10.1016/j.jfoodeng.2016.06.010
References 511

81. Bernewitz R, Guthausen G, Schuchmann HP (2016) Imaging of double emulsions. In: Sozer N
(ed) Imaging technologies and data processing for food engineers. Springer International,
Cham, pp 69–98. https://doi.org/10.1007/978-3-319-24735-9_3
82. Morris VJ (2007) Atomic force microscopy (AFM) techniques for characterising food
structure. In: McClements DJ (ed) Understanding and controlling the microstructure of complex
foods. Woodhead Publishing, pp 209–231. https://doi.org/10.1533/9781845693671.2.209
83. Gunning AP, Morris VJ (2018) Getting the feel of food structure with atomic force microscopy.
Food Hydrocoll 78:62. https://doi.org/10.1016/j.foodhyd.2017.05.017
84. Middendorf D, Bindrich U, Mischnick P, Juadjur A, Franke K, Heinz V (2016) Atomic Force
Microscopy study on the effect of different lecithins in cocoa-butter based suspensions. Colloids
Surf A Physicochem Eng Aspects 499:60. https://doi.org/10.1016/j.colsurfa.2016.03.057
85. Gunning PA (2013) Light microscopy: principles and applications to food microstructures. In:
Morris VJ, Groves K (eds) Food microstructures. Woodhead Publishing, pp 62–95. https://doi.
org/10.1533/9780857098894.1.62
UV and X-rays
13

Frequencies above the visible range of the electromagnetic spectrum are invisible to
the human eye. The wavelength ranging from 380 nm to 1 nm is called ultraviolet
radiation, wavelengths shorter than 1 nm are called X-rays. When UV radiation hits a
material, the valence electrons are energetically excited. For this reason, UV spec-
troscopy is also called electron spectroscopy. X-rays, on the other hand, are absorbed
by internal electrons. The propagation speed c of electromagnetic waves is the
product of frequency f and wavelength λ.
c=λ
f ð13:1Þ

c propagation speed in m s-1

λ wavelength in nm
f frequency in s-1

13.1 UV

There are different perspectives on the subdivision of the ultraviolet range. Physi-
cally, it is divided into near UV (380–200 nm), far UV (200–10 nm), and extremely
far UV (1–31 nm) (Table 13.1). Regarding the filtering of the Sun’s UV radiation
through the Earth’s atmosphere, the distinction is made between UV-A
(380–315 nm), UV-B (315–280 nm), and UV-C (280–100 nm). The UV-A range
includes the wavelengths that are hardly absorbed by the atmosphere and thus reach
the Earth’s surface. Wavelengths below 315 nm are largely absorbed by atmospheric
ozone and wavelengths below 280 nm by atmospheric oxygen. Laboratory
experiments with UV below a wavelength of 200 nm must therefore be carried out
under vacuum, which is why the range of 200 to 1 nm is also called vacuum UV
(Fig. 13.1).

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 513
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_13
514 13 UV and X-rays

Table 13.1 Electromagnetic radiation, UV, and X-ray

Frequency in Hz Designation Vacuum wavelength in m
3 ∙ 1021 10-13
3 ∙ 1020 X-ray 10-12
3 ∙ 1019 10-11
3 ∙ 1018 10-10
3 ∙ 1017 10-9
3 ∙ 1016 Ultraviolet (UV) 10-8
3 ∙ 1015 10-7
3 ∙ 1014 Light infrared (IR) 10-6
3 ∙ 1013 10-5
3 ∙ 1012 10-4

Fig. 13.1 Classification of ultraviolet radiation (UV)

Because the UV absorption of materials provides information about valence

electrons and thus about molecular properties, UV radiation is used in analytical
spectroscopy. UV radiation allows electrons to change their energetic level. For
example, from the lowest unoccupied molecular orbital (LUMO) in a molecule to the
highest occupied molecular orbital (HOMO) [1].
Spectrometers that use both visible and UV frequencies are called UV/vis
spectrometers. UV radiation can also lead to fluorescence of molecules, which is
used in fluorescence spectroscopy. Numerous sensors are based on the interaction of
analytes with UV radiation [2].
In addition, UV radiation is used to disinfect water and food. Between 200 and
300 nm, most proteins, amino acids, and nucleic acids absorb UV radiation. With
sufficiently high energy intake, this can lead to molecular structural changes that lead
to the loss of original function. For this reason, it is possible to inactivate
microorganisms with short-wave UV radiation, as well as viruses and
bacteriophages [3–6]. With short-wave UV radiation, bacteria in food can be
inactivated and food shelf lives increased [4, 7–9]. UV radiation can be used to
disinfect food surfaces or work surfaces and machines in processing plants. In
addition, gases or liquids with low layer thickness can be treated. In the field of
drinking water treatments, there are numerous UV applications, some in combina-
tion with other inactivation processes [10–18]. The application of pulsed short-wave
UV [19–22] along with powerful LEDs with specific frequency spectrum is
promising [3, 7].
13.2 X-ray 515

13.2 X-ray

Electromagnetic radiation with wavelengths in the range of 10-11 m to 10-9 m or

frequencies of 1013 Hz to 1017 Hz is called X-rays. X-rays with wavelengths smaller
than 10-10 m are called hard X-rays, with wavelengths larger than 10-10 m soft
X-rays. Radiation with wavelengths below 10-11 m can be described as ultrahard
X-rays and radiation with wavelengths above 10-9 m as ultrasoft X-rays.
Conceptually, the transition from ultrasoft X-rays to UV radiation is fluid (Fig. 13.2).
The energy of X-ray photons is several orders of magnitude higher than the
energy of visible light photons. The energy of an X-ray photon is proportional to the
frequency of the radiation. It can be expressed in electron volts (eV) instead of
Joules (J).
The conversion is: 1 eV = 1.609 ∙ 10-19 J.
It is

E=h
f ð13:2Þ

E energy in J or eV
h Planck’s constant h = 6.62607015 ∙ 10-34 J s
f frequency in s-1

Example
Wavelength and photon energy of X-rays having a frequency 1018 Hz
is with c = λ
f it is

c 3
108 ms - 1
λ= = = 3
10 - 10 m = 30 nm
f 1018 s - 1
and with E = h
f the energy is

E = h
f = 6:6
1034 J s
1018 s1 = 6:6
1016 J = 4:1 eV

By absorbing X-rays, ions can be generated in the absorbent material. X-rays are
therefore assigned to ionizing radiation. Due to their high penetration depth, X-rays
can be used to elucidate the structure of materials. X-ray diffraction can be used to
elucidate the lattice structures of crystalline materials. Amorphous solid materials
that can be formed during the drying of food [23, 24] can be distinguished from
crystalline materials by powder X-ray diffractometry [25].
516 13 UV and X-rays

Fig. 13.2 Categories of 100Å 10Å 1Å 0.1Å 0.01Å

X-rays

ultra weak weak hard ultra hard

108m 109m 1010m 1011m 1012m

13.2.1 X-ray Absorption

The main absorption mechanisms for X-rays are the photo effect, elastic and inelastic
scattering as well as pair formation. Pairing is the formation of electron–positron
pairs, which occurs with X-ray photons with energies above a few MeV. X-ray
diffraction is used to elucidate the structure of solids, while the photoelectric effect is
the main cause of absorption in biological materials. In the photo effect, the intake of
the energy of an X-ray photon leads to the release of an electron from the hit atom. At
a sufficiently high X-ray frequency, internal electrons of an atom are also released.
Fluorescence or the appearance of Auger electrons occurs when the electron places
have been filled up by external electrons. Both can be used spectroscopically to
determine the composition of the absorbent material. The Lambert-Beer law applies
to the attenuation of X-rays by the photo effect. Accordingly, the intensity decreases
exponentially with the depth of penetration:

I = I 0
e - μx ð13:3Þ

I intensity at x
I0 intensity at x = 0
μ absorption coefficient
x penetration depth in m

The absorption coefficient μ for X-rays depends on the number density and cross-
section of the absorbing atoms and thus strongly on the atomic number of the
absorbing material. It can be shown that the absorption coefficient is proportional
to the third power of the atomic number [26].

Example
Lead shows about 30 times greater X-ray absorption than iron. So a 1 mm
thick layer of lead has the same shielding effect as a 3 cm thick layer of iron. If
we compare the substances on the basis of their density and atomic numbers, it

(continued)
13.2 X-ray 517

becomes apparent that the third power of the atomic numbers of both materials
also differs by a factor of about 30. The density of lead, on the other hand, is
not 30 times but only about 40% higher than that of iron:

Lead Iron
Density in g∙cm-3 11.3 7.9
Atomic number Z 82 26
Z3 551,368 17,576

The different absorption coefficients of materials are used in X-ray diagnostics

and X-ray inspection. Here, the X-ray transmission of a fan-shaped beam is detected
by suitable detectors. Depending on the absorption coefficient and the layer thick-
ness of the radiated material, an image with different shades of gray is created in the
detector. The grayscales can be technically converted into colors for better visual
interpretation.
Attention The rule of thumb “the higher the density of a material, the higher its
X-ray absorption” must not be taken too strictly and certainly not misunderstood as a
linear relationship. There are numerous examples that contradict this rule of thumb,
e.g. Beryllium has a lower X-ray absorption than potassium, although its density is
twice as high.

13.2.2 Imaging Techniques

With the help of X-ray imaging techniques, shapes of materials can be detected. In
this way, e.g. by X-ray inspection, suspicious objects in baggage or freight
containers can be detected. In the presence of strong absorption differences,
materials can also be recognized or distinguished from the surrounding matrix,
e.g. glass, wood, bones in food. For this reason, X-ray scanners are used in food
production to inspect packaged foods. Applications range from X-ray systems for
entire trucks to fast production line scanners for ready-packed food. Contrasts in the
generated images, i.e. the different X-ray absorption of adjacent areas, are evaluated.
High contrasts occur with large differences in atomic number Z, i.e. when there is a
glass, bone, or metal in a food. Atomic numbers in food are in the range of atomic
numbers of the main components, i.e. carbon (Z = 6), oxygen (Z = 8), hydrogen
(Z = 1), and nitrogen (Z = 7), while the atomic number of calcium in the bones is
Z = 20, of the silicon in mineral glass is Z = 14, and the atomic number of metallic
materials is even higher. Materials that consist of atoms like those of food will show
similar X-ray absorption and, therefore, will form only a faint contrast to food. They
are therefore not easy to identify. These include materials such as wood, fruit
kernels, insects, and plastics such as polypropene, which have a similar element
composition to the surrounding food matrix. Therefore, X-ray scanners cannot
518 13 UV and X-rays

differentiate certain material combinations. X-ray scanners have numerous

advantages over metal detectors in contaminant detection. While metal detectors
have a high sensitivity to metals with high magnetic susceptibility, sensitivity to
metals with low high magnetic susceptibility (Al, Cu, brass) is lower and is further
reduced by foods with high electrical conductivity (refer Chap. 9). X-ray scanners
for detection of foreign bodies do not have these limitations. In addition to steel and
all non-ferrous metals, they can also be used to detect non-metallic impurities such as
fragments of glass, stone and bone, and some plastics. Contaminant detection is also
possible in aluminum-packed foods.
Parallel to the inspection of a food product for foreign bodies, the completeness of
a product can be determined from the X-ray image and also the layer thickness of the
food on the basis of the grayscale. In this way, complete packs can be distinguished
from incompletely filled ones. With appropriate calibration, the setpoint of the mass
of the packaged food can be checked from the grayscale X-ray image [27].

Definition
CT stands for computed tomography. Tomography comes from Greek. tóμoς
(section) and Greek γράφειν (write) and refers to the production of cross-
sectional images. X-ray CT is a standard procedure of medical diagnostics.

In a medical CT technique a fan-like X-ray beam rotates around the body to be

examined. In this way, numerous transmission images from different directions are
obtained, which are combined to form a cross-sectional image of the body. By
slowly moving the body through the CT machine, three-dimensional representations
of the inside of the body can be created. To increase the X-ray absorption of certain
areas, solutions with a high atomic number are administered, so-called contrast
agents like iodine-containing solutions. With the help of computed tomography on
small objects (micro-CT, μCT), microstructures of food can be examined [28] and
food processes can be investigated [29, 30].
In this chapter, we have become acquainted with ways to characterize food by
image analysis. In addition to microscopy with electromagnetic waves and electron
beams, there is acoustic spectroscopy (see Chap. 15) and nuclear magnetic resonance
(see Chap. 10). A comprehensive comparison of the different analytical applications
can be found in [31].

Further Reading
Ultraviolet

Solar water disinfection: effect of UV and temperature [32, 33]

Fruits and vegetables: microbial inactivation by UV [5, 9, 34]
Corona-Virus: inactivation by UV-LEDs [3]
Tomato: microbial and color effects during UV-C treatment [35]
Lettuce: microbial and sensory effects during UV-C-treatment [36]

(continued)
References 519

Carrot juice: comparison of UV-C and thermal treatment [37]

UV-LED for inactivation of pathogens [4, 6, 7, 38]
UV absorption of DNA, RNA, and proteins [39]
E. coli and L. plantarum: UV-C-sensitivity [40]
Tomato processing: inactivation of microorganisms by UV-C [41]

X-rays

Identification of X-ray-irradiated hazelnuts by electron spin resonance (ESR) [42]

Density of foods by X-ray Imaging [27]
Lactose: Crystal structure by X-ray diffraction [43–45]
Coffee beans: Micro X-ray tomography (μCT) during roasting [29]
Frozen food: μCT on ice crystals [46]
Bread: three-dimensional quantitative by μCT [47]
Apple: pore space structure by μCT [48]
Meat: structure by μCT [30]
Foams: structure by μCT [49]
Milk: analysis of chemical elements by X-ray fluorescence [50]
Cacao: analysis of chemical elements by X-ray fluorescence [51]
Corn: kernel hardness and density by μCT [52]
X-ray imaging methods for internal quality evaluation of agricultural produce [53]

Summary
Ultraviolet radiation can be used for disinfection. In this chapter, basic
concepts and the differences between atmospheric and technical UV are
clarified in concise words. At the end of the section, application examples
are listed from which the potential of future UV-LEDs can be seen. In the
electromagnetic spectrum, the field of X-rays adjoins the short-wave UV. The
basic features of the absorption of X-rays are presented and applications of
X-ray scanners and imaging methods are described. At the end of the section,
numerous application examples from the field of micro-CT are listed, which
can be used for further studies and as suggestions for your own scientific work

References
1. Hinderer F (2020) UV/Vis-Absorptions- und Fluoreszenz- Spektroskopie. Springer,
Heidelberg. https://doi.org/10.1007/978-3-658-25441-4
520 13 UV and X-rays

2. Fraden J (2016) Handbook of modern sensors physics, designs, and applications. https://doi.
org/10.1007/978-3-319-19303-8
3. Gerchman Y, Mamane H, Friedman N, Mandelboim M (2020) UV-LED disinfection of
Coronavirus: wavelength effect. J Photochem Photobiol B Biol 212:112044. https://doi.org/
10.1016/j.jphotobiol.2020.112044
4. Koutchma T, Popović V, Green A (2019) Overview of ultraviolet (UV) LEDs technology for
applications in food production. In: Koutchma T (ed) Ultraviolet LED technology for food
applications. Academic, pp 1–23. https://doi.org/10.1016/B978-0-12-817794-5.00001-7
5. Baykuş G, Akgün MP, Unluturk S (2021) Effects of ultraviolet-light emitting diodes
(UV-LEDs) on microbial inactivation and quality attributes of mixed beverage made from
blend of carrot, carob, ginger, grape and lemon juice. Innovative Food Sci Emerg Technol 67:
102572. https://doi.org/10.1016/j.ifset.2020.102572
6. Souza VR, Koutchma T (2021) Repair mechanisms of UV-induced damage of microorganism
in foods. In: Knoerzer K, Muthukumarappan K (eds) Innovative food processing technologies.
Elsevier, Oxford, pp 385–397. https://doi.org/10.1016/B978-0-08-100596-5.22943-0
7. Koutchma T, Popović V (2019) UV light-emitting diodes (LEDs) and food safety. In:
Koutchma T (ed) Ultraviolet LED technology for food applications. Academic, pp 91–117.
https://doi.org/10.1016/B978-0-12-817794-5.00005-4
8. Popović V, Koutchma T, Pagan J (2021) Emerging applications of ultraviolet light-emitting
diodes for foods and beverages. In: Knoerzer K, Muthukumarappan K (eds) Innovative food
processing technologies. Elsevier, Oxford, pp 335–344. https://doi.org/10.1016/B978-0-08-
100596-5.22667-X
9. Barut Gök S, Gräf V, Stahl MR (2021) Engineering aspects of UV-C processing for liquid
foods. In: Knoerzer K, Muthukumarappan K (eds) Innovative food processing technologies.
Elsevier, Oxford, pp 171–181. https://doi.org/10.1016/B978-0-08-100596-5.23000-X
10. Xu R, Gong L, Wang H, Zhang G (2020) Disinfection effect of short-wave ultraviolet radiation
(UV-C) on ASFV in water. J Infect 80(6):671. https://doi.org/10.1016/j.jinf.2020.02.021
11. Wang C-P, Chang C-S, Lin W-C (2020) Efficiency improvement of a flow-through water
disinfection reactor using UV-C light emitting diodes. J Water Process Eng 40:101819. https://
doi.org/10.1016/j.jwpe.2020.101819
12. Zou X-Y, Lin Y-L, Xu B, Cao T-C, Tang Y-L, Pan Y, Gao Z-C, Gao N-Y (2019) Enhanced
inactivation of E. coli by pulsed UV-LED irradiation during water disinfection. Sci Total
Environ 650:210. https://doi.org/10.1016/j.scitotenv.2018.08.367
13. Wen G, Deng X, Wan Q, Xu X, Huang T (2019) Photoreactivation of fungal spores in water
following UV disinfection and their control using UV-based advanced oxidation processes.
Water Res 148:1. https://doi.org/10.1016/j.watres.2018.10.028
14. Li X, Cai M, Wang L, Niu F, Yang D, Zhang G (2019) Evaluation survey of microbial
disinfection methods in UV-LED water treatment systems. Sci Total Environ 659:1415.
https://doi.org/10.1016/j.scitotenv.2018.12.344
15. Ayoub GM, Malaeb L (2019) Solar water disinfection: UV radiation transmittance of various
solar reactor tubes. Energy Procedia 157:498. https://doi.org/10.1016/j.egypro.2018.11.214
16. Nyangaresi PO, Qin Y, Chen G, Zhang B, Lu Y, Shen L (2018) Effects of single and combined
UV-LEDs on inactivation and subsequent reactivation of E. coli in water disinfection. Water
Res 147:331. https://doi.org/10.1016/j.watres.2018.10.014
17. Nocker A, Shah M, Dannenmann B, Schulze-Osthoff K, Wingender J, Probst AJ (2018)
Assessment of UV-C-induced water disinfection by differential PCR-based quantification of
bacterial DNA damage. J Microbiol Methods 149:89. https://doi.org/10.1016/j.mimet.2018.
03.007
18. Li G-Q, Wang W-L, Huo Z-Y, Lu Y, Hu H-Y (2017) Comparison of UV-LED and low pressure
UV for water disinfection: photoreactivation and dark repair of Escherichia coli. Water Res 126:
134. https://doi.org/10.1016/j.watres.2017.09.030
References 521

19. Izmirlioglu G, Ouyang B, Demirci A (2020) Utilization of pulsed UV light for inactivation of
Salmonella Enteritidis on shelled walnuts. LWT 134:110023. https://doi.org/10.1016/j.lwt.
2020.110023
20. Garvey M, Rowan NJ (2019) Pulsed UV as a potential surface sanitizer in food production
processes to ensure consumer safety. Curr Opin Food Sci 26:65. https://doi.org/10.1016/j.cofs.
2019.03.003
21. Pedrós-Garrido S, Condón-Abanto S, Clemente I, Beltrán JA, Lyng JG, Bolton D, Brunton N,
Whyte P (2018) Efficacy of ultraviolet light (UV-C) and pulsed light (PL) for the
microbiological decontamination of raw salmon (Salmo salar) and food contact surface
materials. Innovative Food Sci Emerg Technol 50:124. https://doi.org/10.1016/j.ifset.2018.
10.001
22. Keklik NM, Krishnamurthy K, Demirci A (2012) Microbial decontamination of food by
ultraviolet (UV) and pulsed UV light. In: Demirci A, Ngadi MO (eds) Microbial decontamina-
tion in the food industry. Woodhead Publishing, pp 344–369. https://doi.org/10.1533/
9780857095756.2.344
23. Mendonsa N, Almutairy B, Kallakunta VR, Sarabu S, Thipsay P, Bandari S, Repka MA (2020)
Manufacturing strategies to develop amorphous solid dispersions: An overview. J Drug Deliv
Sci Technol 55:101459. https://doi.org/10.1016/j.jddst.2019.101459
24. Thakral S, Terban MW, Thakral NK, Suryanarayanan R (2016) Recent advances in the
characterization of amorphous pharmaceuticals by X-ray diffractometry. Adv Drug Deliv Rev
100:183. https://doi.org/10.1016/j.addr.2015.12.013
25. Figura LO, Epple M (1995) Anhydrous α-lactose A study with DSC and TXRD. J Therm Anal
44(1):45. https://doi.org/10.1007/BF02547132
26. Meschede D (2006) Gerthsen Physik. Springer, Berlin. https://doi.org/10.1007/978-3-662-
45977-5
27. Kelkar S, Boushey CJ, Okos M (2015) A method to determine the density of foods using X-ray
imaging. J Food Eng 159:36. https://doi.org/10.1016/j.jfoodeng.2015.03.012
28. Schoeman L, Williams P, du Plessis A, Manley M (2016) X-ray micro-computed tomography
(μCT) for non-destructive characterisation of food microstructure. Trends Food Sci Technol 47:
10. https://doi.org/10.1016/j.tifs.2015.10.016
29. Frisullo P, Barnaba M, Navarini L, Del Nobile MA (2012) Coffea arabica beans microstructural
changes induced by roasting: an X-ray microtomographic investigation. J Food Eng 108(1):232.
https://doi.org/10.1016/j.jfoodeng.2011.07.036
30. Frisullo P, Laverse J, Marino R, Nobile MAD (2009) X-ray computed tomography to study
processed meat microstructure. J Food Eng 94(3–4):283. https://doi.org/10.1016/j.jfoodeng.
2009.03.020
31. Metilli L, Francis M, Povey M, Lazidis A, Marty-Terrade S, Ray J, Simone E (2020) Latest
advances in imaging techniques for characterizing soft, multiphasic food materials. Adv Colloid
Interf Sci 279:102154. https://doi.org/10.1016/j.cis.2020.102154
32. Castro-Alférez M, Polo-López MI, Marugán J, Fernández-Ibáñez P (2017) Mechanistic
modeling of UV and mild-heat synergistic effect on solar water disinfection. Chem Eng J
316:111. https://doi.org/10.1016/j.cej.2017.01.026
33. Lawrie K, Mills A, Figueredo-Fernandez M, Gutierrez-Alfaro S, Manzano M, Saladin M (2015)
UV dosimetry for solar water disinfection (SODIS) carried out in different plastic bottles and
bags. Sens Actuators B Chem 208:608. https://doi.org/10.1016/j.snb.2014.11.031
34. Pala ÇU, Toklucu AK (2012) Effects of UV-C light processing on some quality characteristics
of grape juices. Food Bioproc Tech 6(3):719. https://doi.org/10.1007/s11947-012-0808-7
35. Song H-J, Choi D-W, Song KB (2011) Effect of aqueous chlorine dioxide and UV-C treatment
on the microbial reduction and color of cherry tomatoes. Hortic Environ Biotechnol 52(5):488.
https://doi.org/10.1007/s13580-011-0043-6
36. Lee JH, Chun HH, Oh DH, Park J, Won M, Song KB (2012) Sensory and microbiological
qualities of romaine lettuce and kale affected by a combined treatment of aqueous chlorine
dioxide and ultraviolet-C. Hortic Environ Biotechnol 53(5):387. https://doi.org/10.1007/
s13580-012-0013-7
522 13 UV and X-rays

37. Riganakos KA, Karabagias IK, Gertzou I, Stahl M (2017) Comparison of UV-C and thermal
treatments for the preservation of carrot juice. Innovative Food Sci Emerg Technol 42:165.
https://doi.org/10.1016/j.ifset.2017.06.015
38. D’Souza C, Yuk H-G, Khoo GH, Zhou W (2017) Light-emitting diodes in postharvest quality
preservation and microbiological food safety. In: Dutta Gupta S (ed) Light emitting diodes for
agriculture: smart lighting. Springer Singapore, Singapore, pp 191–235. https://doi.org/10.
1007/978-981-10-5807-3_9
39. Lavilla M, Lasagabaster A, Martínez-de-Marañón I (2019) Impact of ultraviolet processing on
food composition. In: Roohinejad S, Koubaa M, Greiner R, Mallikarjunan K (eds) Effect of
emerging processing methods on the food quality: advantages and challenges. Springer Inter-
national Publishing, Cham, pp 173–196. https://doi.org/10.1007/978-3-030-18191-8_8
40. Müller A, Pietsch VL, Schneele JS, Stahl MR, Greiner R, Posten C (2015) Effect of temperature
and pH value on the UV-C sensitivity of Escherichia coli and Lactobacillus plantarum. LWT
Food Sci Technol 64(2):699. https://doi.org/10.1016/j.lwt.2015.06.013
41. Abadias M, Colás-Medà P, Viñas I, Bobo G, Aguiló-Aguayo I (2021) Application of an
innovative water-assisted ultraviolet C light technology for the inactivation of microorganisms
in tomato processing industries. Food Microbiol 94:103631. https://doi.org/10.1016/j.fm.2020.
103631
42. Chiappinelli A, Mangiacotti M, Tomaiuolo M, Trotta G, Marchesani G, Chiaravalle AE (2019)
Identification of X-ray-irradiated hazelnuts by electron spin resonance (ESR) spectroscopy. Eur
Food Res Technol 245(10):2323. https://doi.org/10.1007/s00217-019-03349-2
43. Fan F, Roos YH (2015) X-ray diffraction analysis of lactose crystallization in freeze-dried
lactose-whey protein systems. Food Res Int 67:1. https://doi.org/10.1016/j.foodres.2014.10.023
44. Fan F, Roos YH (2015) X-ray diffraction analysis of lactose crystallization in freeze-dried
lactose–whey protein systems. Food Res Int 67:1. https://doi.org/10.1016/j.foodres.2014.
10.023
45. Wijayasinghe R, Bogahawaththa D, Chandrapala J, Vasiljevic T (2020) Crystallization behav-
ior and crystal properties of lactose as affected by lactic, citric, or phosphoric acid. J Dairy Sci
103(12):11050. https://doi.org/10.3168/jds.2020-18375
46. Mousavi R, Miri T, Cox PW, Fryer PJ (2005) A novel technique for ice crystal visualization in
frozen solids using X-ray micro-computed tomography. J Food Sci 70
47. Falcone PM, Baiano A, Zanini F, Mancini L, Tromba G, Dreossi D, Montanari F, Scuor N,
Nobile MAD (2005) Three-dimensional quantitative analysis of bread crumb by X-ray
microtomography. J Food Sci 70(4):E265. https://doi.org/10.1111/j.1365-2621.2005.tb07182.x
48. Mendoza F, Verboven P, Mebatsion HK, Kerckhofs G, Wevers M, Nicolai B (2007) Three-
dimensional pore space quantification of apple tissue using X-ray computed microtomography.
Planta 226(3):559. https://doi.org/10.1007/s00425-007-0504-4
49. Lim KS, Barigou M (2004) X-ray micro-computed tomography of cellular food products. Food
Res Int 37(10):1001. https://doi.org/10.1016/j.foodres.2004.06.010
50. Pashkova GV, Smagunova AN, Finkelshtein AL (2018) X-ray fluorescence analysis of milk
and dairy products: a review. TrAC Trends Anal Chem 106:183. https://doi.org/10.1016/j.trac.
2018.06.014
51. Oliveira LB, dos Santos WPC, Teixeira LSG, Korn MGA (2020) Direct analysis of cocoa
powder, chocolate powder, and powdered chocolate drink for multi-element determination by
energy dispersive X-ray fluorescence spectrometry. Food Anal Method 13(1):195. https://doi.
org/10.1007/s12161-019-01565-w
52. Guelpa A, du Plessis A, Kidd M, Manley M (2015) Non-destructive estimation of maize (Zea
mays L.) Kernel hardness by means of an X-ray micro-computed tomography (μCT) density
calibration. Food Bioprocess Tech 8(7):1419. https://doi.org/10.1007/s11947-015-1502-3
53. Kotwaliwale N, Singh K, Kalne A, Jha SN, Seth N, Kar A (2014) X-ray imaging methods for
internal quality evaluation of agricultural produce. J Food Sci Technol 51(1):1. https://doi.org/
10.1007/s13197-011-0485-y
Radioactivity
14

Nuclides (also: atomic nuclei) are characterized by their number of protons (atomic
number) and neutrons. Numerous proton–neutron combinations cause nuclides to be
unstable and convert into other nuclides by emitting characteristic radiation. This
property of spontaneous conversion is called radioactivity. Of the chemical elements
with numbers from 1 to 112, more than a thousand different nuclides are known, the
vast majority is unstable, i.e. radioactive. The nuclides with an atomic number above
83 are consistently unstable.

Definition
A nuclide is an atomic nucleus with a specific number of protons and neutrons.
Unstable nuclides are called radionuclides.

Unstable nuclides found in nature lead to the natural radioactivity of materials. In

the field of biological materials, 14 40
6 C and 19 K belong to the naturally occurring
radionuclides. Therefore, foods made from plant or animal raw materials have a
measurable natural radioactivity. In addition, technical radionuclides such as 137
55 Cs
can occur in food as unwanted contamination.

Example
The designation 14
6 C identifies the nuclide with atomic number 6 and atomic
mass number 14. The nuclide contains 6 protons (atomic number) and
8 neutrons (6 + 8 = 14).

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 523
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_14
524 14 Radioactivity

Definition
Isotopes are nuclides with the same atomic number, e.g. 12 13
6 C and 6 C. Since
both nuclides have the same electron configuration, they both behave like the
same chemical element, in this example here carbon. From Greek: ισo, τóπoς,
same place.

In order to be able to estimate the natural radiation exposure emanating from

food, we start this chapter with a look at different types of radioactive decay.

14.1 Types of Radiation

The emission of particles consisting of 2 neutrons and 2 protons is called α-radiation.

These so-called α-particles ejected during decay are positively charged particles with
comparatively low speed and reach. Emission of electrons and positrons is referred
to as β-radiation. β-particles have a speed in the order of magnitude of the speed of
light and correspondingly a long reach. γ-radiation consists of quanta of high energy,
it is electromagnetic radiation with frequencies above X-rays. In addition, there is the
emission of neutrons from the atomic nucleus, the so-called neutron radiation
(Table 14.1).
Radioactive substances that predominantly emit α-radiation are called α-
radiators, analogously one speaks of β-radiator and γ-radiator. Since incoming rays
produce secondary particles as they pass through the material, radioactive decay
often results in a mixture of different types of radiation. In addition to the types of
radiation shown in Table 14.1, in connection with artificial radioactivity emission of
protons, positrons and neutrinos can occur.
The reach of radiation depends on interaction with the material in question. Since
α-particles are highly ionizing to their immediate environment, they have only a very
short reach. β-particles can form secondary ions at a greater distance, depending on

Table 14.1 Types of radioactive radiation

Electric
Designation Description charge Example
α-radiation Emission of slow He 2+ α (Radium), 4.9 MeV
nuclides Reach in air: some centimeters
β-radiation Emission of fast electrons 1- β 40
19 K 1.4 MeV
or positrons or Reach in air: some meters
1+
γ-radiation Emission of high-energy 0 Wavelength < 10 pm, few keV to few
γ-quanta MeV per quant
Reach in air: some hundreds m
Neutron Emission of fast or slow 0
radiation neutrons
14.1 Types of Radiation 525

their energy or velocity. γ-quants have a speed close to that of light or X-rays and,
depending on their energy, have an extremely long reach. Reach of the different
types of radiation thus depends on the absorption ability of the environment for the
respective type of radiation. Reaches of respective types of radiation in air, water,
metal, or biological tissue are completely different. For characterization of γ-radia-
tion or related radiation protection materials, the so-called penetration depth of
radiation can be specified as the length after which its intensity has fallen to
1/e = 36.7%. Or the length after which the intensity has fallen to 50% (so-called
half-value depth).

14.1.1 Activity and Decay

Radioactive transformation of a nucleus occurs spontaneously, i.e. without any

external influence. Therefore, the temporal prediction of decay in nuclei cannot be
carried out for individual cases, but only with statistical methods for a large number
of atomic nuclei. If we look on a large number of nuclei, we see that the decay rate is
proportional to the number of existing nuclei. Gradually, this number of remaining
nuclei decreases and so does the rate of decay and intensity of radioactive radiation.
Since radioactive radiation is produced during each individual decay, substances
radiate more strongly when many nuclei decay per second. A high rate of activity is
expressed by specifying a high decay constant λ. Substances with a low decay
constant, on the other hand, decay very slowly, therefore show only a comparatively
low level of radioactive radiation and it takes a very long time for their activity to
become immeasurably small. Activity has the unit Becquerel, 1 Bq = 1 s-1. An
activity of 103 Bq means that 1000 decays take place per second.
With the assumption that the decay rate dN dt is proportional to the number of
existing nuclei N, the decay law is

dN
- =λ
N ð14:1Þ
dt
with the decay constant

- dN
λ= dt
ð14:2Þ
N
The activity is thus

dN
A= - =N
λ ð14:3Þ
dt
If we integrate the decays over time,
526 14 Radioactivity

N t
dN
- =λ dt ð14:4Þ
dt
N0 t=0

i.e.,

- ðln N - ln N 0 Þ = λ
ðt- 0Þ ð14:5Þ

or

N
ln = -λ
t ð14:6Þ
N0
we get an exponential relationship:

N = N 0
e - λ
t ð14:7Þ

N number of atomic nuclei

N0 initial number of atomic nuclei
A activity in Bq
λ decay constant in s-1
T1/2 half-life in s, d, y
m mass in kg
M molar mass in kg∙mol-1
n amount of substance in mol
NA Avogadro’s constant NA = 6.02214076 ∙ 1023 mol-1

The half-life T is the time after which half of all originally existing nuclei of the
material under consideration have decayed. This is also the time after which the
activity has decreased to half the value of the initial activity. It results from:

N0
N= ð14:8Þ
2
to

N 1
- T 1=2
λ = ln = ln ð14:9Þ
N0 2
that is

ln 2
T 1=2 = ð14:10Þ
λ
Table 14.2 shows some examples of half-lives.
14.1 Types of Radiation 527

Table 14.2 Examples of Nucleus T1/2 λ/Bq

decay constants λ and half- 40
19 K
1.25 ∙ 109 a 1.8 ∙ 10-17
lives T1/2 of nuclei [1]
14
6 C
5730 a 3.8 ∙ 10-12
226
88 Ra
1580 a 1.4 ∙ 10-11
137
55 Cs
30 a 7.3 ∙ 10-10
90
38 Sr
28.1 a 7.8 ∙ 10–10
131
53 I
8.05 d 1 ∙ 10-6

Substances with high activity, i.e. having a high decay constant λ, therefore have
a short half-life. Conversely, substances with low activity have a long half-life.
A quantity derived from the activity is the so-called specific activity. It is the ratio
of the activity A of a substance and its mass m. In the case of an isotope-pure sample
of the mass m with N radioactive nuclei, the specific activity can be calculated as
follows:

A λ
N λ
n
NA NA
a= = = =λ
ð14:11Þ
m m n
M M
The specific activity has the SI unit Bq
kg-1. If materials are mixtures of
radioactive and non-radioactive isotopes, the quotient of activity and total mass of
the mixture is referred to as specific activity.

Example
88 Ra is 3.7 ∙ 10
The specific activity of the nuclide radium 226 13
Bq
kg-1 .This
means that 1 g of this substance has an activity of 3.7 ∙ 10 Bq. The former
10

unit of activity Curie (Ci) corresponds exactly to this value: 1 g of the nuclide
88 Ra has the activity 1 Ci = 3.7 ∙ 10 Bq
226 10

Radioactive conversions are associated with a high energy release, which often is
not seen because it occurs over long periods of time (Table 14.2).

Example
The decay of 1 g radium, including all radioactive derivatives, leads to a heat
dissipation to the environment of 0.2 W. Over the course of a half-life of
1580 years, this corresponds to a heat amount of

Q = 0:2 W
3600 s
24
365
1580 = 107 kJ:

For comparison: The combustion of 1 g of coal provides 32 kJ, the calorific

value of 1 g of sucrose is 17 kJ (Chap. 8).
528 14 Radioactivity

14.1.2 Measurement of Ionizing Radiation (a-, b-, g-)

The detection and quantitative determination of radioactive radiation is carried out

by recording the effects generated by the radiation on ionization detectors which
determine the number of ions generated.
To distinguish between α-, β-, and γ-radiation, the radiation can be filtered before
entering an ionization detector. For this purpose, electric and magnetic fields are
used. The Coulomb force of an electric field causes a deflection of electrically
charged α-particles and β-particles, but does not affect γ-quanta (Fig. 14.1). In the
electric field, electrons (β-) and positrons (β+) can also be separated. The Lorentz
force of a magnetic field supports the differentiation according to charge number,
sign of charge, and particle speed (Fig. 14.2).
Another possibility is selective absorption with suitable materials. While α-
radiation can be shielded by a sheet of cardboard, a rough distinction between β-
and γ-radiation can be made with metal plates of suitable thickness, depending on the
energy of the radiation.
As a result of the interaction of γ-rays with matter, the photo effect or the
Compton effect occurs. Both are associated with the release of an electron from
the atomic shell of the absorbent material. These so-called secondary ions can be
used to detect γ-radiation. Another effect is the so-called pair formation, where an
electron–positron pair is formed when a γ-quant hits a material. The reverse effect is
called pair annihilation, here an electron–positron pair recombines to a γ-quant.

Definition
Wave-particle dualism: γ-radiation is an electromagnetic wave with typical
frequency and wavelength. The radiation occurs quantized, a quantum has the
energy E = h
f. These γ-quants behave like high-energy particles.

Fig. 14.1 Separation of

radioactive radiation in the J ,n
electric field


E E

S
14.1 Types of Radiation 529

Fig. 14.2 Deflection of

radioactive radiation in the
magnetic field perpendicular
to the direction of movement,
schematic. Low-weight
electrons β--and positrons
β-+ are more strongly
deflected than heavier
α-particles. γ-quants and
neutrons (n) are not affected

14.1.2.1 Counter Tubes

In a mixture of noble gases such as Helium, Neon, and Argon, ions are produced by
inciting radioactive radiation. In the strong electric field of the counter tube these
primary ions generate further ions (secondary ions) which cause a short-term
measurable electric pulse via the photoelectric effect. If the electrical field in the
counter tube is not too high, the number of detected pulses is proportional to the
number of primary ions. A counter tube operated in this way is called a proportional
counter tube. A well-known design is the Geiger-Müller counter tube.

14.1.2.2 Semiconductor Detectors

In semiconductor detectors, the ions generated by radioactive radiation lead to
countable voltage pulses. For example, ultrapure Germanium or pn-junctions in
Lithium-doped Germanium are used for γ-radiation and Li-doped silicon detectors
for X-rays. With semiconductor detectors it is possible to determine the energy of the
γ-quants. This makes it possible not only to detect nuclear decays and to count them
across the board, but also to get an energy spectrum of the radiation. Figure 14.3
shows such a spectrum schematically.
With the help of such spectra, conclusions can be drawn about the type and source
of the radiation. If radioactive contamination of food is found, the clarification of the
type and origin of the radionuclides is an essential question. In scintillation detectors,
the radiation generates a light pulse. The light emission is amplified and counted. As
scintillators, crystals are used (e.g., anthracene for β-radiation, ZnS/Ag for
530 14 Radioactivity

Fig. 14.3 Energy spectrum

of radioactive radiation:
Intensity of the absorbed γ-
quants with respect to their
energy E

Table 14.3 Isotopes of Isotope Natural prevalence in % T1/2

potassium and their half- 39
19 K
93.2581 Stable
lives T1/2 [2]
40
19 K
0.0117 1.25 ∙ 109 a
41 6.7302 Stable
19 K
42 0 12.36 h
19 K
43 0 22.3 h
19 K

α-radiation, NaI/Tl for γ-radiation). Lithium containing glasses are used as scintilla-
tor material for slow neutrons.
Liquid scintillators can be mixed with the sample and measured in cuvettes.
Because of low absorption losses between sample and scintillator, this is advanta-
geous for low-energy radiation (e.g., 14C).

14.1.3 Natural Radioactivity

Natural potassium consists mainly of the stable isotopes 39 41

19 K and 19 K and to a very
40 42 43
small extent of the unstable, i.e. radioactive, isotopes 19 K, 19 K, 19 K. In Table 14.3,
the natural prevalence of K- isotopes is listed. Because of the short half-life of 4219 K
and 43
19 K these isotopes have already completely decayed on the planet. Since they are
hardly newly formed, their occurrence is given as 0% in Table 14.3.
The metabolism of humans needs Potassium. The Potassium requirement can be
completely covered by food. The radioactive natural isotope 40K enters our body by
intake of Potassium. Due to a balance between the intake of such natural
radionuclides from food or air and their biological excretion, the value of natural
radioactivity in the human body is largely constant, also in animals, plants, and
foods.
14.1 Types of Radiation 531

Example
Specific activity of Potassium: Natural Potassium consists of 0.01117% radio-
active isotope 40K with a half-life of 1.28 billion years. The number N of atoms
in a 1 g potassium (molecular weight M = 39.098 g mol-1) is

1g
N = n
NA =
6:022
1023 atoms mol - 1 = 1:54
39:098 g mol - 1

1022 atoms

Of these, 0.0117% are 40K isotopes, i.e.

Nð40 KÞ = 0:000117
1:54
1022 = 1:8
1018

The decay constant of 40K is:

ln 2 ln 2
λ= = = 1:72
10 - 17 s - 1
T 1=2 1:277
109 a

The activity of one gram thus results in:

A=λ
N
A = 1:72
10 - 17 s - 1
1:80
1018 = 31:0 s - 1

The specific activity of Potassium according to this approximation

a = 31 Bq
g-1 is essentially caused by the isotope 40K. This estimation
neglected the fact that only about 90% of the 40K isotopes emit β-radiation and
pass into 40Ca. About 10% of the 40K isotopes are converted to 40Ar by β+
-emission and electron capture.

The Potassium content of the human body depends on age, gender, and diet. It is
about 2 g per kg body weight. A more precise calculation is possible by means of the
following empirical formulas:

Potassium content c in g per kg body weight

Men c = 2.38658 - (0.00893
age)
Women c = 1.9383 - (0.00675
age)

Example
A 30-year-old woman has a Potassium content of
c = 1.9383 - (0.00675
30) = 1.7g Potassium per kg body weight.
532 14 Radioactivity

Table 14.4 Activity of Nuclide Activity in Bq

natural radionuclides in 40
K 4000
humans [4] 14
C 3800
87
Rb 650
210
Pb <100
210
Bi <100
210
Po <100
3
H <100
7
Be <100
Rn decay products <100

Table 14.5 specific 40K Material a/Bq kg-1

activity a of some materials,
Human body 40–70
examples [5, 6]
Milk 40–60
Milk powder 400–500
Fruits, vegetables 30–150
Fruit juice concentrate 600–800
Hazelnuts 190–270
Honey 30–200
Instant coffee powder 1000
Potash fertilizer 16,000

With an average Potassium content of the human body (1.7–2.2 g/kg body
weight), an average body weight results in an activity of about 4000 Bq of the
human body, which is only due to the isotope 40K [3]. Due to other radionuclides
occurring in the environment, the total activity of the human body is about twice as
high, Table 14.4 lists the orders of magnitude of the individual radionuclides.
Due to the Potassium content of plant and animal foods, the human body contains
a natural content of Potassium and thus also of the nuclide 40
19 K. Table 14.5 shows the
magnitude of the 40 19 K specific activities of some materials in comparison. When
calculating data like this, attention must be paid to whether the activity relates to the
weight or to the dry matter of the material.

Example
Estimation of the natural radioactivity of spinach via the potassium content:
Spinach contains about 0.5 g Potassium per 100 g [7]. With the specific
activity of potassium.
a = 31.0 Bq ∙ g-1 it is A = 31.0 Bq ∙ g-1 ∙ 0.5g ∙ 100 g-1 = 15.5 Bq ∙ 100 g-1.
14.1 Types of Radiation 533

Attention
For example, a 100 g sample of spinach leaves will have an activity of 16 Bq.
If the radioactivity measurement is carried out with a dried sample, a 100 g
sample shows an activity of 180 Bq. The first measurement referred to 100 g of
weight, the second to 100 g of dry matter. The water content of the sample
before drying was 91.2% (m/m). In order to avoid communication errors, it is
advisable to specify precise information, such as “180 Bq per 100 g dry
matter.”

14.1.3.1 Exposure to the Human Body

For calculation of the radiation exposure of humans, the energy absorbed by the
tissue of the body (absorbed dose) must be determined as follows:
absorbed energy of radiation
absorbed dose of ionizing radiation =
mass
The absorbed dose of ionizing radiation has units of J∙kg-1. One J∙kg-1 = 1 Gy
(Gray). Conversion from the previously used unit of rad (radiation adsorbed dose)
into the unit of Gray is simply a factor of 100. One rad = 10-2 Gy, or one Gy = 100
rads. For conversion of units, see also Appendix. In addition to the amount of
absorbed energy from exposure to radiation, it is just as important to take into
account the kind of radiation which is being absorbed. For this purpose, a
multiplying factor called a quality factor has been introduced (see Table 14.6).
The absorbed dose of ionizing radiation is multiplied by this quality factor in order
to calculate the dose equivalent, as follows:

dose equivalent = qualityfactor
adsorbed dose of ionizing radiation

The dose equivalent of the received dose of radiation reflects the effect of the
adsorbed dose of ionizing radiation on the human body, regardless of the type of
radiation from which it came (X-ray, γ-ray or β-radiation).
The unit of the dose equivalent is Sv (Sievert). Conversion from the previously
used unit of rem (roentgen equivalent man) into the new unit of Sv is simply a factor
of 100. One rem = 10-2 Sv, or one Sv = 100 rem (further quantities for use in
dosimetry are given in Table 14.7).

Table 14.6 Quality Radiation Quality factor

factors for different types of
Photons (X-ray, γ-radiation) 1
radiation [8]
Electrons, positrons (β-radiation) 1
Neutrons
10 keV 5
10–100 keV 10
100 keV–2 MeV 5
2–20 MeV 20
Protons 5–15
α-particles 20
534 14 Radioactivity

Table 14.7 Terms used in dosimetry

Quantity Definition SI unit Equivalent with
Activity number decays Bq s-1
time
Specific activity number decays Bq∙kg-1 s-1 ∙kg-1
mass
time
Energy J, eV
Absorbed dose of ionizing radiation absorbed energy Gy J∙kg-1
mass
Power of absorbed dose of ionizing radiation absorbed energy Gy∙s-1 J∙kg-1 s-1
mass
time
Dose equivalent absorbed energy
Q Sv J∙kg-1
mass
Power of dose equivalent absorbed energy
Q Sv∙s-1 J∙kg-1 s-1
mass
time
-1
Dose of ionization absorbed charge C∙kg A∙s∙kg-1
mass
Power of dose of ionization absorbed charge C∙kg-1∙s-1 A∙kg-1
mass
time

Contributions to the natural radiation exposure of humans are:

cosmic radiation
terrestrial radiation (soil)
ingestion (food)
inhalation (breathing air)

Average natural radiation exposure to humans thus depends on their geographical

location (altitude, soil, air quality) and way of life, including diet and age. A value
1 - 10 mSv
a-1 can be taken as an order of magnitude for the average natural
radiation exposure to an adult [9]: In Germany, the average value for natural
radiation exposure is about 2 mSv
a-1 from which about 0.2 - 0.4 mSv
a-1
comes from food, including 0.165 mSv
a-1 from Potassium [5]. The magnitude of
the contribution of the natural radioactivity of the air we breathe (222Rn) is
1 mSv
a-1 and varies with local conditions. In addition to natural radiation
exposure, there is the so-called civilizational radiation exposure, e.g. through occu-
pational exposure, medical diagnostics, or air travel.
Sensors that work on basis of the absorption of radioactive radiation use
fabricated radiation sources, such as 63Ni foils. For example, the thickness of a
sample can be determined by measuring the absorption of radioactive radiation.
Some radiation sources are listed in Table 14.8.

14.1.3.2 Irradiation of Food and Packaging Material

By irradiating biological materials with γ-radiation or β-radiation, enzymes, viruses,
fungi, pathogenic germs, and pests can be inactivated and thus the shelf life of food
and agricultural products can be increased. Table 14.9 shows the magnitude of the
radiation doses used here and their effect. Details on the technology and legal
regulations can be found at [10–16]. When irradiating packaged food, possible
effects on the physio-chemical properties of the packaging material must also be
considered.
14.1 Types of Radiation 535

Table 14.8 Measurement methods with radioactive emitters

Measurement Example Radiator e.g.
β-transmission On-line coating thickness measurement of paper, 90
Sr
204
polymers, metals Tl
147
Pm
γ-absorption On-line coating thickness measurement of thick 241
Am
60
samples such as glass, metal, plastic on conveyor Co
137
belts. On-line level measurement Cs
β- Thickness measurement of thin films (few μm) on a 90
Sr
204
backscattering substrate or metal coatings Tl
γ-backscattering Thickness measurement of glass, plastic, light metal 137
Cs
241
alloys Am
210
Static Removal of static charges from electrical Po
elimination components, films, films, etc. with air-ionizing
(dischargers) nozzles (packaging technology, printing industry)
63
Electron capture Used in gas chromatography. The detector detects a Ni
(ECD) stream of electrons, which is altered by analytes
β-radiography Making watermarks visible, making distribution of a 14
C Polymethyl-
drug in an organ visible by radiolabeling the drug methacrylate
(PMMA)
γ-radiography Non-destructive inspection of welds 60
Co
192
Ir
252
n-radiography Neutron scattering for non-destructive investigation Cf
of materials with low mass numbers (turbine blades,
valve parts, bondings)
60
Radiotherapy Tumor control, sterilization of medical devices, Co
agricultural products, food

Table 14.9 Effects of ionizing radiation [12, 18]

Dose in kGy Effect of radiation
0.005–0.01 Lethal to humans
0.05–0.15 Germination inhibition in potatoes, onions, garlic, ginger roots
0.15–0.75 Pest control (insects, parasites)
0.25–1 Delay in ripening of fruit and vegetables
1–10 Inactivation of pathogenic microorganisms (pasteurization)
10–50 Sterilization
10–200 Inactivation of viruses
20–1000 Inactivation of enzymes

The detection of irradiation can be done with the help of electron spin resonance
(ESR) spectroscopy. The method of ESR spectroscopy is explained in Chap. 10. It is
based on the detection of radicals produced in the material by ionizing irradiation.
The lifetime of these radicals is short in aqueous materials, but in dry materials such
as bones or cardboard packaging the lifetime may be long enough for radicals to be
detected [17].
536 14 Radioactivity

14.1.3.3 Industrial Applications of Irradiation in Food Processing

Doses of radiation less than 10 Gy can be used to destroy microorganisms in food
products without using preservatives or other chemicals. When this is done, the food
is neither left contaminated nor radioactive from the process, nor is the quality or
functionality of the food altered in any way.
When a food is exposed to a dose of ionizing radiation, pathogenic and food
spoilage microorganisms are destroyed by molecular bombardment from either β-
radiation (high-speed electrons) from electron beams or γ-radiation (high-energy
photons) from radioactive isotopes.
Electron beams that emit β- radiation are machine-generated through linear
accelerators controlled by an on–off switch. The limited penetration depth of β-
radiation makes this type of radiation treatment suitable only for relatively small
food items as they pass beneath an electron beam on a conveyor belt. In contrast, γ-
radiation is highly penetrating and can be used to treat large pallet loads of
prepackaged foods when they are exposed to a radioactive isotope, such as Cobalt
60 or Cesium 137. The disadvantage of γ-radiation is that it cannot be turned on and
off. Instead, exposure to the γ- radiation is controlled by shielding the radioactive
isotope under 4 m of water, 2 m of concrete, or 0.15 m of lead.
Contract irradiation facilities now constitute an important sector of the food
processing industry. Some of these facilities provide electron beam irradiation
services for β-radiation treatments, and others provide γ-radiation treatment from
radioactive isotopes. These facilities are used, e.g., by meat processors who ship
prepackaged frozen hamburgers to fast food restaurants. Delivery trucks stop at
contract radiation facilities to have the frozen hamburgers treated by γ-radiation to be
sure they are free of any food-borne pathogenic microorganisms. Pallet loads of
frozen hamburgers are unloaded from the trucks and conveyed through the facility
for exposure to the radioactive isotope. The trucks are then reloaded and continue on
to complete their delivery.

Further Reading

ESR detection of irradiated foods containing crystalline sugars [19]

ESR detection of irradiated foods containing fat [20]
Foodstuffs—Thermoluminescence detection of irradiated food from which [21]
silicate minerals can be isolated
ESR spectroscopy for detection of irradiated bone or bone-containing foods [22]
ESR detection of irradiated food containing cellulose [23]
Detection of irradiated food using photostimulated luminescence [24]
Preservation of food by ionizing radiation [10, 11]
Influence of γ-irradiation on food properties [13]
γ-sterilization of food [15]
Irradiation methods for meat and fish [11]
Positron emission tomography (PET) for investigation of metabolic processes [25, 26]

(continued)
References 537

Positron emission tomography (PET) to examine obese persons during tasting [27]
Positron anihilation lifetime spectroscopy (PALS) for the characterization of [28]
amorphous materials

Summary
Radioactivity occurs in food in the form of natural radioactivity or in the form
of contamination. In this chapter, the basic concepts of radiation types,
radioactive decay, and the effect of radioactive radiation are explained in
simple words and underpinned by calculated examples. By compiling the
quantities of dosimetry and its units, the magnitude of natural and civiliza-
tional radiation exposure can be understood. At the end of the chapter,
numerous examples of measurement methods for radioactive irradiation are
listed, which are intended to encourage further study in this field.

References
1. Lieser KH (1992) Einführung in die Kernchemie. Weinheim
2. Lexikon der Physik (1998) Spektrum Akademischer Verlag, Heidelberg
3. Umweltradioaktivität und Strahlenbelastung - Jahresbericht 2016 (2016). Bundesministerium
für Umwelt, Naturschutz, Bau und Reaktorsicherheit (BMUB), Bonn
4. BfS (2020) Natural radioactivity in food. https://www.bfs.de/EN/topics/ion/environment/
foodstuffs/radioactivity-food/radioactivity-food.html. Accessed 2020-12-17
5. LGL (2020) Bayerisches Landesamt für Gesundheit und Lebensmittelsicherheit. https://www.
lgl.bayern.de/lebensmittel/chemie/kontaminanten/radioaktivitaet/
6. Stansfield CM (2003) Legislation | contaminants and adulterants. In: Caballero B
(ed) Encyclopedia of food sciences and nutrition. Academic, Oxford, pp 3507–3513. https://
doi.org/10.1016/B0-12-227055-X/00689-1
7. Souci SW, Fachmann W, Kraut H (2016) Die Zusammensetzung der Lebensmittel, Nährwert-
Tabellen. Wissenschaftliche Verlagsgesellschaft (WVG), Stuttgart
8. Krieger H (2004) Grundlagen der Strahlungsphysik und des Strahlenschutzes. Springer,
Wiesbaden. doi:https://doi.org/10.1007/978-3-8348-2238-3
9. UNSCEAR (2000) Sources and effects of ionizing radiation (2000). United Nations, New York
10. Ehlermann DAE (2016) Wholesomeness of irradiated food. Radiat Phys Chem 129:24. https://
doi.org/10.1016/j.radphyschem.2016.08.014
11. Ehlermann DAE (2016) Particular applications of food irradiation: meat, fish and others. Radiat
Phys Chem 129:53. https://doi.org/10.1016/j.radphyschem.2016.07.027
12. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, München
13. Jan K, Bashir K, Maurya VK (2021) Gamma irradiation and food properties. In: Knoerzer K,
Muthukumarappan K (eds) Innovative food processing technologies. Elsevier, Oxford, pp
41–60. https://doi.org/10.1016/B978-0-08-100596-5.23052-7
14. Kocol H (2001) Radioactivity in food and water. In: Hui YH, Kitts D, Stanfield PS (eds)
Foodborne disease handbook. CRC Press, Boca Raton. https://doi.org/10.1201/9781351072113
538 14 Radioactivity

15. Feliciano CP (2018) High-dose irradiated food: current progress, applications, and prospects.
Radiat Phys Chem 144:34. https://doi.org/10.1016/j.radphyschem.2017.11.010
16. Roberts PB (2016) Food irradiation: standards, regulations and world-wide trade. Radiat Phys
Chem 129:30. https://doi.org/10.1016/j.radphyschem.2016.06.005
17. Marchioni E (2006) ESR as a technique for food irradiation detection. In: Webb GA
(ed) Modern magnetic resonance. Springer, Dordrecht, pp 1855–1860. https://doi.org/10.
1007/1-4020-3910-7_211
18. Ehlermann DAE (1999) Die Strahlenkonservierung von Lebensmitteln. Bundesanstalt für
Ernährung, Karlsruhe. https://doi.org/10.5445/IR/128199
19. EN 13708 Foodstuffs- Detection of irradiated foodstuff containing crystalline sugar by ESR
spectroscopy (2022) Beuth, Berlin
20. EN 1785 Foodstuffs - Detection of irradiated food containing fat - Gas chromatographic/mass
spectrometric analysis of 2-alkylcyclobutanones (2003) Beuth, Berlin. https://doi.org/10.31030/
9515264
21. Deutsches Institut für Normung (2002) DIN EN 1788 Foodstuffs - Thermoluminescence
detection of irradiated food from which silicate minerals can be isolated. Beuth, Berlin.
https://doi.org/10.31030/9139521
22. DIN EN 1786 Foodstuffs - Detection of irradiated food containing bone - Method by ESR
spectroscopy (1997) Beuth, Berlin. https://doi.org/10.31030/7311426
23. DIN EN 1787 Foodstuffs - Detection of irradiated foodstuff containing cellulose by ESR
spectroscopy (2022) Beuth, Berlin
24. Detection of irradiated food using photostimulated luminescence (2022) European Commis-
sion, Directorate-General for Health and Food Safety, Brussels
25. Coenen HH, Ermert J (2020) Expanding PET-applications in life sciences with positron-
emitters beyond fluorine-18. Nucl Med Biol. https://doi.org/10.1016/j.nucmedbio.2020.07.003
26. Knuuti J (2004) Positron emission tomography—molecular imaging of biological processes. Int
Congr Ser 1265:248. https://doi.org/10.1016/j.ics.2004.06.014
27. DelParigi A, Chen K, Salbe AD, Reiman EM, Tataranni PA (2005) Sensory experience of food
and obesity: a positron emission tomography study of the brain regions affected by tasting a
liquid meal after a prolonged fast. NeuroImage 24(2):436. https://doi.org/10.1016/j.
neuroimage.2004.08.035
28. Martini F, Hughes DJ, Badolato Bönisch G, Zwick T, Schäfer C, Geppi M, Alam MA, Ubbink J
(2020) Antiplasticization and phase behavior in phase-separated modified starch-sucrose
blends: A positron lifetime and solid-state NMR study. Carbohydr Polym 250:116931.
https://doi.org/10.1016/j.carbpol.2020.116931
Acoustic Properties
15

Acoustic properties determine the behavior of a material when interacting with

sound waves. Sound consists of periodic pressure fluctuations that propagate in the
form of longitudinal waves. Sound frequencies from 16 Hz to 16 kHz are perceptible
to the human ear, lower frequencies (infrasound) and higher frequencies (ultrasound)
are not audible to humans (Fig. 15.1, Table 15.1). The lower and upper limits of
perception are slightly different from individual to individual.
In addition to gases such as air, sound waves can also propagate in solids and
liquids. Depending on the material, this leads to attenuation of the amplitude, to a
change in the propagation speed, or to a phase shift. Measurement of these quantities
can provide information about the material in question, such as composition or
quality. For example, the crispness of baked goods can be acoustically characterized
[1] or the particle size in disperse systems [2]. There are ultrasound-based sensors in
food technology such as distance sensors and flow sensors (Chap. 16). High-power
ultrasound can be used to support processes such as the inactivation of
microorganisms [3] or emulsification and homogenization by cavitation [4]. Let’s
start the chapter with the basic concepts of acoustics.

15.1 Sound

Sound waves are propagated through a material medium in the form of mechanical
longitudinal waves. All materials are capable of transmitting sound waves if they
have any degree of elasticity (elastic bonds between molecules). This is true of most
materials (see Young’s modulus of elasticity in Chap. 4). Therefore, sound can be
transmitted through gases like air, as well as through liquids and solids. By multipli-
cation of frequency and wavelength we get the speed of sound in a material.

c=λ
f ð15:1Þ

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 539
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_15
540 15 Acoustic Properties

Fig. 15.1 Frequency ranges: infrasound (I), audible sound (II), ultrasound (III), hyper sound (IV)

Table 15.1 Sound Designation Frequency range

frequencies
Infrasound 0–16 Hz
Audible sound 16 Hz–16 kHz
Ultrasound 20 kHz–10 GHz
Hyper sound 109–1012 Hz

c Speed of sound in m
s-1

λ Wavelength in m
f Frequency in s

Sound waves can be generated by a mechanical disturbance, such as when we

turn on a machine with moving parts that will make the air vibrate at a given
frequency. These air vibrations are the sound waves that travel (propagate) in all
directions. If we are standing nearby, these sound waves will strike our ear drums
and cause them to vibrate at the same frequency because of the oscillating air
pressure. Our human ear senses the vibrations in our ear drums and transmits these
vibrations through the inner ear as nerve signals to our brain. When this happens, we
hear sound. This process also means that there must be a medium of some material
substance (gas, liquid, solid) between the source and the receiver (the machine and
our ear) for sound waves to be generated and transmitted to produce sound. That is
why sound does not exist in a vacuum. When we decrease the air pressure in a room
we can reduce the emission and perception of sound.

15.1.1 Speed of Sound

The speed of sound depends on the coupling or bonding strength between the
oscillating molecules or atoms within the material medium through which the
sound waves are being transmitted. This bonding strength is much greater in solids
than in liquids and greater in liquids than in gases where molecules have little
interaction. For this reason, sound waves travel much faster in solids than in liquids,
and faster in liquids than in gases (Table 15.2). The speed of sound can also depend
on temperature as shown for the speed of sound in air (Table 15.3).
There are different terms by which we characterize the various features of sound.
We will define these here and explain how they are used. First is sound power. This
is the quantity of energy flow transmitted by the source of the sound waves. Typical
sound power levels for various sound sources are listed in Table 15.4.
15.1 Sound 541

Table 15.2 Speed of Material v/m s-1

sound, examples (15 °C,
Iron 5170
101.3 kPa) [5]
Lead 1250
Water 1464
Hydrogen 1284
Oxygen 316
Air 331

Table 15.3 Speed of Material ϑ/ ° C v/m s-1

sound in air, temperature
Dry air -20 319
dependency
Dry air 0 331
Dry air 20 344
Dry air 100 387
Water vapor 130 450

Table 15.4 Typical sound Sound power

power from various sources
Normal talking 10-5 W
[6]
Human voice, maximum 2 × 10–0.3 W
Trumpet 0.3 W
Auto horn 5W
75-piece orchestra 70 W
Loudspeaker at rock concert >102 W
Siren >103 W
Rocket engine 106 W

Table 15.5 Definition and reference quantity of sound levels [9]

Sound level (in dB) Definition Reference quantity
Sound pressure level Lp = 20
lg pp po = 2.10-5Pa
0

Sound velocity level Lv = 20
lg vv0 v0 = 5 × 10-8m s-1

(particle velocity level)
Sound intensity level LI = 10
lg II0 I0 = 10-12W m-2
Sound power level LP = 10
lg PP0 P0 = 10-12W

In acoustics, we often refer to relative levels of sound. These levels are generally
quantified by taking the logarithms of the ratio of a given sound quantity to a
reference quantity. Although different reference quantities can be used, the mini-
mum quantity which a normal human ear can sense is taken as a reference quantity in
most cases. These levels have no units, but they are given the name decibel
(dB) [7]. Table 15.5 shows some definitions of important levels.
Sound intensity is a measure of the amount of sound power per unit area and can
be expressed as the quotient of the incremental sound power dP that is transmitted
542 15 Acoustic Properties

across an incremental area dA, which is perpendicular to the direction of the energy
flow P.

dP
I= ð15:2Þ
dA
Another term, sound pressure, quantifies the amplitude of the pressure
oscillations and is given in SI units of Pa. The propagation velocity of the pressure
oscillations in a material can be calculated from its density and compression modulus
K of the material.

K
c= ð15:3Þ
ρ

For a gas that can be assumed an ideal gas, we can express the compression
modulus K with the help of the isentropic coefficient κ and pressure p.

K =κ
p ð15:4Þ

So the speed of sound in an ideal gas is

κ
p
c= ð15:5Þ
ρ

with the equation of state for ideal gases

p
V = m
RS
T ð15:6Þ

or

p
ρ= ð15:7Þ
RS
T
we get

κ
p p
c=
RS
T = κ
Rs
T ð15:8Þ
p

p pressure in Pa
κ isentropic exponent
K compression modulus in Pa
Rs specific gas constant in J K-1 kg-1
m mass in kg
ρ density in kg m-3
T temperature in K
ϑ temperature in °C
15.1 Sound 543

Example
According to Eq. (15.8) for air at 22 °C with κ = 1.4 and RS = 287 J
kg-1
K-1
we get

c= 1:4
287J
kg - 1
K - 1
295:15 K = 118, 591 N
m
kg - 1

= 118, 591 kg
m2
s - 2
kg - 1

c = 344:3 m
s - 1

As an approximation for the speed of sound in air in the range of -20 °C to 40 °C

from published literature, we can use the following rule of thumb [8].

c=m s - 1 = 331:5 þ 0:6
ϑ= ° C:

15.1.2 Loudness and Volume

The combination of amplitude (sound pressure) and frequency of a sound wave is

what is responsible for what we perceive as loudness of a sound. Sensitivity of the
human ear to loudness of a sound is based on the sound pressure and the frequency of
the sound. The sound pressure level at a frequency of 1000 Hz has been established
as a technical measure for what we call loudness LS. Relative loudness is also
quantified as a loudness level and is scaled in units of phon. So, loudness is:

p
Ls = 20
lg phon ð15:9Þ
p0

At a frequency of 1000 Hz the value of sound pressure level and loudness are the
same. The human ear may be more sensitive at another frequency, such that a lower
sound pressure level at that frequency could cause the same perception of loudness.
This phenomenon can be illustrated in equal-loudness contours which have been
established in international standards for average persons.
Because loudness of sound is a matter of human perception, different types of
sounds or noises are sensed differently. The so-called weighing curves have been
introduced to take these differences into account when measuring loudness. For
example, the A-weighting curve is used to measure environmental noise and indus-
trial noise, B- and C- curves for louder sounds, and the D-curve for assessing loud
aircraft noise. Giving a loudness with the unit db(A) indicates that weighting curve A
was used [7].
Examples of loudness levels are listed in Table 15.6.
544 15 Acoustic Properties

Table 15.6 Values of Ls/phon

loudness levels
Detection threshold 0
Ticking watch 12
Backyard garden 20
Normal speech (1 m) 65
Heavy traffic (7 m) 80
Automobile horn 90
Jet aircraft (200 m) 115
Pain limit 130

It should be noted that the minimum sound pressure p0 that can be sensed by a
human ear (which is used as a reference in calculating the loudness level) also
depends on the frequency of the sound waves, as well as the age and health of the
person.

15.1.3 Noise

Up to this point, we have learned to understand sound waves as simple oscillations of

air pressure with a given frequency, wavelength, and speed. When simple sound
waves of this type reach our ear, we perceive the sound to be like a tone in music.
Several distinct tones being received at the same time together is perceived as what
we call a clang. When we receive sound which contains a continuous range of
frequencies, it is called noise.
Excessive noise is that type of noise which is disturbing and can impair our
health. Low frequencies enhance the feeling of being disturbed by noise. Permanent
high levels of sound pressure can lead to increasing the minimum sound level a
person can hear (audibility) at different frequencies and can result in losing the
perception of noises at all. Such dangerously permanent high levels of sound
pressure can occur in food processing factories where heavy machinery is operating
continuously. These types of industrial situations require hearing protection for the
factory workers. This is often accomplished by design of machinery housing,
automation to the extent that workers are not needed on the factory floor or by
having the workers wearing ear protection.
A noise can be typical and expected for a given process and can be used to
analyze or to characterize the process. For example, noises like cracking, crushing,
popping, crackling, sizzling, frizzling, and detonating can be an expected character-
istic for a given process or event. Such noises are time-dependent combinations of
tones with different intensities.
15.3 Ultrasonic Sound 545

15.2 Acoustic Quality Assessment

When we wish to characterize a food material for purposes of consumer perception,

we sometimes use terms about the acoustic information provided by taste panelists
during taste testing. Imagine if you bite on a crisp or crusty food and try to describe
the crispness: Our perception of crispness is often taken from the sound that is made
in our mouth. A person without an acoustic sense would not be able to describe food
like this. Such a person would have to rely on dental sensations coming from the
teeth breaking or crushing the food item. Unfortunately, our dental roots are not very
sensitive to such relatively mild forces in crushing foods, they normally start to
respond when we bite on a stone or bone.
So, perception of sound can be used as a descriptor of the textural quality of food.
For example, the sound we hear when biting on food may give us information about
the freshness of vegetables or about under or over cooked, they are. To develop
methods for acoustic characterization of a material, first the process of biting,
breaking, chewing, shaking, etc. must be standardized, and then the resulting noise
has to be measured and analyzed for its acoustic spectrum which is a plot of intensity
versus frequency.
The acoustic spectrum of a mechanical quality test can be used to quantify
sensory perceptions as crisp, crusty, crunchy, etc. So in instrumental texture analysis
from the noise, we can gain information about the quality [10].

Example: Chewing Noise

When chewing lettuce leaves, we can assess whether the lettuce is “fresh” or
“already older.” The assessment is based on the noise that arises on the tooth
surfaces and is transmitted into the hearing.
When chewing heated onion rings, we can tell from the “bite” of the onion
ring whether they are cooked for a long time or only briefly. Here, too, the
acoustic frequency spectra provide information about the material properties.

15.3 Ultrasonic Sound

Sound with high frequencies in the range between 20 kHz and 10 GHz is beyond the
range of frequencies that can be detected by the human ear and is called ultrasonic
sound or ultrasound. According to Eq. (15.1) the related wavelengths in air are
between 33 nm and 1.6 cm, in water between 0.1 μm and 7 cm, in solids between
0.4 μm and 20 cm. The shortest ultrasonic waves are thus in the range of
wavelengths of light. Because of these short wavelengths, ultrasonic waves—
much stronger than audible sound waves—show phenomena of wave optics such
as refraction, diffraction, reflection.
The technically comparatively simple generation of ultrasound with the help of
the piezoelectric effect led to a widespread application of ultrasound. Low-intensity
546 15 Acoustic Properties

ultrasound does not cause physical or chemical material changes. Therefore,

low-intensity ultrasound with frequencies above 1 MHz is used in sensors for
material characterization and process control. Since the speed of sound depends on
the molecular structure of a material, i.e. on the mass of the molecules and the
strength of the chemical bonds, information about the material can be obtained from
the measurement of wave propagation. Interaction between ultrasonic waves and the
material through which they propagate leads to absorption or backscattering of the
acoustic waves with a weakening of the intensity or change in the propagation speed
and phase. By comparing the incident with the reflected ultrasonic wave, the material
composition can be deduced. Possibilities of combining ultrasound methods with
other instrumental methods for quality assessment of fruit and vegetables are
discussed in [11]. Imaging ultrasound methods such as those used in medical
diagnostics (ultrasonography) also can be used to examine food [12]. With the
help of acoustic spectroscopy, particle sizes of emulsions and colloidal systems
can be determined. The advantage of acoustic technology over laser diffraction (see
Chap. 12) is that samples that are not optically transparent can also be measured. In
this way, e.g. emulsions can be characterized undiluted [2, 13, 14].
High-intensity ultrasound, on the other hand, with frequencies of 20–100 kHz is
used, for example, for drilling, milling, cutting, and welding, as well as for disper-
sion, emulsification, or degassing liquids [15–19]. Various applications are based on
the high accelerations to which the oscillating molecules or atoms are exposed
(between 103 g and 106 g). Such acceleration values in oscillating liquids can lead
to cavitation. In the suction phase of ultrasonic oscillation, cavities form in the liquid,
which collapse in the pressure phase [9]. When the cavities collapse, high forces act
locally, which can be used for dispersion, emulsification, cell disintegration, mem-
brane cleaning, etc. [17].
The determination of the transit time of an ultrasonic pulse can be used for
distance determination. Level sensors for liquids and bulk solids, sonar devices
and flowmeters are based on this principle. Ultrasonic waves reflected by moving
bodies show a frequency shift due to the Doppler effect. By measuring this fre-
quency shift, the speed of the reflecting body can be determined (Doppler sonogra-
phy). Chapter 16, on-line sensors, describes how ultrasonic flow sensors work.

Further Reading

Roasted almonds: acoustic assessment of crispness [20]

Dairy industry: ultrasonic techniques [16, 21]
Biscuit: acoustic assessment of crispness [1, 22]
Meat quality by ultrasound [23, 24]
Apple: acoustic characterization of crispness and hardness. [25]
Drying, degassing, and defoaming by ultrasound [15]
Shockwave processing of food [26]
Freezing and thawing supported by power ultrasound [27]
Brining of meat: intensification by ultrasound [24, 28]

(continued)
References 547

Roasted almonds: acoustic assessment of crispness [20]

Fish: non-destructive evaluation by ultrasound [29]
Honey: inspection by ultrasound [30]
Soil: acoustic sensing to measure soil moisture [31]
Pharmaceutical dispersions: characterization by ultrasound spectroscopy [2]
Acoustic cavitation: technical applications [4]
Microorganisms: ultrasound-assisted thermal inactivation [3]
Membrane food processing: ultrasound effects [32]
Milk protein isolate powder: acoustic monitoring during [33]
Vacuum packages: detection of air leakage using acoustic measurements [34]
Ultrasonic and acoustic microscopy on food [12]
Mapping of cultured biological cells by acoustic [35]
Ultrasound technologies for food and bioprocessing [36]

Summary
Sound plays an important role in the quality assessment of food. This chapter
explains the basics of acoustic waves and associated metrological terms in a
simple way. Examples illustrate the sensory and acoustic perception of
quality-relevant properties of food. Ultrasound is used for acoustic spectros-
copy as well as in numerous on-line technical sensors in continuous produc-
tion operations. At correspondingly high intensity, ultrasound can support
food processing processes. Numerous application examples at the end of the
chapter show the range of applications of acoustic methods and are intended to
encourage further studies or one’s own scientific work.

References
1. Saeleaw M, Schleining G (2011) A review: crispness in dry foods and quality measurements
based on acoustic–mechanical destructive techniques. J Food Eng 105(3):387. https://doi.org/
10.1016/j.jfoodeng.2011.03.012
2. Bonacucina G, Perinelli DR, Cespi M, Casettari L, Cossi R, Blasi P, Palmieri GF (2016)
Acoustic spectroscopy: a powerful analytical method for the pharmaceutical field? Int J
Pharm 503(1–2):174. https://doi.org/10.1016/j.ijpharm.2016.03.009
3. Evelyn SFVM (2020) Ultrasound assisted thermal inactivation of spores in foods: pathogenic
and spoilage bacteria, molds and yeasts. Trends Food Sci Technol 105:402. https://doi.org/10.
1016/j.tifs.2020.09.020
4. Gevari MT, Abbasiasl T, Niazi S, Ghorbani M, Koşar A (2020) Direct and indirect thermal
applications of hydrodynamic and acoustic cavitation: a review. Appl Therm Eng 171:115065.
https://doi.org/10.1016/j.applthermaleng.2020.115065
5. Fraden J (2016) Handbook of modern sensors physics, designs, and applications. doi:https://
doi.org/10.1007/978-3-319-19303-8
548 15 Acoustic Properties

6. Kohlrausch F (1996) Praktische Physik Bd.3. Teubner, Stuttgart. doi:https://doi.org/10.1002/

piuz.19870180109
7. DIN 45630 Grundlagen der Schallmessung; Physikalische und subjektive Größen von Schall
(1971) Beuth, Berlin. https://doi.org/10.31030/1234608
8. Hering E, Martin R, Stohrer M (2017) Taschenbuch der Mathematik und Physik. Springer
Vieweg doi:https://doi.org/10.1007/978-3-662-53419-9
9. Kohlrausch F (1996) Praktische Physik Bd.1. Teubner, Stuttgart. doi:https://doi.org/10.1007/
978-3-322-87205-0
10. Povey MJW (2007) Acoustic techniques to characterize food microstructure. In: McClements
DJ (ed) Understanding and controlling the microstructure of complex foods. Woodhead Pub-
lishing, pp 311–333. https://doi.org/10.1533/9781845693671.2.311
11. Shah LBM (2016) Instrumental and ultrasonic techniques in quality evaluation of fresh fruit and
vegetables. Dr. phil., University, Leeds
12. Povey MJW, Watson N, Parker NG (2013) Ultrasonic and acoustic microscopy: principles and
applications to food microstructures. In: Groves VJM (ed) Food microstructures: microscopy,
measurement and modelling. Woodhead, Cambridge
13. Holmes MJ, Povey MJW (2017) Ultrasonic particle sizing in emulsions. https://doi.org/10.
1002/9781118964156.ch2
14. Povey MJW (2013) Ultrasound particle sizing: a review. Particuology 11(2):135. https://doi.
org/10.1016/j.partic.2012.05.010
15. Villamiel M, Riera E, García-Pérez JV (2021) The use of ultrasound for drying, degassing and
defoaming of foods. In: Knoerzer K, Muthukumarappan K (eds) Innovative food processing
technologies. Elsevier, Oxford, pp 415–438. https://doi.org/10.1016/B978-0-08-100596-5.
22957-0
16. Silva M, Chandrapala J (2021) Ultrasound in dairy processing. In: Knoerzer K,
Muthukumarappan K (eds) Innovative food processing technologies. Elsevier, Oxford, pp
439–464. https://doi.org/10.1016/B978-0-08-100596-5.22955-7
17. Bhargava N, Mor RS, Kumar K, Sharanagat VS (2021) Advances in application of ultrasound
in food processing: a review. Ultrason Sonochem 70:105293. https://doi.org/10.1016/j.
ultsonch.2020.105293
18. Chen F, Zhang M, Yang C-h (2020) Application of ultrasound technology in processing of
ready-to-eat fresh food: a review. Ultrason Sonochem 63:104953. https://doi.org/10.1016/j.
ultsonch.2019.104953
19. Musielak G, Mierzwa D, Kroehnke J (2016) Food drying enhancement by ultrasound – a
review. Trends Food Sci Technol 56:126. https://doi.org/10.1016/j.tifs.2016.08.003
20. Varela P, Chen J, Fiszman S, Povey MJW (2006) Crispness assessment of roasted almonds by
an integrated approach to texture description: texture, acoustics, sensory and structure. J
Chemom 20(6–7):311. https://doi.org/10.1002/cem.1029
21. Mohammadi V, Ghasemi-Varnamkhasti M, Ebrahimi R, Abbasvali M (2014) Ultrasonic
techniques for the milk production industry. Measurement 58:93. https://doi.org/10.1016/j.
measurement.2014.08.022
22. Chen J, Karlsson C, Povey M (2005) Acoustic envelope detector for crispness assessment of
biscuits. J Texture Stud 36(2):139. https://doi.org/10.1111/j.1745-4603.2005.00008.x
23. Alarcon-Rojo AD, Carrillo-Lopez LM, Reyes-Villagrana R, Huerta-Jimenez M, Garcia-Galicia
IA (2019) Ultrasound and meat quality: a review. Ultrason Sonochem 55:369. https://doi.org/
10.1016/j.ultsonch.2018.09.016
24. Sergeev A, Shilkina N, Tarasov V, Mettu S, Krasulya O, Bogush V, Yushina E (2020) The
effect of ultrasound treatment on the interaction of brine with pork meat proteins. Ultrason
Sonochem 61:104831. https://doi.org/10.1016/j.ultsonch.2019.104831
25. Demattè ML, Pojer N, Endrizzi I, Corollaro ML, Betta E, Aprea E, Charles M, Biasioli F,
Zampini M, Gasperi F (2014) Effects of the sound of the bite on apple perceived crispness and
hardness. Food Qual Prefer 38:58. https://doi.org/10.1016/j.foodqual.2014.05.009
References 549

26. Aganovic K, Bolumar T, Toepfl S, Heinz V (2021) Fundamentals of shockwave processing for
food. In: Knoerzer K, Muthukumarappan K (eds) Innovative food processing technologies.
Elsevier, Oxford, pp 395–411. https://doi.org/10.1016/B978-0-12-815781-7.00023-8
27. Qiu L, Zhang M, Chitrakar B, Bhandari B (2020) Application of power ultrasound in freezing
and thawing processes: effect on process efficiency and product quality. Ultrason Sonochem 68:
105230. https://doi.org/10.1016/j.ultsonch.2020.105230
28. Gómez-Salazar JA, Galván-Navarro A, Lorenzo JM, Sosa-Morales ME (2021) Ultrasound
effect on salt reduction in meat products: a review. Curr Opin Food Sci 38:71. https://doi.org/
10.1016/j.cofs.2020.10.030
29. Tokunaga K, Saeki C, Taniguchi S, Nakano S, Ohta H, Nakamura M (2020) Nondestructive
evaluation of fish meat using ultrasound signals and machine learning methods. Aquac Eng 89:
102052. https://doi.org/10.1016/j.aquaeng.2020.102052
30. González-Mohino A, Jiménez A, Paniagua MJ, Perez-Palacios T, Rufo M (2019) New
contributions of ultrasound inspection to the characterization of different varieties of honey.
Ultrasonics 96:83. https://doi.org/10.1016/j.ultras.2019.02.010
31. Gorthi S, Chakraborty S, Li B, Weindorf DC (2020) A field-portable acoustic sensing device to
measure soil moisture. Comput Electron Agric 174:105517. https://doi.org/10.1016/j.compag.
2020.105517
32. Córdova A, Astudillo-Castro C, Ruby-Figueroa R, Valencia P, Soto C (2020) Recent advances
and perspectives of ultrasound assisted membrane food processing. Food Res Int 133:109163.
https://doi.org/10.1016/j.foodres.2020.109163
33. Wu S, Fitzpatrick J, Cronin K, Ahmed MR, Fitzpatrick D, Miao S (2019) Application of
broadband acoustic resonance dissolution spectroscopy (BARDS) to the gas release behaviour
during rehydration of milk protein isolate agglomerates. J Food Eng 253:14. https://doi.org/10.
1016/j.jfoodeng.2019.02.010
34. Koruk H, Sanliturk KY (2019) Detection of air leakage into vacuum packages using acoustic
measurements and estimation of defect size. Mech Syst Signal Process 114:528. https://doi.org/
10.1016/j.ymssp.2018.05.023
35. Hozumi N, Yoshida S, Kobayashi K (2019) Three-dimensional acoustic impedance mapping of
cultured biological cells. Ultrasonics 99:105966. https://doi.org/10.1016/j.ultras.2019.105966
36. Feng H, Barbosa-Cánovas GV, Weiss J, Link S (2011) Ultrasound technologies for food and
bioprocessing. Springer, New York
On-line Sensors
16

Many quality-related properties like color, viscosity, or moisture content in most

cases are measured off-line by sending samples from a production line to a labora-
tory. Results from laboratory measurements often take several hours to be reported.
This delay leads to costly interruption of continuous production. If products not
meeting quality specifications leave the plant during that lag time, consumer
complaints over inferior products could have a damaging effect in the marketplace
for those products. In-line sensors which take measurements and produce results in
real time can be of great value in improving production efficiency. When an in-line
sensor at a critical control point detects a quality deviation, this detection can be
immediately transferred to the production line control center so that products not
meeting quality specifications can be ejected or removed from the production line
without interruption. In this way, continuous production of products with consistent
quality can be achieved.
In most cases, in-line sensors are designed to measure physical properties. This is
because physical properties can be more easily detected and measured more quickly
than chemical or biological properties. While food ingredients and microorganisms
are normally measured in laboratories, physical properties like optical refraction,
resonance frequency, impedance, NIR absorption can be measured quickly in real
time. The challenge is to develop on-line sensors which can translate detected
defects in physical properties to related chemical or microbiological properties.
Those types of sensors are often called chemo-sensors and biosensors [1–4]. In
addition to high sensitivity, such sensors must be robust enough to withstand varying
conditions of processing and machine cleaning operations on the factory floor.
Requirements for sensors like these are difficult to achieve and can only be met by
interdisciplinary teamwork of experienced technologists, engineers, and scientists.
Reliable analytics play a central role in reaching food safety and quality assurance.
Creative ideas are often needed for fusion of technical solutions that will integrate
analytics into the production process. This type of technology falls into the realm of
biomedical engineers along with teams of chemical, electrical, and mechanical
engineers that work in industries that manufacture such products as medical

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 551
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8_16
552 16 On-line Sensors

diagnostic equipment, hospital products, and pharmaceutical and cosmetic products,

as well as food products [5–7].
This chapter begins with the study of terms from control and measurement
disciplines. Then, the chapter covers principles that lie behind some on-line sensors
and their future options. With the understanding of those basic principles, the chapter
moves on to chemo -and biosensors. At the end of the chapter there is a list of sensor
applications.

16.1 Terms

A sensor is generally said to be a device that turns a stimulus into a display

[8]. However, this definition is too broad for our concern. When developing
in-line sensors, we need sensors that produce an electrical signal as an output that
can be forwarded, processed, memorized, and used for production control. Since we
only want to talk about fast (in-line) sensors in this chapter, we will focus on sensors
that measure physical quantities. One definition that meets this is:

Definition
A sensor is a device that converts a physical effect into an electrical signal. An
in-line sensor will transmit this signal in real time.

In Table 16.1 examples of some physical properties and physical effects that can
be easily measured by in-line sensors are given.

Table 16.1 Physical properties and effects capable of being measured by in-line sensors [8, 9]
Force Velocity Concentration
Torque Acceleration Water content
Mechanic stress Angular velocity Viscosity
Momentum of inertia Angular acceleration Turbidity
Distance Rotational frequency Foreign matter
Angle Flow velocity Roughness, gloss
Length, area, number Volume flow Color
Position Mass flow
Volume Light
Density Infrared
Object recognition Ionizing radiation
Filling height Ultrasound
Strain, deformation Vibration, acoustic signals
Weight, mass Resonance frequency, damping
Temperature Heat, heat flow
Pressure
pH
16.1 Terms 553

Table 16.2 Classification of sensors

Sensor Description
On-line Get data in real time
In-line Sensor is integrated with the production line
At-line/near-line Sensor is operated close to production line
Off-line Sensor is operated far from production line, e.g. in an on-site
laboratory or off-site laboratory

Table 16.3 Applications of sensors for process- and quality monitoring

Sensor for Process property Product property
Goal Process monitoring, control, Quality and safety of product
continuous uninterrupted
production
Examples Filling height of a tank, oven Getting concentration data by refraction
temperature, monitoring measurement, color of roasted product,
machine vibration X-ray absorption for detection of foreign
matter.

Table 16.2 shows a different classification of sensors according to their location.

When defining the term “in-line sensor,” we take note that the adjective “in-line”
refers to the real-time transmission of the electrical measurement signal. Here “in-
line” stands for the electrical communication connection. However, if “line” means
the production line, then sensors can be classified according to whether they are
integrated into the production line (in-line), operated in proximity (at-line), or
completely separate from it in a laboratory (off-line). This differentiation of sensors
based on their spatial arrangement corresponds directly with the time delay between
measurement and feedback of the measurement result. When using sensors in a
laboratory, sampling and sample transport are necessary, which creates a time
interval between sampling and feedback of the measurement result. With a measur-
ing stand near the production line (at-line), the test transport is omitted, so the time
interval is smaller. An in-line sensor, on the other hand, records measurement data in
the production line and sends it in real time to an electrical network. An in-line
sensor thus meets the requirement for real-time data.
There are also other possibilities for categorizing sensors, such as how active or
passive they are. Active sensors have their own power supply, while passive sensors
take the energy to generate their signal from the material to be measured. Examples
of passive sensors are impeller counters and thermocouples. Sensors are also
categorized according to whether their output signal is analog or digital, as well as
those that measure process variables or product sizes. Table 16.3 gives some
examples.
Since we want to focus on on-line sensors in this chapter, i.e. sensors that can be
used for process control and process automation, Table 16.4 lists some in-line
sensors with their associated measurement principles.
554 16 On-line Sensors

Table 16.4 On-line sensors in food production, examples

Measured property Measuring principle
Mass Elastic deformation, piezoresistive force measurement, weight force
Pressure Elastic deformation, capacitive or resistive measurement
Density Measurement of the resonance frequency of an oscillator
Mass flow Magnetic induction, Coriolis effect, thermal measurement
Volume flow Traveling time of an ultrasonic pulse, Doppler frequency
Viscosity Attenuation of an oscillator, pressure drop, torque on a rotating body
Position Magnetic inductive, light reflection, triangulation, ultrasound
Thickness Absorption of electromagnetic waves
Particle size and shape Laser diffraction, ultrasound spectroscopy, in-situ microscopy,
camera
Surface tension Capillary pressure measurement (bubble point)
Temperature IR radiation, thermal expansion, electrical resistance
Thermal conductivity Temperature difference and heat flow
Level, filling height Pressure, ultrasound echo, radar echo, electrical resistance, electrical
capacitance, attenuation
Rotational frequency Count light signal or inductive signal
Concentration, acidity Electrical resistance, electrical impedance, pH measurement
Sugar concentration, Refraction measurement
salinity
Concentration of Turbidity measurement
disperse phase
Water content Permittivity measurement
Water, fat, protein, NIR absorption, FTIR spectrometry, NMR
composition
Fat in meat Impedance measurement, X-ray absorption
Chemical or biological Change in electrical, optical or mechanical properties of a sensor with
substances a bio-sensitive or chemo-sensitive receptor
Metallic foreign bodies Change in electromagnetic induction
Non-metallic foreign X-ray absorption, NMR, thermography
bodies
Color Optical absorption
Texture Force, tension, strain, acoustic properties
Water activity Humidity measurement
Humidity of air Electric resistive, capacitive, change in length, dew point
measurement, psychrometry
Dew point Optical or electrical properties of a cooled surface
Psychrometrics Temperature difference between dry and wet temperature sensor

Table 16.4 shows the variety of on-line sensors. However, if we want to develop a
process control or automation system for a specific process, the question is reversed.
Then it is necessary to clarify:
16.2 Control–Directing–Monitoring 555

1. What characteristics must the product have that could provide quality-relevant
information about the product?
2. Which in-line sensors are available for these features?

The meaning of the first question can be clarified by looking at Table 16.5 which
lists some examples of how measurable physical properties can be related to product
quality.

16.2 Control–Directing–Monitoring

A sensor is used for the purpose of control, directing, or process monitoring. To

differentiate these terms from each other and understand their meaning, we will look
at a simple food process (“baking bread rolls”) and go through the terms one by one.
Achieving continuous quality assurance requires the ability to control the process by
knowing which process parameters must be measured and monitored and compared
with process specifications for those parameters.

Example
A conveyor belt transports dough pieces with a defined mass and composition
through a baking line. In order for the baked dough to reach the specified
quality specification, it must be in an oven at a temperature of 180 °C for
3 min. The product property is the volume of the baked dough. Process
parameters are temperature and time.

Monitoring a process means continuously reading or determining the process

parameters and product properties. You can then define a value range for process
parameters and product properties in which the process should run. If the values run
outside this range, an alarm can be issued.

Example
Oven temperature and residence time in the oven are monitored. The residence
time in the oven results from the speed of the conveyor belt and the length of
the oven. The temperature (of all zones) of the oven and the speed of the
conveyor belt or the rotational frequency of its drive motor are continuously
read and logged. Excursions from defined limit values can thus be detected.

Determining product properties while processing is underway is called process

analytics. While samples of products are often taken from the production line to be
examined in the laboratory, some product properties can also be determined during
the ongoing process. This requires fast sensors.
556 16 On-line Sensors

Table 16.5 Physical properties and related quality properties

Physical properties Connection to quality (examples)
Mechanical properties
Length Particle size in powders, emulsions, foams. Texture parameters of
chocolate, ice cream, pulp
Volume Porosity, bulk density of powders
Density Concentration of sugars in fruit juices, ethanol in beverages, water in
milk, starch in potatoes
Hardness, elasticity Texture of food
Viscosity, yield point Flow behavior, consistency, stability
Mass, weight Filling weight
Electrical properties
Conductivity Water content in cereals, honey, milk. Electrolyte concentration in milk,
particle counting
Impedance Microbial cell growth, fat content meat
Induction Metallic foreign bodies in food
Permittivity Water content
Acoustic properties
Acoustic impedance Moisture, ripeness of apples
Sound spectrum Texture of baked goods and cereals
Ultrasonic Material composition, drop size of emulsions
impedance
Optical properties
Refraction Content of water, sugar, dry matter in beverages
Color Condition and ripeness of fruits
Polarization angle Sugar concentration
NIR-Absorption Composition of food
Electromagnetic properties
Microwave Water content
absorption
X-ray absorption Metallic and non-metallic foreign bodies in food
Thermal properties
Temperature Transport conditions, process conditions (sterilization, pasteurization)
Solidification Water content of milk
temperature
Glass transition Stability
temperature
Melting temperature Identity and purity
Magnetic properties
Induction Metallic foreign bodies
LR-NMR Solid fat content (SFC), water content, fat content
HR-NMR Isotope ratio, authenticity, origin of food
16.2 Control–Directing–Monitoring 557

Example
As a supplement to the off-line laboratory measurement of the product prop-
erty “volume of the baked dough pieces,” that same property can be randomly
determined at-line by measuring displacement of the baked dough pieces at a
testing station near to the production line.

If at-line process analysis is too slow, products with undesirable quality

deviations can be manufactured before countermeasures are taken. Economic losses
are the result. If you want to prevent this, you need faster process analysis. A
measurement in real time would be optimal, i.e. you need an on-line sensor for the
product property.

Example
One finds that, e.g., by using a CCD camera, which takes lateral pictures of the
baked dough pieces, one can also deduce the volume of the baked dough. A
camera could therefore take over the function of the on-line sensor.

If it turns out that there is an alternative product property for quality characteriza-
tion that may be easier to measure on-line or more cost-effective, consider switching
process analytics to it. Before doing this, it is safe to determine in advance in
laboratory tests that this alternative product property is suitable and that the intended
on-line sensor has a suitable sensitivity, resolution, and stability. This is tested by
creating a new quality specification that matches the alternative product characteris-
tic, testing samples and validating the measurements with a reference method.

Example
You want to make an on-line measurement of the color of the baked dough
pieces instead of the at-line volume determination. You need a correlation
between the color of the baked dough pieces and their volume in advance. This
shows whether color measurement is suitable for replacing previous analytics.
With the new quality specification, which specifies color measures and their
tolerance range, sample measurements are carried out to test the on-line
colorimeter.

Process control is understood to change process parameters with the help of

measured deviations between product property and quality specification in such a
way that these deviations are reduced. A typical element of a process control is thus
the feedback of this actual-setpoint-comparison between the measured variable
(actual value) and the target variable (setpoint) in the running process. The difference
between the measured variable and the target variable results in the necessary
558 16 On-line Sensors

controlling intervention in the process, i.e. the sign and the size of the change of the
respective process parameter.

Example
Color measurement shows that the dough pieces are too dark. Comparison of
this color measurement with the target value or set point shows that the baking
time must be reduced by 10 s. As a result, the speed of the conveyor belt is
increased to such an extent that the baking time is reduced by 10 s.

In control technology, a distinction is made between measuring, directing, and

controlling. The difference between directing and controlling lies in the feedback
(closed loop): In the case of controlling, the result of the actual-setpoint comparison
is used to intervene in the process in a controlling manner. Without feedback—e.g.,
if there is no measurement tool installed—we do not speak of closed loop control
with feedback, but of open loop control (directing feedforward).

Example
If you increase the flow through a pipeline by opening a valve a little further,
you are directing the flow. You are not controlling it. A prerequisite for control
is a measurement and comparison with an actual-setpoint, from which the
controlling intervention is derived.

Let’s take a closer look at the terms control and directing from the point of view
of time: In principle, control can take place very quickly. This requires a fast
measurement, a fast actual-setpoint comparison, fast data transmission, and a
delay-free directing action. In practice, however, it turns out that each of these
sub-steps requires a certain amount of time, i.e., even an on-line sensor has a certain
inertia (time constant). For this reason, a control is always carried out with a small
time lag, the minimization of that time lag is a constant challenge. On the other hand,
it may also be desirable to have a sensor that does not pass on short-term fluctuations
of the measured value (noise) so that the control unit is working uselessly. For this
purpose, the measured values of the sensors can be averaged over a certain period of
time and only these averaged values need to be used for the control. This is referred
to as integrating sensors. If you opt for such a procedure, the resolution of the sensor
should not be unnecessarily high.

Example
The hot air temperature of an oven statistically fluctuates between 178 °C and
182 °C. A fast temperature sensor with a resolution of 0.1 °C displays this
highly noisy temperature signal. However, it does not make sense to regulate

(continued)
16.3 Working Principles of On-line Sensors 559

the oven heater in the same rhythm, e.g., in microsecond cycles. Here, a
temperature sensor with lower resolution, e.g., 0.5 °C and a greater inertia
(e.g., integration time 1 s) would be sufficient. It would provide a “quieter”
temperature signal and avoids unnecessarily frequent response of the oven
heater.

On the other hand, too long a period between the acquisition of measured values
and the controlling intervention of the control system is disadvantageous. This can
lead to the production of inferior quality until the intervention of the regulation has
become effective. Depending on the technique used, milliseconds, minutes, hours or
longer can pass between the measurement and the control action. In the case of
off-line analytics, with which we started this consideration, hours could pass before
production is interfered with to make necessary changes.

Example
In a baking line without on-line process analysis, product samples are ran-
domly taken and transported to the operating laboratory for quality assess-
ment. On Monday morning, the laboratory reports that the samples from
Friday afternoon are outside the quality specification.

16.3 Working Principles of On-line Sensors

In this section we will take a closer look at how some on-line sensors work. By
understanding the function, it is possible to recognize possibilities and limits of
automation ideas for your own production.

16.3.1 Piezoresistive Sensors

If an electrical conductor is stretched in length, its electrical resistance changes:

Since the strain via Hooke’s law is related to the force, this piezoresistive effect can
be used for force measurement. The piezoresistive effect occurs not only in metals
but also in semiconductors. By installing piezoresistive elements (also called strain
gauges in machine elements), the mechanical load on the element can be easily
tracked by measuring the electrical resistance. After appropriate calibration, stresses
and forces can be measured in this way or analogous strains and compressions.
Common applications include piezoresistive pressure measurement and
piezoresistive weighing. By installing a piezoresistive sensor in a weighing device,
weight forces or masses can be determined (see Chap. 2). Such electronic balances
are used in numerous designs, for example to determine the weight of loaded trucks
560 16 On-line Sensors

when receiving raw materials, the weight of transport units (pallets, containers),
stationary storage tanks, and products on a conveyor belt. In-line belt balances for
checking the filling weight of prepackages are also known as checkweighers. Multihead
weighers in production contain a variety of electronic scales that are used to weigh in
parallel. They are an example of in-line sensors that can be used to automate production.
Areas of application of piezoresistive elements include production, storage, tank
farms, transport equipment, cranes, and material testing.
By measuring inertial forces, piezoresistive sensors can also be used to measure
accelerations and angular accelerations. These are called piezoresistive
accelerometers. With the help of miniaturized measuring systems (Micro-Electro-
Mechanical Systems, MEMS), piezoresistive sensors based on semiconductors can
be built very small and sensitive [8], e.g. in nanoscale design [10].

16.3.2 Mechanical Oscillation

16.3.2.1 Density Measurement

There are in-line sensors that can be used to determine the density of fluids as they
flow through a pipe. A U-shaped tube is set into bending vibrations. The resonance
frequency of the tube depends on its mass, i.e., on the density of the fluid. This
arrangement is shown schematically in Fig. 16.1. The resonance frequency of the
bending oscillator is measured, which is related to the density of the material in the
tube with the help of a previous calibration of the system.
The oscillation period of the bending oscillator is

m
T = 2π ð16:1Þ
D

with

mpipe þ msample
T = 2π ð16:2Þ
D
i.e.,

Fig. 16.1 Construction of a

U-tube bending oscillator,
schematic. An
electromagnetic system
(E) causes the U-shaped tube
(U) to vibrate while the liquid
sample is flowing through it
E
U
16.3 Working Principles of On-line Sensors 561

Fig. 16.2 Bending oscillator:

relationship between
oscillation period and density

U
A

B 0
2
T

mpipe þ ρsample V pipe

T 2 = 4π 2 ð16:3Þ
D
so

T 2D mpipe D 4π 2 mpipe
ρsample = 2
- = 2 T 2- ð16:4Þ
4π V pipe V pipe 4π V pipe D

With the following abbreviations:

D 4π 2 mpipe
A= and B=
4π 2 V pipe D

the relationship between vibration period and density is (see Fig. 16.2)

ρ = A  T 2- B

D elastic constant in N m
mpipe mass of empty u-shaped pipe in kg
msample mass of sample in u-shaped pipe in kg
T oscillation period in s
V volume in m3

Constants A and B are determined by calibrating the instrument with samples of

known density. As from Fig. 16.2 A is the slope in the ρ-T2 -diagram and B the axis
section on the T2 axis. To create a two-point calibration, the vibration periods of at
562 16 On-line Sensors

least two samples with densities ρ1 and ρ2 must be determined. These calibration
standards should be chosen in such a way that the subsequent measured values are
within the calibrated range. As an example, if densities around 1200 kg m-3 are to be
measured, it is not recommended to use ethanol (ρ = 780 kg m-3) and water
(ρ = 998.2 kg m-3) as calibration standards. Is this case sucrose solutions with
densities of, e.g., 1100 kg m-3 and 1300 kg m-3 can be used as calibration
standards.
A major advantage of density determination by mechanical oscillation (bending
oscillator technology, vibration technology) is the possibility of obtaining measured
values while the fluid flows through the oscillator. In this way, the oscillator can be
used as an in-line sensor for process automation or as a fast at-line measuring device.

Example
Calibration of a bending oscillator to determine the density of beer. As
calibration standards, water and 20% (m/m) sucrose solution are used:

Calibration standard Density (from literature) Reading

H2O (20 °C) ρ1 = 998.2 kg/m-3 T1 = 4.0044 units
Sucrose solution ρ2 = 1080.96 kg/m-3 T2 = 4.0594 units
20% (m/m) 20 °C

for A and B we have

dρ Δρ ρ - ρ1
A= ≈ = 22
dT 2
ΔT 2
T 2 - T1
2

ρ1 ρ
B = T 21 - = T 22 - 2
A A
i.e.,

ρ2 - ρ 1 ð1080:96 - 998:2Þ kg m - 3
A= = = 186:598 kg m - 3 units - 2
T2 - T1
2 2
4:05942 - 4:00442 units2

ρ2 1080:96 kg m - 3 units2
B = T 22 - = ð4:0594 unitsÞ2 - = 10:68575 units2
A 186:598 kg m - 3

Example measurement of a beer sample: Averaged vibration period is

T20° = 4.0594 units

so with ρ = A T 2- B

we get

-3
= 998:35 kg m - 3
0
ρ20
beer = 186:598 × 4:0045 - 10:685754 kg m
2
16.3 Working Principles of On-line Sensors 563

From the principle of measuring the resonance frequency of a mechanical oscil-

lator, there are miniaturized technical solutions, such as oscillating quartz crystals on
a semiconductor chip for electronic measurement of frequency and attenuation.
Components that combine micromechanical structures and electronic logic elements
in one chip are called micro-electro-mechanical systems (MEMS) [11].
They are based, for example, on a resonator with a piezoresistive electrical
measuring bridge and can provide the density and viscosity of the sample by
measuring frequency and admittance. Such a “microbalance” (quartz crystal micro-
balance, QCM) requires only small sample volumes, e.g. 10–100 μl [12, 13].

16.3.2.2 Flow Measurement (Coriolis Effect)

Coriolis sensors also use the mechanical oscillation of a pipe. The Coriolis force
deforms the shape of vibration with increasing flow velocity of a fluid as it flows
through the vibrating tube. In-line flow sensors can be built on basis of this principle.
→
The Coriolis force on a moving mass m in a system with angular velocity ω is
→
proportional to the translation velocity v of the mass:

→ → →
FC = -2 m× v × ω ð16:5Þ

ω = 2π  f ð16:6Þ

FC Coriolis force in N
m mass of the moving material in kg
v translation speed of mass in m s-1
ω angular velocity in s-1
f frequency in s-1

If a fluid is allowed to flow through a pipe that vibrates with the frequency
f perpendicular to the flow direction, then the Coriolis force leads to a deformation
of the oscillating system (see Fig. 16.3) which is electronically measurable between

Fig. 16.3 The Coriolis force

FC leads to the deformation of
an oscillating tube (with a
fluid flowing through it, which
can be measured as a phase
shift Δφ between points A and
B. ω is the angular frequency
of the oscillation
564 16 On-line Sensors

Table 16.6 Mechanical oscillation for in-line characterization of a flowing fluid

Measured quantity Influencing factor To determine ...
Frequency Mass Density
Coriolis deformation Velocity of mass Mass flow
Damping Non-valent interactions Viscosity

two stationary sensors on the pipe [8]. The measurement signal of the Coriolis flow
sensor is proportional to the flow velocity and mass of the flowing fluid is related to
the mass flow of the fluid. This is an advantage over many other flow sensors, which
provide the volumetric flow by measuring the flow velocity, which can then only be
converted into the mass flow rate with the help of the density of the fluid.
Another advantage of the Coriolis flow sensor is that no pipe fittings and seals are
required. As a result, an in-line flow sensor based on the Coriolis principle does not
bring any hygienic disadvantages to the process. In addition to the phase shift, the
attenuation of the oscillation can be determined in parallel, which allows conclusions
to be drawn about the viscosity of the fluid. A concurrent measurement of the
temperature of the fluid is then advisable. Table 16.6 summarizes the measured
variables that can be recorded simultaneously in-line.

16.3.3 Induction (Flow Measurement, Metal Detection)

16.3.3.1 Magnetic-Inductive Flow Sensors

In the case of electrically conductive fluids, on-line flow measurement is also
possible due to electromagnetic detection. In a magnetic field, the Lorentz force
acts on moving electric charge carriers, ref. Chap. 10.

→ →
jF L j = Q  v  B  sin∢ v ; B

The mode of operation of a magnetic-inductive flow sensor (magmeter) is

schematically shown in Fig. 16.4. A fluid flows through a pipe at the velocity v.
Perpendicular to the direction of flow is a magnetic field B. The Lorentz force leads
to the deflection of positive charge carriers to the right (seen in the direction of flow)
and negative charge carriers to the left. This creates an electrical voltage UH
perpendicular to the flow direction. If the magnetic flux B and the composition of
the fluid are constant, the voltage UH is proportional to the flow velocity v.
The volume flow therefore is

V_ = v  A

For a non-compressible fluid, the mass flow results from the flow velocity v with
16.3 Working Principles of On-line Sensors 565

Fig. 16.4 Magnetic-

inductive flow sensor.
Perpendicular to the flow
velocity v of the charged
particles and perpendicular to
the direction of magnetic field
B, the electrical voltage UH is
generated

m_ = ρ  v  A

V_ volume flow in m3s-1

m_ mass flow in kgs-1
ρ density in kgm-3
v flow velocity in ms-1
A area in m2
FL Lorentz force in N
Q electric charge in C
B magnetic flux in Vsm-2

Magnetic-inductive flow sensors are only suitable for electrically conductive

fluids. The electrical conductivity of drinking water is already sufficient to generate
a reliable measuring signal for sensors with high sensitivity and is therefore suitable
for most liquid foods.
In practice, the magnetic-inductive flow sensor is operated in such a way that UH
is not generated as a DC voltage but as an alternating voltage. An alternating voltage
for UH has the advantage that there are no chemical changes to the electrodes. This is
achieved by an alternating magnetic field with a frequency range of 0.1 to 1 kHz [8].

16.3.3.2 Metal Detection

Metal detectors also work according to the principle of electromagnetic induction.
When an electrically conductive body enters an alternating electromagnetic field, an
electrical voltage is induced into this body, which leads to the formation of eddy
currents, ref. Chap. 10. This effect can be measured as a disturbance of the primary
alternating electromagnetic field. Scanning metal detectors have the shape of a coil
(transmitter and receiver), with which the objects to be examined are scanned.
Tunnel-shaped metal detectors allow investigation during the tunnel passage with
the help of a conveyor belt, see Fig. 16.5. When the metal detector has a positive
signal, the food in question is ejected from the production line in real time and taken
for laboratory examination to determine the reason for contamination with metallic
particles.
566 16 On-line Sensors

Fig. 16.5 Tunnel-shaped

metal detector for inspecting
products on a conveyor belt,
schematic

Fig. 16.6 Ultrasonic flow E S

sensor, schematic. The transit
time of an ultrasonic pulse
from sensor S to receiver E is
shorter in the flow direction
than against the flow direction

S E

16.3.4 Ultrasound Transit Time

Flow Sensor
By measuring the transit time of ultrasound waves (see Chap. 15), distances can be
measured, and positions and speeds can be determined. Ultrasonic flow meters make
use of the difference in transit times of ultrasonic pulses, which are emitted in and
against the direction of fluid flow, see Fig. 16.6. From the transit time difference,
flow velocity can be calculated.
16.3 Working Principles of On-line Sensors 567

Ultrasonic transmitters and receivers are piezoelectric elements that vibrate at a

frequency of a few MHz. A piezoelectric element can alternately act as transmitter
and receiver, so in practice only two elements are needed and not four as in the
schematic (Fig. 16.6).
An alternative method is to measure the phase difference between the ultrasonic
pulses emitted in and against the direction of the flow. With continuous ultrasonic
waves instead of the pulse method, the Doppler effect can be used to measure flow
velocity, see Chap. 14.

Position, Distance, Level

By measuring the transit time of ultrasonic pulses, distances can be measured and
thus positions of objects or filling levels can be determined.

16.3.5 Capacity and Permittivity

Electrical capacitance can be measured just as quickly as electrical resistance, so

capacitive sensors are just as widely used as electro-resistive sensors. Since the
electrical capacitance of a component depends on the material that is in the electric
field (ref. Chap. 9), material properties can also be tracked via capacitance
measurements. To understand the principle, let’s start this section by calculating
the capacitance of a simple capacitor for measuring the level of liquid or fluid in a
tank or vessel.

16.3.5.1 Capacitive Measurement of Fluid Level

The capacitance of a square plate capacitor as in Fig. 16.7 is:

a2
C = ε  ε0 
d

Fig. 16.7 Partially filled

plate capacitor, schematic

H h
a

d
568 16 On-line Sensors

Fig. 16.8 Parallel circuit of Cempty part

two capacitors

Cfilled part

The capacitance of a plate capacitor partially filled with product (Fig. 16.7) can be
calculated by considering the filled part and the empty part of the capacitor as two
capacitors connected in parallel (Fig. 16.8).

C capacity in F
ε permittivity number
ε0 electric field constant
a plate length in m
h filling level in m
d plate distance in m
χ electrical susceptibility

The total capacity results from

C total = Cfull þ C empty ð16:7Þ

the individual capacities are

ah
C full = εfull  ε0  ð16:8Þ
d
ð a - hÞ  a
C empty = εempty  ε0  ð16:9Þ
d
The total capacity is therefore

ε0  a
Ctotal =  ðεfull  h þ ða - hÞÞ ð16:10Þ
d
transformed

ε0  a
Ctotal =  ðh  ðεfull - 1Þ þ aÞ ð16:11Þ
d
i.e.,
16.3 Working Principles of On-line Sensors 569

Fig. 16.9 Partially filled

cylinder capacitor

ri
ra

ε0  a
Ctotal =  ða þ h  χ Þ ð16:12Þ
d
respective

ε0  a  χ
Ctotal = C empty þ h ð16:13Þ
d
The capacitance of the measuring arrangement increases linearly with the filling
level h in the capacitor. Therefore, such an arrangement in principle is suitable as a
level sensor.
Sometimes cylindrical capacitors are advantageous, then analogously:
For the capacitance of a cylinder capacitor as in Fig. 16.9

2π  ε0  ε  l
C= ð16:14Þ
ln rrai

The total capacity results from

C total = Cfull þ C empty ð16:15Þ

the individual capacities are

2π  ε0
C full = ε  h: ð16:16Þ
ln rrai
570 16 On-line Sensors

2π  ε0
C empty = ðl- hÞ ð16:17Þ
ln rrai

i.e.,

2π  ε0 2π  ε0
Ctotal = εhþ ðl- hÞ ð16:18Þ
ln rrai ln rrai

2π  ε0
C total =  ðε  h þ ðl- hÞÞ ð16:19Þ
ln rrai

2π  ε0 2π  ε0
C total =  ðhðε - 1Þ þ lÞ =  ðh  χ þ l Þ ð16:20Þ
ln rrai ln rrai

2π  ε0  l 2π  ε0  χ
C total = þ h ð16:21Þ
ln rrai ln rrai

2π  ε0  χ
C total = C empty þ h ð16:22Þ
ln rrai

Here, too, there is a linear relationship between filling level h and capacitance
Ctotal of the capacitor.

16.3.5.2 Measurement of Humidity

If the capacitor is filled with a material whose electrical permittivity strongly
depends on the water vapor equilibrium of the ambient atmosphere, the humidity
can be measured capacitively. Such a material may be a film of a hydrophilic or
hygroscopic polymer with metallized outer surfaces. Miniaturized sensors can also
be used to carry out moisture measurements in processes [14], such as during
fluidized bed drying [15].

16.3.5.3 Cell Growth

The material in the capacitor can also be a liquid in which microorganisms multiply.
Since a growing number of biological cells change the electrical capacitance of the
suspension, cell growth can be monitored capacitively [16]. The measurement is
carried out with alternating voltage. So, strictly speaking, the measured quantity is
not the capacitance but the impedance of the solution, ref. Chap. 9. Impedance
measurement can also be based on the functional principle of biosensors. Selective
adsorption of microorganisms on a functionalized surface changes their impedance
[17]. In this way, detection of microorganisms by measuring impedance is possible,
refer Sect. 16.4.
16.3 Working Principles of On-line Sensors 571

16.3.6 Refraction and Absorption

In the chapter on optical properties (Chap. 12) we learned that different propagation
speeds of electromagnetic waves lead to the refraction of this radiation. Since the
propagation speed of electromagnetic waves in matter depends on the composition
of the materials, information about the material can be obtained by measuring the
refractive index. This is often used to measure the concentration of dissolved solute
in an aqueous solution, such as when evaporating fruit juices or other beverages.
Refractive on-line sensors measure the refractive index of a solution as it flows
through a measurement channel. At-line sensors exist as thermostatic desktop
devices or as battery powered handheld devices. In the field of optics, numerous
values for the refractive index of sugar solutions at a wavelength of 589 nm are
tabulated (see Annex) or available by numerical value equations. Therefore, elec-
tronic refractometers are often operated at this wavelength. The measured quantity is
the change in direction of an electromagnetic wave during transition from the optical
glass of the sensor to the solution to be measured, ref. Chap. 12.
With the help of optical waveguides, distances between location of the in-line
measurement and the sensor electronics can be bridged. There are numerous optical
biosensors based on refractometric sensors [18, 19]. An example of an optical
refraction sensor is also the SPR-biosensor (surface plasmon resonance, SPR)
[20, 21]. Here, antibodies are used on a thin metal layer, such as nanogold particles.
Light falling on this layer leads to a characteristic coloration of the layer by
interacting with the surface plasmons of the metal-dielectric layer. The binding of
antigen on this layer leads to a tiny change in the refractive index of the layer and, as
a result, to a change in resonance frequency of the surface plasmons, combined with
a change in the color of the layer. In this way, the presence of the analyte can be
indicated by a color change. This principle of changing the refractive index of thin
metal layers can also be used by attaching the bio-sensitive metal layer to a prism. In
this way, antigen–antibody, as well as protein–DNA or protein–protein interactions
can be detected [22]. Numerous bio-sensitive layers for different analytes can be
optically evaluated simultaneously (SPR-Imaging, multichannel bio-sensor)
[23]. Further examples of optical bio-sensors can be found in [24].
In addition to refraction, absorption of electromagnetic waves offers numerous
approaches for the construction of on-line sensors. These include sensors for on-line
color measurement [25], see Chap. 12, and on-line sensors for NIR analysis of food
[26]. Electronic cameras and associated on-line image processing techniques also
belong to this category. An overview of miniaturized sensors on an optical basis for
on-line quality assurance can be found in [27].
572 16 On-line Sensors

16.4 Chemo- and Bio-sensors

In the previous sections, we were briefly introduced to biosensors. Now we want to

study the basic principles that underlie the construction of a bio-sensor or chemo-
sensor. We want to focus on systems that are suitable as on-line sensors, i.e. fast
sensors whose output signal is transmitted electrically.

Definition
A chemo-sensor is a sensor for detecting a specific chemical molecule or group
of molecules. A bio-sensor is a sensor for detecting a certain biological
material, such as an antibody. Since the detection is often based on a typical
protein, bio-sensors are a special group of chemo-sensors.

If the substance to be detected (analyte) is a microbiological body or a chemical

molecule, a specific receptor for the analyte is required at the sensor. Alternatively, a
specific reaction on a sensor is required, which takes place as soon as the analyte
reaches the sensor.
The function of the receptor can be realized by a bio-sensitive or chemo-sensitive
layer, which is attached to the sensor. This sensitive layer contains the components
necessary for bio-recognition of an enzyme or other biochemical. This sensitive
layer is firmly attached to the sensor and ensures the immobilization of the
components necessary for bio-detection so that they are not desorbed or washed out.

Definition
Analyte is the generic term for what is to be detected, e.g. a chemical element,
a group of substances, a gas, a protein, a microorganism, a virus, a cell.

If this sensitive layer comes into contact with a fluid (gas phase or liquid)
containing the analyte, the analyte is bound to the film as a result of
bio-recognition, which slightly changes the properties of the film. This change in
property must now be measured and converted into an electrical signal. Possibilities
that we have already learned are the measurement of mechanical resonance fre-
quency, or electrical capacitance or impedance, or optical properties such as
refraction.
There are numerous possibilities for the selective or specific binding of the
analyte to the receptor. These include physisorption, chemisorption (see Chap. 1),
enzymatic reaction, antibody–antigen binding, binding to surface proteins of
microorganisms, viruses, and much more. In addition to high sensitivity, high
specificity is often needed for recognition of the analyte (bio-recognition). This
recognition must be as precise as possible to avoid mistaken recognition of related
structures with similar properties.
16.4 Chemo- and Bio-sensors 573

Definition
Specificity refers to the ability to distinguish analyte and non-analyte.
Detecting a particular protein requires a higher specificity of the detection
method than if a group of similar proteins are to be detected in a sample.

Definition
Selectivity is the property of selecting specific objects from a set of objects.
Example: An ion-selective sensor detects only substances present as ions.

High specificity of the sensitive layer of chemo-sensors is achieved, e.g., by

catalysts that only allow certain reactions or nanoscale cavities with a size and shape
in which only the analyte molecule fits. Polymers with molecularly matching
cavities are called MIP (molecular imprintable polymer). In biosensors, specificity
is achieved by enzyme reactions or specific protein bonds. In this way, specificity
can even be achieved for a specific DNA section, or nucleic acid, virus, bacterium,
antibody, immunoglobulin, or similar [28]. When using aptamers (Latin: aptus,
fitting, Greek: μερoς: part) as binding partners in the sensitive layer, biosensors
are also referred to as apta-sensors. Aptamers can be short single-stranded nucleic
acid oligomers, which lead to high specificity in detecting certain proteins, such as
antibodies or viruses and thus play an increasing role in food safety [29–35].

Definition
Aptamer and analyte are two molecules that fit together like a lock and a key.
Aptamers can be DNA or RNA segments.

Once a suitable mechanism for bio-recognition has been found, the next step is to
generate a physical measurement signal from it. Table 16.7 lists examples of sensor
principles that are available for this purpose.

Definition
Sensitivity of a measurement method refers to the slope of the calibration
curve, i.e. the ratio of the measurement signal of the sensor to the concentra-
tion of the analyte.

Table 16.7 shows that there are numerous principles for measuring the physical
changes of an analyte-sensitive layer. They differ in sensitivity and technical effort in
miniaturization. Miniaturized, electro-mechanical systems (Micro-Electro-Mechani-
cal Systems, MEMS), in which the analyte-sensitive layer and the measurement
574 16 On-line Sensors

Table 16.7 Measurement principles for chemo- and biosensors, examples

Sensor principle Description What the analyte changes . . .
Thermal, Temperature sensor with chemo- Temperature of the sensor
microcalorimetry sensitive layer, e.g. with catalyst or
enzyme
Mechanical Bending oscillator with sensitive layer Resonance frequency
oscillation
Optical Interaction of the analyte with Intensity, polarization, wave
electromagnetic radiation propagation speed
Infared (IR) and Analytes absorb typical IR or NIR bands Intensity of certain
NIR wavelengths
Color change Reaction of the analyte causes color Intensity of certain
change wavelengths, e.g., 615 nm,
530 nm, 460 nm
Optical absorption Absorption of typical wavelengths is Ratio of absorption of two
ratio compared wavelengths
Metal oxide Analyte is chemisorbed on MOS surface Electrical resistance of MOS
semiconductor
(MOS)
Molecular Analyte is absorbed in molecule- Mass of MIP
imprintable matching cavities of the polymer
polymer (MIP)
Polymer Polymer film absorbs analyte Electrical resistance of the
conductive PC
composites (PC)
Chemo-capacitive Nonconductor or dielectric polymer Electrical permittivity
absorbs analytes
ChemFET Modified field-effect transistor (FET), Electrical resistance between
gate with chemo-sensitive layer source and drain
ISFET Ion-selective ChemFET Electrical resistance between
source and drain
PID (photo UV radiation produces ions from analyte Electric current of the
ionization detector) electrometer
Potentiometric Analyte reacts to electrochemical Electrical voltage at the
electrode electrode
Amperometric Analyte reacts to electrochemical Electric current at the
electrode electrode
Conductometric Analyte reacts to electrochemical Electrical conductivity at the
electrode electrode

technology are close together, can be used to manufacture sensor chips [36–41]. For
example, small mechanically oscillating platelets, the so-called micro-cantilevers
with a thickness of less than 1 μm, can be provided with a functionalized surface.
When analyte molecules are attached to the functionalized surface, the resonance
frequency of the oscillating platelet changes. Such systems are also referred to as
microbalance sensors (gravimetric sensors) or MEMS cantilever sensors. If the
oscillation is between a few kHz and a few GHz, they are also referred to as acoustic
16.4 Chemo- and Bio-sensors 575

devices. The resonance is measured physically by laser light, but can also be
piezoresistive, capacitive or thermal, which significantly reduces the technical effort.

Definition
A functionalized surface in bio-sensors and chemo-sensors refers to a coating
upon which the surface specifically binds analytes, such as proteins.

Miniaturization of sensors has led to the fact that not only one, but numerous
sensors can be mounted on a chip. The so-called multi-sensor arrays allow simulta-
neous measurement of numerous analytes or multiple measurements of an analyte to
reduce measurement uncertainty [42]. The so-called e-nose and the electronic tongue
are examples of such multi-sensor arrays with which numerous flavoring substances
and flavors are determined at the same time [43, 44].
The main advantage of on-line chemo-sensors and bio-sensors lies in the combi-
nation of high specificity through the functionalization of the sensitive surface
coating with sensitive and fast physical measurement technology. A challenge is
to make the sensitive layer sufficiently stable and robust so that many measurements
can be carried out and even harsh process conditions can be overcome. Cleaning
cycles with hot solutions, acids, and alkali often are difficult hurdles for sensor
development.
Up to this point, we have already used numerous terms of measurement technol-
ogy (metrology) such as sensitivity, selectivity, specificity. Short definitions to
facilitate the distinction between these terms are summarized in Table 16.8.
As stated earlier, the term accuracy is often used in the sense of trueness (see
Table 16.8) and there is no distinction between them. However, the terms accuracy
and precision have significantly different meanings and must be kept apart to avoid
confusion. The definitions of the terms in Table 16.8 the International Vocabulary of
Metrology of BIPM [45]. Figure 16.10 helps to understand and distinguish between
the terms precision and trueness.
Another conceptual distinction is the absolute measurement and relative measure-
ment of quantities. Relative measurements are quotients of measured value and the
value of a reference material or a reference point. Thus, the results of relative
measurements are only ratios without physical unit. Absolute measurements, on
the other hand, have a physical unit. For passing on to third parties, quantities with
unit are recommended. For internal quality monitoring or for the electronic control
of processes, the use of absolute measurements is not necessary and relative
measurements are preferred because they are easier to handle.
In food technology, the term “sensorial analysis” is used for organoleptic testing
of food, such as in the assessment of color, smell, taste, texture, etc. Organoleptic
sensing is a method of testing samples using human senses and should not be
confused with the type of sensor technology presented in this chapter. Occasionally
there are points of contact between the two disciplines, such as when comparing
576 16 On-line Sensors

Table 16.8 Terms of metrology, according to [45]

Selectivity Property to select multiple specific objects from a set of similar objects. Ability
to distinguish specific substances from other similar substances. Determining
the suitability of an analytical method (IUPAC definition)
Sensitivity Ratio of signal strength of a sensor to the concentration of the analytes
Specificity Property to select an object from a set of similar objects. Ability to clearly
identify the analyte among other similar substances.
Detection limit Lowest concentration that can be reliably detected
Accuracy Degree of agreement between displayed reading and true value. High accuracy
requires high precision and high trueness [46].
Precision Spread of the measured values obtained from repeated measurements, “how
close the measured values are to each other.”
Trueness Degree of agreement between measured value and true value. A low
systematic measurement error leads to a high trueness
Stability Property of a sensor to maintain its sensitivity and specificity over a long
period of time
Resolution Smallest difference between two measured values that the measuring
instrument can display
Repeatability Quantitative indication of obtaining the same measurement result repeatability
at the same location, with the same experimenter, under the same conditions
Reproducibility Quantitative indication of obtaining the same measurement result at another
time and location with another experimenter.

High trueness and high precision High trueness and low precision

Low trueness and high precision Low trueness and low precision

Fig. 16.10 Illustration of precision and trueness in measurement. The diagrams show the related
frequency density q0 the measured value x, schematic. The spread among the dots on the target disks
and spread along the x-axis illustrate precision. The closeness T of the measured value to the true
value (R) on the graph illustrates trueness

measurements of electrical noses and tongues with organoleptic assessments [47–

51].
Detection methods that work at the molecular level are referred to as
metabolomics or lipidomics, proteomics, genomics, or transcriptomics depending
16.4 Chemo- and Bio-sensors 577

on the analytical objective. The term foodomics has been recently introduced for
food-related detection methods at the molecular level, and the term flavoromics
methods for flavor-related detection methods. Detection methods for investigating
the molecular causes of organoleptic-sensory effects of food are accordingly referred
to as sensomics [52, 53].

Definition
Sensomics is a generic term for molecular analytical methods that help us to
understand the organoleptic-sensory effects of food.

Further Reading

Topic Refs.
Magnetic resonance tomography for in-line detection of unwebbed eggs [54]
Electrochemical sensor based on nano Fe3O4 for detection of azo colorant [55]
E110
Walnuts, kernels, and shells: Quality assurance using terahertz spectroscopy [56]
Getting selectivity by magnetic nanoparticles in biosensors [57]
Simple LED optics for detecting bio-active compounds in fruits [58]
Semiconductor material as refraction sensor in terahertz band [59]
Electrochemical immunosensor for the detection of acrylamide [60]
Biosensoren based on acoustic resonatoren [61]
Detection of pathogenic microorganisms using MOS sensors [62]
Biosensor for the detection of Staphylococcus aureus [63]
Nanosensors for the detection of pesticides in water [64]
Use of nanofibers in biosensors for food and agar products [65]
Chemo-sensors for food analysis based on nano materials [66]
Bacteria detection by nano biosensor [67, 68]
Piezoelectric immunosensor: kinetic investigation of biosensor function [69]
Apta-sensors for the detection of mycotoxins and antibiotics [70, 71]
Protein-sensitive chips with terahertz technology [72]
On-line monitoring of the freezing process by means of microwaves [73]
NMR-Sensor for the fat content in tuna fillet [74]
Riboflavin: Voltammetric sensor [75]
Mobile acoustic sensor for moisture determination in soils [76]
MEMS cantilever biosensor [77–79]
Immunosensor for the detection of antibiotics in milk [80]
On-chip moisture measurement based on permittivity [14]
Moisture measurement in the fluidized bed using a dielectric in-line sensor [15]
Refraction sensors with fiber optics [81]
Immunoglobulin biosensor based on ellipsometry [28]
Refraction sensors by surface plasmon resonance on gold-coated nanofilm [20, 82]

(continued)
578 16 On-line Sensors

Topic Refs.
Sweet cherry: Hue fingerprinting during storage [83]
LED-based optical sensor for the detection of vitamin C in fruits [58]
Surface plasmon resonance biosensor for sucrose in urine [84]
On-line NIR for process analysis in the dairy industry [85]
Electrical impedance tomography with artificial intelligence [86]

16.5 Abstract

On-line sensors are fast sensors that transmit an electrical signal in time to take action
that assures uninterrupted continuous production in automated process operations. In
this chapter, associated metrological terms are presented and explained using simple
examples. The function of some on-line sensors is presented to illustrate the different
measurement principles. In the section of chemo- and bio-sensors, the acquired
knowledge flows together. At the end of the chapter, works on novel sensors are
listed, which can be used for further studies and as suggestions for one’s own
scientific work. Readers with advanced knowledge may choose to read this last
chapter first, and then study the underlying basic principles that are covered in the
previous chapters.

References
1. Varzakas T, Nikoleli GP, Nikolelis DP (2015) Biosensors in food technology, safety, and
quality control, pp 675–690. doi:https://doi.org/10.1201/b19397
2. Zou X, Huang X, Zhai X, Shi J, Povey M (2020) Non-invasive sensing for food reassurance.
Physics in Food Manufacturing. IOP Publishing. doi:https://doi.org/10.1088/978-0-7503-2596-
7ch10
3. Dumitru LM, Irimia-Vladu M, Sariciftci NS (2016) In: Scognamiglio V, Rea G, Arduini F,
Palleschi G (eds) Biocompatible integration of electronics into food sensors, vol 74. Elsevier, pp
247–271. https://doi.org/10.1016/bs.coac.2016.04.009
4. Gaudin V (2017) Advances in biosensor development for the screening of antibiotic residues in
food products of animal origin - a comprehensive review. Biosens Bioelectron 90:363. https://
doi.org/10.1016/j.bios.2016.12.005
5. Rahman MSA, Mukhopadhyay SC, Yu P-L (2014) Novel sensors for food inspection:
modelling, fabrication and experimentation. Springer, Cham
6. Li J, Li Z, Wang N, Raghavan GSV, Pei Y, Song C, Zhu G (2020) Novel sensing technologies
during the food drying process. Food Eng Rev 12(2):121. https://doi.org/10.1007/s12393-020-
09215-2
7. Compagnone D, Baldini F, Di Natale C, Betta G, Siciliano P (2014) Sensors. Second national
conference on sensors. Rome
8. Fraden J (2016) Handbook of modern sensors physics, designs, and applications. https://doi.
org/10.1007/978-3-319-19303-8
References 579

9. Reindl LM, Tränkler H-R (2014) Sensortechnik. Springer-Vieweg, Heidelberg. https://doi.org/

10.1007/978-3-642-29942-1
10. Cullinan MA, Culpepper ML (2010) Carbon nanotubes as piezoresistive
microelectromechanical sensors: theory and experiment. Phys Rev B 82(11):115428. https://
doi.org/10.1103/PhysRevB.82.115428
11. Itoh A, Motoko I (2011) Separate measurement of the density and viscosity of a liquid using a
quartz crystal microbalance based on admittance analysis (QCM-A). Meas Sci Technol 22
12. Manzaneque T, Ruiz-Diez V, Hernando-Garcia J, Wistrela E, Kucera M, Schmid U, Sanchez-
Rojas JL (2014) Piezoelectric MEMS resonator-based oscillator for density and viscosity
sensing. Sens Actuator A Phys 220:305. https://doi.org/10.1016/j.sna.2014.10.002
13. Cerimovic S, Beigelbeck R, Antlinger H, Schalko J, Jakoby B, Keplinger F (2012) Sensing
viscosity and density of glycerol-water mixtures utilizing a suspended plate MEMS resonator.
Microsyst Technol Micro Nanosyst -Inf Storage Process Syst 18(7-8):1045. https://doi.org/10.
1007/s00542-012-1437-9
14. Kozhevnikov A (2010) Wideband radio-frequency device for measurements of dielectric
properties of small volumes of liquids. Meas Sci Technol 21(4):043001. https://doi.org/10.
1088/0957-0233/21/4/043001
15. Peters J, Bartscher K, Doscher C, Taute W, Hoft M, Knochel R, Breitkreutz J (2017) In-line
moisture monitoring in fluidized bed granulation using a novel multi-resonance microwave
sensor. Talanta 170(Suppl C):369. https://doi.org/10.1016/j.talanta.2017.03.105
16. Krommenhoek EE, Gardeniers JGE, Bomer JG, Van den Berg A, Li X, Ottens M, van der
Wielen LAM, van Dedem GWK, Van Leeuwen M, van Gulik WM, Heijnen JJ (2006)
Monitoring of yeast cell concentration using a micromachined impedance sensor. Sensors
Actuators B Chem 115(1):384. https://doi.org/10.1016/j.snb.2005.09.028
17. Yang L, Bashir R (2008) Electrical/electrochemical impedance for rapid detection of foodborne
pathogenic bacteria. Biotechnol Adv 26(2):135. https://doi.org/10.1016/j.biotechadv.2007.
10.003
18. Chen Y, Liu J, Yang Z, Wilkinson JS, Zhou X (2019) Optical biosensors based on refractomet-
ric sensing schemes: a review. Biosens Bioelectron 144:111693. https://doi.org/10.1016/j.bios.
2019.111693
19. Khansili N, Rattu G, Krishna PM (2018) Label-free optical biosensors for food and biological
sensor applications. Sens Actuator B Chem 265:35. https://doi.org/10.1016/j.snb.2018.03.004
20. Abdullah H, Ahmed K, Mitu SA (2020) Ultrahigh sensitivity refractive index biosensor based
on gold coated nano-film photonic crystal fiber. Results Phys 17:103151. https://doi.org/10.
1016/j.rinp.2020.103151
21. Zhao J, Cao SQ, Liao CR, Wang Y, Wang GJ, Xu XZ, Fu CL, Xu GW, Lian JR, Wang YP
(2016) Surface plasmon resonance refractive sensor based on silver-coated side-polished fiber.
Sens Actuator B Chem 230:206. https://doi.org/10.1016/j.snb.2016.02.020
22. Kretschmann E (1971) Die Bestimmung optischer Konstanten von Metallen durch Anregung
von Oberflächenplasmaschwingungen. Eur Phys J A Hadron Nucl 241:313. https://doi.org/10.
1007/BF01395428
23. Guo X (2012) Surface plasmon resonance based biosensor technique: a review. J Biophotonics
5(7):483. https://doi.org/10.1002/jbio.201200015
24. Narsaiah K, Jha SN, Bhardwaj R, Sharma R, Kumar R (2012) Optical biosensors for food
quality and safety assurance—a review. J Food Sci Technol 49(4):383. https://doi.org/10.1007/
s13197-011-0437-6
25. Shende P, Prabhakar B, Patil A (2019) Color changing sensors: a multimodal system for
integrated screening. TrAC Trends Anal Chem 121:115687. https://doi.org/10.1016/j.trac.
2019.115687
26. Porep JU, Kammerer DR, Carle R (2015) On-line application of near infrared (NIR) spectros-
copy in food production. Trends Food Sci Technol 46(2, Part A):211. https://doi.org/10.1016/j.
tifs.2015.10.002
580 16 On-line Sensors

27. Rodriguez-Saona L, Aykas DP, Borba KR, Urtubia A (2020) Miniaturization of optical sensors
and their potential for high-throughput screening of foods. Curr Opin Food Sci 31:136. https://
doi.org/10.1016/j.cofs.2020.04.008
28. Li K, Wang S, Wang L, Yu H, Jing N, Xue R, Wang Z (2017) Fast and sensitive ellipsometry-
based biosensing. Sensors (Basel, Switzerland) 18(1):15. https://doi.org/10.3390/s18010015
29. Taghdisi SM, Danesh NM, Ramezani M, Alibolandi M, Nameghi MA, Gerayelou G, Abnous K
(2021) A novel electrochemical aptasensor for ochratoxin a sensing in spiked food using strand-
displacement polymerase reaction. Talanta 223:121705. https://doi.org/10.1016/j.talanta.2020.
121705
30. Liu R, Zhang Y, Ali S, Haruna SA, He P, Li H, Ouyang Q, Chen Q (2021) Development of a
fluorescence aptasensor for rapid and sensitive detection of Listeria monocytogenes in food.
Food Control 122:107808. https://doi.org/10.1016/j.foodcont.2020.107808
31. Rajabnejad S-H, Badibostan H, Verdian A, Karimi GR, Fooladi E, Feizy J (2020) Aptasensors
as promising new tools in bisphenol A detection - an invisible pollution in food and environ-
ment. Microchem J 155:104722. https://doi.org/10.1016/j.microc.2020.104722
32. Lu Y, Yuan Z, Bai J, Lin Q, Deng R, Luo A, Chi Y, Deng S, He Q (2020) Directly profiling
intact Staphylococcus aureus in water and foods via enzymatic cleavage aptasensor. Anal Chim
Acta 1132:28. https://doi.org/10.1016/j.aca.2020.07.058
33. Jia Y, Wu F, Liu P, Zhou G, Yu B, Lou X, Xia F (2019) A label-free fluorescent aptasensor for
the detection of Aflatoxin B1 in food samples using AIEgens and graphene oxide. Talanta 198:
71. https://doi.org/10.1016/j.talanta.2019.01.078
34. Feng J, Shen Q, Wu J, Dai Z, Wang Y (2019) Naked-eyes detection of Shigella flexneri in food
samples based on a novel gold nanoparticle-based colorimetric aptasensor. Food Control 98:
333. https://doi.org/10.1016/j.foodcont.2018.11.048
35. Majdinasab M, Hayat A, Marty JL (2018) Aptamer-based assays and aptasensors for detection
of pathogenic bacteria in food samples. TrAC Trends Anal Chem 107:60. https://doi.org/10.
1016/j.trac.2018.07.016
36. Neethirajan S, Weng X, Tah A, Cordero JO, Ragavan KV (2018) Nano-biosensor platforms for
detecting food allergens – new trends. Sens Bio-Sens Res 18:13. https://doi.org/10.1016/j.sbsr.
2018.02.005
37. Lv M, Liu Y, Geng J, Kou X, Xin Z, Yang D (2018) Engineering nanomaterials-based
biosensors for food safety detection. Biosens Bioelectron 106:122. https://doi.org/10.1016/j.
bios.2018.01.049
38. Omanović E, Maksimović M (2016) Nanosensors applications in agriculture and food industry.
Bull Chem Technol Bosnia Herzegovina 47:59
39. Bülbül G, Hayat A, Andreescu S (2015) Portable nanoparticle-based sensors for food safety
assessment. Sensors 15(12). https://doi.org/10.3390/s151229826
40. Kobun R (2021) Nanotechnology-based optical biosensors for food applications. In: Advanced
food analysis tools. Academic Press, pp 147–165. https://doi.org/10.1016/B978-0-12-
820591-4.00008-6
41. Toydemir G (2020) Nanosensors for foods. In: Hebbar U, Ranjan S, Dasgupta N, Kumar Mishra
R (eds) Nano-food engineering. Springer, Cham. https://doi.org/10.1007/978-3-030-44552-2_
12
42. Geng Y, Peveler WJ, Rotello VM (2019) Array-basierte Sensorik mit der “chemischen Nase” in
der Diagnostik und Wirkstoffentdeckung. Angew Chem 131(16):5244. https://doi.org/10.1002/
ange.201809607
43. Banerjee R, Tudu B, Bandyopadhyay R, Bhattacharyya N (2016) A review on combined odor
and taste sensor systems. J Food Eng 190:10. https://doi.org/10.1016/j.jfoodeng.2016.06.001
44. Zhong J, Wang X (eds) (2019) Evaluation technologies for food quality. Woodhead publishing
series in food science, technology and nutrition. Woodhead Publishing. https://doi.org/10.1016/
C2017-0-01187-4
45. JCGM (2008) International vocabulary of metrology — Basic and general concepts and
associated terms (VIM). Bureau International de Poids et Mesures BIPM, Versailles
References 581

46. DIN EN 60051 Direkt wirkend anzeigende analoge elektrische Messgeräte und ihr Zubehör
(2017) Beuth, Berlin
47. Xu J, Liu K, Zhang C (2021) Electronic nose for volatile organic compounds analysis in rice
aging. Trends Food Sci Technol 109:83. https://doi.org/10.1016/j.tifs.2021.01.027
48. Teixeira GG, Dias LG, Rodrigues N, Marx ÍMG, Veloso ACA, Pereira JA, Peres AM (2021)
Application of a lab-made electronic nose for extra virgin olive oils commercial classification
according to the perceived fruitiness intensity. Talanta 226:122122. https://doi.org/10.1016/j.
talanta.2021.122122
49. Fedorov FS, Yaqin A, Krasnikov DV, Kondrashov VA, Ovchinnikov G, Kostyukevich Y,
Osipenko S, Nasibulin AG (2021) Detecting cooking state of grilled chicken by electronic nose
and computer vision techniques. Food Chem 345:128747. https://doi.org/10.1016/j.foodchem.
2020.128747
50. Chen Q, Hu Y, Wen R, Wang Y, Qin L, Kong B (2021) Characterisation of the flavour profile of
dry fermented sausages with different NaCl substitutes using HS-SPME-GC-MS combined
with electronic nose and electronic tongue. Meat Sci 172:108338. https://doi.org/10.1016/j.
meatsci.2020.108338
51. Ghasemi-Varnamkhasti M, Apetrei C, Lozano J, Anyogu A (2018) Potential use of electronic
noses, electronic tongues and biosensors as multisensor systems for spoilage examination in
foods. Trends Food Sci Technol 80:71. https://doi.org/10.1016/j.tifs.2018.07.018
52. Vrzal T, Olšovská J (2019) Sensomics - basic principles and practice. KVASNÝ PRŮMYSL
65(5):166. https://doi.org/10.18832/kp2019.65.166
53. Chapman J, Elbourne A, Truong VK, Newman L, Gangadoo S, Rajapaksha Pathirannahalage P,
Cheeseman S, Cozzolino D (2019) Sensomics - from conventional to functional NIR spectros-
copy - shining light over the aroma and taste of foods. Trends Food Sci Technol 91:274. https://
doi.org/10.1016/j.tifs.2019.07.013
54. Durchbruch bei Suche nach Alternative zum Kükentöten. (2018) TU München. https://www.
tum.de/nc/die-tum/aktuelles/pressemitteilungen/details/34775/
55. Shaikshavali P, Reddy TM, Lakshmi Narayana A, Hussain OM, Venkataprasad G, Venu Gopal
T (2020) A powerful electrochemical sensor based on Fe3O4 nanoparticles-multiwalled carbon
nanotubes hybrid for the effective monitoring of sunset yellow in soft drinks. J Food Meas
Charact 14(6):3319. https://doi.org/10.1007/s11694-020-00569-z
56. Wang Q, Hameed S, Xie L, Ying Y (2020) Non-destructive quality control detection of
endogenous contaminations in walnuts using terahertz spectroscopic imaging. J Food Meas
Charact 14(5):2453. https://doi.org/10.1007/s11694-020-00493-2
57. Wang L, Lin JH (2020) Recent advances on magnetic nanobead based biosensors: from
separation to detection. Trac Trends Anal Chem 128:115915. https://doi.org/10.1016/j.trac.
2020.115915
58. Miranda JIS, Lopes MMA, de Souza KO, Pinho HA, Rabelo MC, de Moura LP, Eça KS, de
Miranda MRA, Oliveira LS, Guimarães GF (2020) A low-cost optical sensor to quantify
bioactive compounds in fruit. J Food Meas Charact 14(6):3580. https://doi.org/10.1007/
s11694-020-00601-2
59. Aslinezhad M (2020) High sensitivity refractive index and temperature sensor based on
semiconductor metamaterial perfect absorber in the terahertz band. Opt Commun 463:
125411. https://doi.org/10.1016/j.optcom.2020.125411
60. Wu MF, Wang Y, Li S, Dong XX, Yang JY, Shen YD, Wang H, Sun YM, Lei HT, Xu ZL
(2019) Ultrasensitive immunosensor for acrylamide based on chitosan/SnO2-SiC hollow sphere
nanochains/gold nanomaterial as signal amplification. Anal Chim Acta 1049:188. https://doi.
org/10.1016/j.aca.2018.10.041
61. Zhang Y, Luo J, Flewitt AJ, Cai Z, Zhao X (2018) Film bulk acoustic resonators (FBARs) as
biosensors: a review. Biosens Bioelectron 116:1. https://doi.org/10.1016/j.bios.2018.05.028
62. Viswanath B, Kristine YM, Kim S (2018) Recent trends in the development of complementary
metal oxide semiconductor image sensors to detect foodborne bacterial pathogens. TrAC
Trends Anal Chem 98:47. https://doi.org/10.1016/j.trac.2017.10.019
582 16 On-line Sensors

63. Rubab M, Shahbaz HM, Olaimat AN, Oh D-H (2018) Biosensors for rapid and sensitive
detection of Staphylococcus aureus in food. Biosens Bioelectron 105:49. https://doi.org/10.
1016/j.bios.2018.01.023
64. Steffens C, Steffens J, Marcia Graboski A, Manzoli A, Lima Leite F (2017) Nanosensors for
detection of pesticides in water. Academic, pp 595–635. https://doi.org/10.1016/b978-0-12-
804299-1.00017-5
65. Mercante LA, Scagion VP, Migliorini FL, Mattoso LHC, Correa DS (2017) Electrospinning-
based (bio)sensors for food and agricultural applications: a review. Trac Trends Anal Chem 91:
91. https://doi.org/10.1016/j.trac.2017.04.004
66. Correa DS, Pavinatto A, Mercante LA, Mattoso LHC, Oliveira JE, Riul A (2017) Chemical
sensors based on hybrid nanomaterials for food analysis. Academic, pp 205–244. https://doi.
org/10.1016/b978-0-12-804301-1.00006-0
67. Bhardwaj N, Bhardwaj SK, Nayak MK, Mehta J, Kim KH, Deep A (2017) Fluorescent
nanobiosensors for the targeted detection of foodborne bacteria. Trac Trends Anal Chem 97:
120. https://doi.org/10.1016/j.trac.2017.09.010
68. Stephen Inbaraj B, Chen BH (2016) Nanomaterial-based sensors for detection of foodborne
bacterial pathogens and toxins as well as pork adulteration in meat products. J Food Drug Anal
24(1):15. https://doi.org/10.1016/j.jfda.2015.05.001
69. Chauhan R, Singh J, Solanki PR, Manaka T, Iwamoto M, Basu T, Malhotra BD (2016) Label-
free piezoelectric immunosensor decorated with gold nanoparticles: kinetic analysis and
biosensing application. Sens Actuators B Chem 222:804. https://doi.org/10.1016/j.snb.2015.
08.117
70. Zhang N, Liu B, Cui X, Li Y, Tang J, Wang H, Zhang D, Li Z (2021) Recent advances in
aptasensors for mycotoxin detection: on the surface and in the colloid. Talanta 223:121729.
https://doi.org/10.1016/j.talanta.2020.121729
71. Jiang Y, Sun D-W, Pu H, Wei Q (2020) A simple and sensitive aptasensor based on SERS for
trace analysis of kanamycin in milk. J Food Meas Charact 14(6):3184. https://doi.org/10.1007/
s11694-020-00553-7
72. Mickan S, Lee K-S, Lu T-M, Munch J, Abbott D, Zhang XC (2013) Double modulated
differential THz-TDS for thin film dielectric characterization. Microelectron J 33:1033.
https://doi.org/10.1016/S0026-2692(02)00108-8
73. Pandey G, Vandermeiren W, Dimiccoli L, Stiens J (2018) Contactless monitoring of food
drying and freezing processes with millimeter waves. J Food Eng 226:1. https://doi.org/10.
1016/j.jfoodeng.2018.01.003
74. Nakashima Y (2020) Development of a hand-held magnetic resonance sensor for the nonde-
structive quantification of fat and lean meat of fresh tuna. J Food Meas Charact 14(6):2947.
https://doi.org/10.1007/s11694-020-00539-5
75. Monnappa AB, Manjunatha JGG, Bhatt AS, Ananda PP (2020) Fabrication of a sensitive and
selective electrochemical sensing platform based on poly-l-leucine modified sensor for
enhanced voltammetric determination of Riboflavin. J Food Meas Charact 14(6):3633.
https://doi.org/10.1007/s11694-020-00608-9
76. Gorthi S, Chakraborty S, Li B, Weindorf DC (2020) A field-portable acoustic sensing device to
measure soil moisture. Comput Electron Agric 174:105517. https://doi.org/10.1016/j.compag.
2020.105517
77. Pinto RMR, Chu V, Conde JP (2020) Label-free biosensing of DNA in microfluidics using
amorphous silicon capacitive micro-cantilevers. IEEE Sensors J 20(16):9018. https://doi.org/
10.1109/JSEN.2020.2986497
78. Priyadarsini S, Das JK, Dastidar A (2016) Analysis of MEMS cantilever geometry for designing
of an array sensor. In: 2016 International conference on signal processing, communication,
power and embedded system (SCOPES), 3–5 Oct 2016, pp 625–628. https://doi.org/10.1109/
SCOPES.2016.7955515
References 583

79. Parsediya DK, Singh J, Kankar PK (2014) Simulation and analysis of highly sensitive MEMS
cantilever designs for “in vivo label free” biosensing. Proc Technol 14:85. https://doi.org/10.
1016/j.protcy.2014.08.012
80. Strasser A (2003) Entwicklung eines Biosensors zum Nachweis von Antibiotika und
Sulfonamiden in Milch - Herstellung der immunchemischen Komponenten. Doctoral, Techni-
cal University, Munich
81. Li J (2020) A review: development of novel fiber-optic platforms for bulk and surface refractive
index sensing applications. Sens Actuators Rep 2(1):100018. https://doi.org/10.1016/j.snr.
2020.100018
82. Wu Y, Liu B, Wang J, Wu J, Mao Y, Ren J, Zhao L, Sun T, Nan T, Han Y, Liu X (2021) A
novel temperature insensitive refractive index sensor based on dual photonic crystal fiber. Optik
226:165495. https://doi.org/10.1016/j.ijleo.2020.165495
83. Le Nguyen LP, Visy A, Baranyai L, Friedrich L, Mahajan PV (2020) Application of hue spectra
fingerprinting during cold storage and shelf-life of packaged sweet cherry. J Food Meas Charact
14(5):2689. https://doi.org/10.1007/s11694-020-00515-z
84. Yadav A, Sharan P, Kumar A (2020) Surface plasmonic resonance based five layered structure-
biosensor for sugar level measurement in human. Results Opt 1:100002. https://doi.org/10.
1016/j.rio.2020.100002
85. Pu Y-Y, O'Donnell C, Tobin JT, O'Shea N (2020) Review of near-infrared spectroscopy as a
process analytical technology for real-time product monitoring in dairy processing. Int Dairy J
103:104623. https://doi.org/10.1016/j.idairyj.2019.104623
86. Khan TA, Ling S (2019) Review on electrical impedance tomography: artificial intelligence
methods and its applications. Algorithms 12:88. https://doi.org/10.3390/a12050088
Appendix A

A.1 The International System of Units (SI)

The international system of units (SI, French: le Systeme International d’Unites) is

the modern metric system of measurement and used worldwide. There are 7 basic
units, which are defined by fundamental physical constants, see Tables A.1 and A.2.
From these basic units numerous derived units can be formed, see Table A.3, A.4,
A.5, A.6, A.7, A.8, A.9, and A.10.

Table A.1 Units of SI and their definition [1]

Basic unit Symbol for unit Definition
Kilogram kg 1kg = 1:475521
1040
h
Δν
c2
Meter m 1 m = 30.663318c/Δν
Second s 1 s = 9192631770/Δν
Mol mol 1 mol = 6.022140857
1023/NA
Ampere A 1 A = 6.789687
108
Δν
e
Kelvin K 1 K = 2.266665
Δν
h/k
Candela cd 1 cd = 2.614830
1010
Δν2
h
Kcd

Table A.2 Fundamental physical constants used for SI basic units

h The Planck constant: h = 6.62607015
10–34 J
s
c Speed of light in vacuum: c = 299792458 m
s–1
Δν Frequency of the hyperfine transition of 133Cs: Δν = 9192631770 s–1
e Elementary charge: e = 1.602 176 634
10–19 C
k The Boltzmann constant: k = 1.380 649
10–23 J
K–1
NA The Avogadro constant: NA = 6.02214076
1023 mol–1
Kcd The luminous efficacy of monochromatic radiation of frequency 540
1012 Hz: Kcd =
683 lm
W-1

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 585
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8
586 Appendix A

Table A.3 Derived SI Quantity Symbol

units
Area m2
Volume m3
Velocity m/s
Acceleration m/s2
Wavenumber m-1
Density kg m-3
Specific volume m3 kg-1
Electric current density A m-2
Magnetic field strength A m-1
Substance concentration mol m-3
Luminance cd m-2

Table A.4 Derived units with special names and symbols

SI derived unit
Expressed in terms of Expressed in terms of
Quantity Name Symbol other SI units SI base units
Plane angle radian rad m m-1 = 1
Solid angle steradian Sr m2 m-2 = 1
Frequency Hertz Hz s-1
Force Newton N m kg s-2
Pressure, stress Pascal Pa N m-2 m-1 kg s-2
Energy, work, Joule J Nm m2 kg s-2
heat
Power Watt W J s-1 m2 kg s-3
Electric charge Coulomb C As
Electric potential Volt V W A-1 m kg s-3 A-1
difference
Capacitance Farad F C V-1 m-2 kg-1 s4 A2
Electric resistance Ohm Ω V A-1 m2 kg s-3 A-2
Electric Siemens S A V-1 m-2 kg-1 s3 A2
conductance
Magnetic flux Weber Wb Vs m2 kg s-2 A-1
Magnetic flux Tesla T Wb m-2 kg s-2 A-1
density
Inductance Henry H Wb A-1 m2 kg s-2 A-2
Celsius °Celsius °C K
temperature
Luminous flux lumen lm cd sr
Illuminance lux lx lm/m2 cd sr m-2
Appendix A 587

Table A.5 Derived SI units with special names used in Chap. 14

SI derived units
Expressed in terms of Expressed in terms of
Derived quantity Name Symbol other SI units SI base units
Activity (of a Becquerel Bq s-1
radionuclide)
Absorbed dose Gray Gy J/kg m2 s-2
Dose equivalent Sievert Sv J/kg m2 s-2

Table A.6 Examples of SI derived units and their expression in terms of SI base units
Derived quantity SI derived unit Expressed in terms of SI base units
Angular velocity rad s-1
Angular acceleration rad s-2
Dynamic viscosity Pa s m-1 kg s-1
Momentum Nm m2 kg s-2
Surface tension N m-1 kg s-2
Heat flow (flux) density W m-2 kg s-3
Radiation intensity W sr-1 m2 kg s-3 sr-1
Radiance W m-2 sr-1 kg s-3 sr-1
Heat capacity, entropy J K-1 m2 kg s-2 K-1
Specific heat capacity, specific entropy J kg-1 K-1 m2 s-2 K-1
Specific energy J kg-1 m2 s-2
Thermal conductivity W m-1 K-1 m kg s-3 K-1
Energy density J m-3 m-1 kg s-2
Electric field strength V m-1 m kg s-3 A-1
Electric charge density C m-3 m-3 s A
Electric flux density, C m-2 m-2 s A
Permittivity F m-1 m-3 kg-1 s4 A2
Permeability H m-1 m kg s-2 A-2
Molar energy J mol-1 m2 kg s-2 mol-1
Molar entropy, molar heat capacity J mol-1 K-1 m2 kg s-2 K-1 mol-1
Exposure (X and γ- rays) C kg-1 kg-1 s
Absorbed dose rate Gy s-1 m2 s-3
588 Appendix A

Table A.7 SI prefixes for Factor Name Symbol Factor Name Symbol
orders of magnitude
1024 yotta Y 10-1 deci d
1021 zetta Z 10-2 centi c
1018 exa E 10-3 milli m
1015 peta P 10-6 micro μ
1012 tera T 10-9 cano n
109 giga G 10-12 pico p
106 mega M 10-15 femto f
103 kilo k 10-18 atto a
102 hecto h 10-21 zepto z
10 deca da 10-24 Yocto y

Table A.8 Non-SI units accepted for use with the International System
Name Symbol Value in SI units
minute (time) min 1 min = 60 s
hour h 1 h = 60 min = 3600 s
day d 1 d = 24 h = 86,400 s
gon gon 1 gon = (1/200) rad
degree (plane angle) ° 1° = (π/180) rad
minute (plane angle) ´ 1´= (1/60)° = (π/10,800) rad
second (plane angle) ´´ 1´´= (1/60)°= (π/648,000) rad
liter l,L 1 l = 1 dm3 = 10-3 m3
metric ton t 1 t = 103 kg

Table A.9 Other non-SI units currently accepted for use with SI
Name Symbol Value in SI units
nautical mile n.m. 1 nautical mile = 1852 m
knots kn 1 nautical mile per hour = (1852/3600) m s-1
Ångstrøm Å 1 Å = 0.1 nm = 10-10 m
acre a 1 a = 1 dam2 = 102 m2
hectare ha 1 ha = 1 hm2
barn b 1 b = 10-28 m2
bar bar 1 bar = 0.1 MPa = 105 Pa
Curie Ci 1 Ci = 3.7 1010 Bq
Roentgen R 1 R = 2.58 10-4 C/kg
rad rad 1 rad = 1 cGy = 10-2 Gy
rem rem 1 rem = 1cSv = 10-2 Sv
Appendix A 589

Table A.10 Old English units and their conversion into SI units
Length Area
1 mile = 1.609 34 km 1 sq mile = 2.589 99 km2
1 furlong = 0.201 168 km 1 acre = 4046.86 m2
1 chain = 20.116 8 m 1 rood = 1011.71 m2
1 yd (yard) = 0.914 4 m 1 yd2 (square yard) = 0.836 127 m2
1 ft (foot) = 0.304 8 m 1ft2 (square foot) =0.092903 m2
1 in (inch) = 2.54 cm 1 in2 (square inch) =6.4516 cm2
1 nautical mile = 1.853 18 km
1 fathom = 1.828 8 m Mass
1 ton (short ton is 909.11kg) = 1016.05 kg
Volume 1 cwt (hundredweight) = 50.8023 kg
1 yd3 (cubic yard) = 0.764 55 m3 1 short hundredweight = 45.3592 kg
3
1 ft (cubic foot) = 28.316 8 dm3 1 quarter = 12.7006 kg
3
1 in (cubic inch) = 16.387 1 cm3 1 stone = 6.35029 kg
1 bu (bushel) = 36.368 7 dm3 1 lb (pound) = 0.45359237 kg
1 pk (peck) = 9.092 18 dm3 1 oz (ounce) = 28.3495 g
1 imperial gal (gallon) = 4.546 09 dm3 1 dr (dram) = 1.77185 g
1 US gal = 3.785 41 dm3 1 gr (grain) = 64.7989 mg
1 qt (quart UK) = 1.136 52 dm3 1 troy pound = 373.24172 g
1 pt (pint UK) = 0.568 261 dm3 1 oz tr (troy ounce) = 31.1035 g
1 gill = 0.142 065 dm3 1 drachm = 3.88793 g
1 fl oz = 28.413 1 cm3 1 scruple = 1.29598 g
1 fluid drachm = 3.551 63 cm3 1 dwt (pennyweight) = 1.55517 g
1 minim = 59.193 9 mm3 1 slug = 14.5939 kg
Mass per unit area Mass concentration
1 ton mile-2 = 392.298 kg/km2 1 gr/100 ft3 = 0.022883 5 g cm-3
= 3.922 98 kg/ha 1 oz/gal = 6.23602 g l-1
1 ton acre-1 = 0.251 071 kg/m2 1 gr/gal = 14.2538 mg l-1
= 2510.71 kg/ha
1cwt acre-1 = 0.012 553 5 kg/m2 Moment of inertia
= 125.535 kg/ha 1 lb ft2 = 0.0421401 kg m2
1 lb ft-2 = 4.882 43 kg/m2 1 lb in2 = 2.92640 kg cm2
-2
1 lb in = 70.307 0 g/cm2 1 oz in2 = 0.182900 kg cm2
-2
1 oz yd = 33.905 7 g/m2 1slug ft2 = 1.35582 kg m2
-2
1 oz ft = 305.152 g/cm2
Area per unit mass Angular moment of inertia
1 sq mile/ton = 2 549.08 m2 kg 1 lb ft/s = 0.138255 kg m s-1
1 yd2/ton = 0.822922 m2 t-1
Specific volume Moment
1 in3/ lb = 36.1273 cm3 kg-1 1 lb ft2/s = 0.042140 1 kg m2 s-1
3
1 ft /ton = 0.027869 6 dm3 kg-1 force
3
1 ft /lb = 62.4280 dm3 kg-1 1 tonf (ton-force) = 9 964.02 N
(continued)
590 Appendix A

Table A.10 (continued)

1 gal/lb = 10.0224 dm3 kg-1 1lbf = 4.44822 N
(pound-force)
Fuel consumption 1 ozf = 0.278014 N
1 ga mile-1 = 2.82481 l km-1 1pdl (poundal) = 0.138255 N
1 Usgal mile-1 = 2.35215 l km-1
1 mile gal-1 = 0.354006 km l-1 Angular
moment
1 mile USgal-1 = 0.425144 km l-1 1 lbf ft = 1.35582 N m
1 ton f ft = 3037.03 N m
Density 1 pdl ft = 0.0421401 N m
1 ton yd-3 = 1.32894 t m-3 1 ozf in = 0.706155 N cm
-3
1 lb ft = 16.0185 kg m-3 1 lbf in = 0.112985 N m
-3
1 lb in = 27.6799 g cm-3
-1
1 lb gal = 0.099776 3 kg l-1
-3
1 slug ft = 515.379 kg m-3
Pressure, stress Energy
1 tonf ft-2 = 107.252 kPa 1 therm = 105.506 MJ
1 tonf in-2 = 15.4443 MPa 1 hp h (horsepower hour) = 2.68452 MJ
1 lbf ft-2 = 47.8803 Pa 1 Btu (British thermal unit) = 1.05506 kJ
1 lbf in-2 = 6 894.76 Pa 1 ft lbf = 1.35582 J
1 ft H2O = 1.48816 Pa 1 ft pdl = 0.0421401 J
1 in H2O = 249.089 Pa
1 in Hg = 3386.39 Pa
Dynamic viscosity Power
1 pdl s ft-2 = 1.48816 Pa s 1 hp (horsepower) = 745.700W
1 lbf h ft-2 = 0.172369 MPa s 1 lf lbf s-1 = 1.35582 W
-2
1 lbf s ft = 47.8803 Pa s Heat flux
Heat flow 1 Btu ft-2 h-1 = 3.15459 W m-2
-1
1 Btu h = 0.293071 W
Specific energy Overall heat transfer
coefficient
1 Btu lb-1 = 2326 J kg-1 1 Btu ft-2 h-1 °F-1 = 5.678 26 W/m2 K
-1 -1
1 ft lbf lb = 2.989 07 J kg Thermal conductivity
Caloric value per unit volume 1 Btu ft ft-2 h-1 °F-1 = 1.73073 W m-1
K-1
1 therm gal-1 = 23.2080 MJ l-1 1 Btu in ft-2 h-1 °F-1 = 0.144228 W m-1
K-1
1 Btu ft-3 = 0.0372589 J cm-3 1 Btu in ft-2 s-1 °F-1 = 519.220 W m-1
K-1
Specific thermal resistance
1 ft2 h °F Btu-1 in-1 = 6.933 47 K m W-1
Specific heat capacity 1 ft2 h °F Btu-1 ft-1 = 57.778 9 K cm W-
1

(continued)
Appendix A 591

Table A.10 (continued)

1 Btu lb-1 °F-1 = 4 186.8J kg-1 K-1 Luminous intensity
1 ft lbf lb+-1 ° = 5.38032 J kg-1 K-1 1 lm ft-2 = 1 foot-candle = 10.763 9 lx
F-1
Specific entropy Luminance
1 Btu lb-1 °R-1 = 4 186.8 J kg-1 K-1 1 cd ft-2 = 10.7639 cd m-2
Heat capacity per unit volume 1 cd in2 = 1 550.00 cd m-2
1 Btu ft-3 °F-1 = 0.0670661 J cm- 1 foot-lambert = 3.426 26 cd m-2
3 -1
K
Pound:
The Troy system is the System of units used for precious metals and gems:
1 troy pound = 12 troy ounces = 240 pennyweight = 5760 grain
The Avoirdupois-system defines: 1 (Avoirdupois-) pound = 0.453 592 27 kg. For further special
units, refer [2]

The unit liter has the symbol l and alternative L in order to avoid the confusion
between the letter l and number 1. Sometimes the symbol ‘ is used to indicate a Liter.

A.2 Distribution Functions

To better understand how distributions are generated and how they work and can be
interpreted, some basic examples of velocity distributions are presented here. The
first one is very simple and can be used to understand the components of distribution
functions without any special mathematical background. Let us assume we have
obtained these data from a police speed check in front of our house on a given day:

v/km h-1
50
40
30
60
60
50
40
30
20
40
50
50
60
60
70
60
592 Appendix A

Example
The arithmetic mean of these N = 16 values is

N
1
v= vi = 48, 1 km h - 1
N i=1

The median of the distribution is the middle of an ordered listing of the data
(from lowest to highest value:

20 30 30 40 40 40 50 50 50 50 60 60 60 60 60 70

So it is apparently 50 km h-1

Now we can make categories of velocity ranges (in this example we will make
5 velocity categories), and put the data into these categories, as shown in Table A.11
Terms:
j
ΔN i
Distribution sum
Q= N = Q0 ðN i Þ = Q0,i Here: number distribution sum, index 0
i=1
k Sum of the fractions in the classes
N= ΔN i Here: the sum of numbers N, not, e.g., of mass
i=1
ΔN i
Distribution density
q0,i = N
Δvi Here: number distribution density
i Category index
j Index of quantity, here velocity
ΔN i
N = ΔQ0,i Number fraction in category i
Here: fraction of number

When Q is drawn versus the measured quantity (here the velocity) we get a
distribution curve, called the sum distribution curve (Fig. A.1). The value of Q is
between 0 and 1, meaning between 0% and 100%. The histogram bars in Fig. A.1
show us how many categories we have. The smooth sigmoid curve in Fig. A.2 what
the profile of the histogram bars would look like if we were able to make a very large

Table A.11 Sorting the measured velocities into 5 categories

ΔN i
i j vj/km h-1 Δvi/km h-1 vi /km h-1 ΔNi ΔQ0,i = N Q0,i q0,i/km h-1
1 0 0 21 10.5 1 0.0625 0.000 3.0 × 10-3
2 1 21 14 28.5 2 0.125 0.0625 8.3 × 10-3
3 2 35 17 43.5 7 0.4375 0.1875 25.7 × 10-3
4 3 52 10 57.0 5 0.3125 0.625 31.3 × 10-3
5 4 62 13 68.5 1 0.0625 0.9375 4.8 × 10-3
5 75 1.0
Appendix A 593

Fig. A.1 Sum distribution 1

curve showing accumulation
of sum fraction of speed
measurements in velocity
0.75
class categories as a function
of velocity

Q 0 0.5

0.25

10 20 30 40 50 60 70 80
–1
v /kmh

Fig. A.2 Velocity 40

distribution density curve
showing the number fraction
as a function of velocity in a 30
radar speed check.

q0 / 20
–1
kmh

10

10 20 30 40 50 60 70 80
v /kmh–1

number of categories (approaching an infinite number) with very large amounts of

measured data.
When q is drawn versus the measured quantity (here the velocity) we will obtain
another distribution curve, called the distribution density curve (see Fig. A.2). In this
example, it is a number distribution density curve. In mathematical terms q is the
derivative of Q. Taking this new histogram shown in Fig. A.2, we see that with only
a limited number of categories we have no smooth curve, but a “bell-shaped” step-
wise curve made up of steps going up and down because of the limited number of
categories. Again, the smooth “bell-shaped” curve is the one we would get if we
could have an infinite number of categories and measured data points.
This example was over a velocity distribution. We measured the velocity, put the
data into categories, and counted the number of occurrences. This resulted in a
number distribution function. This type of distribution function exercise could be
594 Appendix A

carried out with any other type of measurable quantity. For example, instead of
measuring velocities we could have measured the mass of particles in different size
categories to get a mass distribution, or measured length to get a length distribution,
or areas or volumes to get an area distribution function or volume distribution
function. Therefore, it is important to distinguish between what is the distributed
quantity (here velocity) and what set (grouping) we use to characterize an individual
belonging to a category. The set or grouping can be a number but it can also be
another type of set. Recall that when we do a particle size distribution, we perform
sieving experiments and then weigh the sample on each of the sieves. This results in
a mass distribution and not in a number distribution of the particle size (understand
that in the case of using sieves, we at least only have to weigh the sample and not
measure the actual size of the particles to get a particle size distribution).
When we talk about number distributions, then we use terms from probability
theory. In our speed check example, the expression ΔN N = ΔQ0,i gives the fraction of
i

cars found traveling in a velocity category i. For example, in category 4 (speed

between 52 and 62 km h-1) of our example we found five cars, so ΔQ0,4 = 16 5
=
0:3125 = 31:25%. If the data were reliable, such as would be the case for a much
greater number of measured data points (speed checks), then we say that the
probability P to meet a car traveling with a speed between 52 and 62 km h-1 is
about 31%.
So, we can see that probability is nothing more than

ΔN i
P= = ΔQ0,i
N
or

dN i
P= = dQ0,i
N
So when we have a number distribution, we can handle the Q0-v-function like a
probability function. The derivative function then is called a probability density
function q0.

dQ0
q0 = f ð v Þ =
dv
Perhaps the most famous probability density function related to velocity is the
Maxwell-Boltzmann function, which describes the velocity distribution of
molecules and atoms. Let us have a short look at this more complex example of a
velocity function f(v). It is also a number distribution, too. Let us start with some
general properties of this function, as listed in Table A.12.
When the distribution function f(v) is known, many properties of the set of
individuals can be obtained by mathematics, e.g.,
Appendix A 595

Table A.12 Properties of a velocity distribution f(v)

vi fraction of individuals in a category i, having a
f ðvÞ
dv = ΔQ0 velocity between v1 . . . v2. Probability to meet an
v1
individual having a speed between v1 and v2
vmax by condition of standardization: The fraction of
f ðvÞ
dv = 1
v min
individuals having a speed between vmin . . . vmax is
1, so the probability to meet an individual having a
speed between vmin . . . vmax is 100 %
1 is the integral mean of the velocity
f ðvÞ
v
dv = v
0
1 1 N is identical to the known arithmetic mean of the
f ðvÞ
v
dv = dN
N
v ffi N1 ΔN i vi velocity
0 0 i=1
1 is the integral mean of the velocity squared
f ðvÞ
v2
dv = v2
0

• the most probable velocity (modal value)

• the average velocity
• the average of the velocity squared
• the median
• the width (or range) of the distribution
• the standard deviation of the mean

An example of how this can be performed is shown here with the velocity
distribution of atoms or molecules in a gas. The function f(v) was first derived by
Maxwell and was proven later by Boltzmann. Today it is known as the Maxwell-
Boltzmann function defined as follows:
3 1mv2
m
e-
2 2
f ðvÞ = 4π
v2
kT
2πkT
Boltzmann factor

m mass of an atom in kg
k Boltzmann’s constant (1.380658 × 10-23 J K-1)
T temperature in K
v velocity of an atom in m s-1

The maximum of the curve provides us with the velocity which is most probable:
With

df ðvÞ
=0
dv
and the condition
596 Appendix A

mv2 df ðvÞ
e - 2kT = e - 1 ) =0
dv

we get

1 2
mv = kT
2
this is the same as

Ekin = kT

so the most probable velocity is

2kT
vw =
m
The mean velocity (arithmetic mean) can be obtained by forming the integral
mean. The result often is also called the average velocity

1 1
8kT 2
4 2kT
vd = f ðvÞ
v
dv = =

πm π m
0

so it is slightly different from the most probable velocity

4 2kT
vd =
= 1, 13
vw
π m
The integral mean of v2 is

1
3kT
v2 = f ðvÞ
v2
dv =
m
0

or

3kT
vm = v2 =
m
and
Appendix A 597

Table A.13 Characteristics of the Maxwell-Boltzmann velocity distribution function

Velocity Symbol Formula Comparison to vw Description
Most probable vw 2kT 1 vw Maximum of curve
m

Average vd 8 kT 1.128 vw Arithmetic mean

π m

Mean vm = v2 3kT 1.225 vw Mean v2

m

3
v2 =
v = 1, 225
vw
2 w

Now we may recognize vm = 3kT

m which is used in many applications,
such as in the basis of the ideal gas law,

p
V = m
RS
T

which is an outcome of the application of mathematics to the Maxwell-Boltzmann

distribution function. When we look again at this function (see below), we recognize
it is made up of a superposition of a square root function and an exponential
function. The last one sometimes is called the Boltzmann factor. At low temperatures
the Boltzmann factor is nearly 1, so we simplify to the square root function only. On
the other hand, at high temperatures the Boltzmann factor is the dominant part of the
function. In Table A.13 the resulting velocities are summarized.

A.3 Complex Numbers

Complex numbers are numbers which include a real part and an imaginary part.
Imaginary numbers are defined as a multiple of the imaginary unit i.
The imaginary unit i is the square root of -1:
p
i= -1

This definition of the imaginary unit has consequences like this:

i2 = - 1

i3 = - i

i4 = þ 1

Working with real numbers only there is no solution for the quadratic equation
x2 = - 1.
598 Appendix A

Working with imaginary numbers Zahlen we get two solution of the equation:
i and -i.
Because of (-i)2 = - 1, we get further:

i1 = i5 = i9 = i
ð - iÞ1 = ð - iÞ5 = ð - iÞ9 = - i
i2 = i6 = i10 = - 1
ð - iÞ2 = ð - iÞ6 = ð - iÞ10
i3 = i7 = i11 = - i
ð - iÞ3 = ð - iÞ7 = ð - iÞ11 = - i
i8 = i12 = 1
ð - iÞ8 = ð - iÞ12 = 1

Definition
The square root of -1 is called imaginary unit i. Multiples of i are imaginary
numbers.

The position of the imaginary part of the complex numbers in mathematics is

shown in Table A.14.
Complex numbers with the same real part but opposite imaginary part are called
conjugate complex numbers.
A complex number is usually expressed in a Cartesian form. The real part is
plotted on the abscissa (real axis) and the imaginary part (as a multiple of i) is plotted
on the ordinate (imaginary axis), as shown in Fig. A.3.
The representation of complex numbers in polar coordinates looks similar
(Fig. A.4) and is also called a trigonomical representation. Here real part a and
→
imaginary part b are components of vector r , which points to the complex number.
→
The corresponding angle φ to vector r (see Fig. A.4) can be calcluated by

Table A.14 System of numbers in mathematics, with examples

Real numbers Imaginary
Rational numbers Irrational numbers numbers
Whole Fractional Algebraic irrational Transcendent Multiples of i
p
numbers number numbers numbers i= -1
p p p
1; 2; 3; 1 1
3; 2 2; 3 π = 3.1416. . . - 4 = 2i
18;-18 0.25; -6.3 5 27 e = 2.718. . .
4

... ... ... sin2 = 0.9093. . .

Complex numbers: Numbers with real and imaginary parts, e.g. 4 + 3i
Appendix A 599

Fig. A.3 Graphical 4

representation of the complex
number 4 + 3i in a planar
diagram (two dimensions). 4 3i
3

imaginary 2
axis

1 2 3 4 5
real axis

Fig. A.4 Trigonomical

a (real part)
representation of a complex
number in polar coordinates

b (imaginary part)
imaginary
axis d ul)
mo
r(

real axis

a
cos φ =
r
resp.

a = r
cos φ

and

b
sin φ =
r
resp.

b = r
sin φ

So the complex number is

600 Appendix A

z = a þ i
b = r
ðcos φ þ i
sin φÞ

or

z = rðcos φ þ i
sin φÞ

Furthermore with

b
tan φ =
a
→
the modulus of r is

r= a 2 þ b2

Example
The complex number 3 + 4i has the Cartesian coordinates a = 3 and b = 4, that
means

j zj = a2 þ b 2 = 3 2 þ 42 = 5

In the trigonometric form this is

cos φ = 5 = 0:6
3 φ = 53.13°
sin φ = 5 = 0:8
4 φ = 53.13°

vice versa it is

z = rðcos φ þ i
sin φÞ = 5
ðcos 53:13 ° þ i
sin 53:13 ° Þ = 5

ð0:6 þ i
0:8Þ = 3 þ 4i

A very simple expression of complex numbers is to use the relations from Euler.
Euler’s formula are:

eix = cos x þ i
sin x

e - ix = cos x - i
sin x

With φ from the trigonomical representation (Fig. A.4) we have:

eiφ = cos φ þ i
sin φ

Appendix A 601

e - iφ = cos φ - i
sin φ

With Euler’s number, the complex number z can write as:

z = jzj
eiφ

i.e.,

z = jzj
ðcos φ þ i
sin φÞ

so

z = jzj
cos φ þ i
jzj
sin φ

real part imaginary part

The modulus r of the trigonomical representation in (Fig. A.4) is the same as the
amount of the complex number z.

r = j zj

Definition
The Euler’s number e is the base of the natural logarithm.

The advantage of describing a complex number with the Euler’s formula is its
simplicity. There only the quantity ϕ is necessary to express the complex number,
see the right column in Table A.15.

Table A.15 Possibilities for expression of the complex number 3 + 4i

Algebraic form Trigonometric form Exponential form
Real part Imaginary part Angle φ in ° Angle φ in rad Euler Euler
3 4 53.13° 0.927 z = ei
53.13° z = ei
0.927
602 Appendix A

Example
z = 3 þ 4i
z=a þ b
i
z = jzj
cos φ þ i
jzj
sin φ
3 4
b 4
tan φ = =
a 3
φ = 53:13 °

z = eiφ
z = ei
53:13 °
p
jzj = a2 þ b2 = 9 þ 16
j zj = 5

In words, we can say:

imaginary part
tan φ =
real part

jzj = ðreal partÞ2 þ ðimaginary partÞ2

that means that we can estimate from the magnitude of φ or tanφ ratio the ratio of real
part and imaginary part of the complex number. In Table A.16 some exampls are
given.

Table A.16 Several phase angles and their interpretation, examples

φ/° tanφ Description
0 0 Real part >> imaginary part
(e.g., when imaginary part = 0)
p
30 1
3 Real part > imaginary part
3
45 1 Real part = imaginary part
p
60 3 Real part < imaginary part
90 1 Real part >> imaginary part
(e.g., when real part is nearly zero)
135 -1 Imaginary part = - (real part)
180 0 See φ = 0°
Appendix A 603

A.3.1 Complex Physical Quantities

Physical quantities can also be complex quantities which consist of a real part and an
imaginary part. Such a physical quantity is called an apparent quantity, which
consists of an effective component (real part) and a loss component (imaginary
part). Figure A.5 illustrates a complex physical quantity with the help of these terms.

Remark
If we calculate "real" (i.e., without using imaginary numbers), we implicitly
assume that the imaginary part of the complex quantity is zero and thus the
apparent quantity is identical to the real quantity.

To help us work with some of the complex properties used in this book, some
examples are presented here for the purpose of illustration. We should also recognize
that different scientific disciplines or fields of engineering may use their own special
terms for the real and imaginary parts, but the mathematics are the same.

A.3.2 Electrical Properties

In electric circuits with alternating current, we have real resistances and imaginary
resistances, called active resistance and reactive resistance. The apparent quantity—
the apparent resistance—is called the impedance, as shown in Table A.17 and A.18.

Z, R resistance in Ω
Y, G admittance in S
C capacitance in F
L inductance in H
ω angular frequency in s-1

Fig. A.5 Trigonomical effective component

representation of a physical
quantity
loss component

y
titi
imaginary uan
q
axis nt
are
p
ap

real axis
604 Appendix A

Table A.17 Terms of complex electric resistance

Impedance = Effective resistance + Reactive resistance
Z = R + i
ωL - 1
ωC

Table A.18 Terms of complex electric conductance

Admittance = Admittance + Susceptance
Y = G + i
ωC - 1
ωL

There are cases, in which the reactive resistance is zero, e.g. if the capacitance is
zero or the angular frequency is ω = 0. The latter is the case in direct current circuits.
We see that all observed resistances are also active resistances, such as ohmic
resistances, and we do not have to deal with complex quantities. On the other
hand in alternating fields, impedance and resistance is not the same, and we have
to take into account both real and imaginary parts of the complex quantities. In
Chap. 11 we learned about the complex permittivity.

A.3.3 Rheology

By means of mechanic oscillation, it is possible to differentiate between elastic and

non-elastic (viscous) properties of a material. The observed behavior can be
described as a complex quantity (e.g., complex modulus), which includes both
properties, elastic and viscous. Then the elastic properties are represented by the
real part and the inelastic (viscous) properties by the imaginary part. In this sense the
non-elastic, viscous behavior of the sample is understood as the loss part. A behavior
like this can be easily described by the loss angle φ or by the loss tangent, tanφ. It is
important to avoid confusion in this regard. For describing rheological behavior, it is
also possible to use the apparent viscosity as a complex property which consists of
real (viscous) parts and imaginary (elastic) parts. Also sometimes it is useful to use
the complex compliance instead of the complex elasticity. In Chap. 4 we learned
about complex compliance and complex viscosity. When further study of complex
rheological properties is needed, it may be helpful to remember to refer to this
appendix section once again.

A.3.4 Heat Flow Calorimetry

Oscillation tests in Thermal Analysis are referred to as temperature-modulated tests

(MDSC). These tests result in a complex heat flow which is composed of a real part
and an imaginary part. The real part is caused by the heat capacity of the sample; it is
a reversible heat flow. The imaginary part is a non-reversible heat flow, in other
words the loss component. It may be caused by chemical reactions in the sample. In
Appendix A 605

an analogous way a complex heat capacity can be defined. In Chap. 8 we learned

about complex heat flow and complex heat capacity.

A.4 Greek Letters (Fig. A.6)

To exercise the notation of Greek letters Fig. A.7 will be helpful:

Fig. A.6 Greek block letters A α a Alpha

and their names
B β b Beta
Г γ g Gamma
Δ δ d Delta
E ε e Epsilon
Z ζ z Zeta
H η e Eta
Θ T,- th Theta
I ι j Jota
K κ k Kappa
Λ λ l Lambda
M μ m Mu
N ν n Nu
Ξ ξ x Xi
O ο o Omikron
Π π p Pi
P ρ r Rho
Σ σ s Sigma
T τ t Tau
Y υ y Ypsilon
Ф φ ph Phi
X χ ch Chi
Ψ ψ ps Psi
Ώ ω o Omega
606 Appendix A

Fig. A.7 Greek script

A.5 Conversion Chart: Temperature (Table A.19)

Table A.19 conversion of temperatures

Given Required quantity in:
quantity in: °C °F °R K
°C 1 ϑ/°F=9/5
ϑ/ ϑ/°R=9/5
ϑ/°C T/K=ϑ/°C
°C+32 +491.67 +273.15
°F ϑ/°C=5/9
(ϑ/ 1 ϑ/°R=ϑ/°F T/K=(5/9
(ϑ/°F-
°F-32) +459.67 32))+273.15
°R ϑ/°C=5/9
(ϑ/ ϑ/°F= ϑ/° 1 T/K=5/9
(ϑ/°
°R-491.67) R-459.67 R-491.67)+273.15
°K ϑ/°C=T/ ϑ/°F=9/5
T/ ϑ/°R=9/5
(T/ 1
K-273.15 K-459.67 K-273.15)
+491.67
°R degree Rankine, °C degree Celsius, °F degree Fahrenheit, K Kelvin
Appendix A 607

A.6 Sugar Conversion Chart: Concentration, Density, Refraction

(Table A.20)

Table A.20 conversion table for sugar content, i.e. solid matter (degree Brix, degree Oechsle,
degree Baumé, Klosterneuburg degrees)
Degree Degree
Sucrose concentration Oechsle Baumé Klosterneuburg
°Bx Relative density d20/20 °Oe °Be degrees
1 1.00389 4 0.56 1.3
2 1.00779 8 1.12 2.0
3 1.01172 12 1.68 2.8
4 1.01567 16 2.24 3.6
5 1.01965 20 2.79 4.4
6 1.02366 24 3.35 5.2
7 1.02770 28 3.91 5.9
8 1.03176 32 4.46 6.9
9 1.03586 36 5.02 7.7
10 1.03998 40 5.57 8.5
11 1.04413 44 6.13 9.3
12 1.04831 48 6.68 10.1
13 1.05252 53 7.24 11.0
14 1.05677 57 7.79 11.9
15 1.06104 61 8.34 12.7
16 1.06534 65 8.89 14.4
17 1.06968 70 9.45 14.3
18 1.07404 74 10.0 15.2
19 1.07844 78 10.55 16.0
20 1.08287 83 11.10 16.9
21 1.08733 87 11.65 17.7
22 1.09183 92 12.20 18.6
23 1.09636 96 12.74 19.5
24 1.10092 101 13.29 20.3
25 1.10551 106 13.84 20.9
26 1.11014 110 14.39 21.9
27 1.11480 115 14.93 22.8
28 1.11949 119 15.48 23.6
29 1.12422 124 16.02 24.4
30 1.12898 129 16.57 25.3
31 1.13378 134 17.11 26.1
32 1.13861 139 17.65 26.8
33 1.14347 143 18.19 27.8
34 1.14837 148 18.73 28.7
35 1.15331 153 19.28 -
(continued)
608 Appendix A

Table A.20 (continued)

Degree Degree
Sucrose concentration Oechsle Baumé Klosterneuburg
°Bx Relative density d20/20 °Oe °Be degrees
36 1.15828 158 19.81 -
37 1.16329 163 20.35 -
38 1.16833 168 20.89 -
39 1.17341 173 21.43 -
40 1.17853 179 21.97 -

A.7 Material Data

Here are some helpful material data. For further physical data of foods, see [3–10]
(Table A.21).

Table A.21 Properties of water: Density ρ, dynamic viscosity η, refraction index at 589 nm nD,
permittivity ε, saturation steam pressure pv, enthalpy of evaporation Δhvap, and surface tension σ at
temperature ϑ
ϑ ρ η pv σ
°C kg m-3 mPa s nD ε kPa mN m-1
0 999.9 1.787 1.3346 87.9 0.610 75.6
5 1000.0 1.519 1.3346 85.9 0.872 74.8
10 999.7 1.307 1.3343 84.0 1.228 74.1
15 999.1 1.139 1.3338 82.1 1.705 73.3
20 998.2 1.002 1.3333 80.2 2.34 72.6
25 997.1 0.890 1.3329 78.4 3.17 71.8
30 995.7 0.798 1.3323 76.6 4.24 71.0
40 992.2 0.653 1.3309 73.2 7.38 69.4
50 988.1 0.547 1.3293 69.9 12.33 67.7
60 983.2 0.466 1.3275 66.8 19.92 66.0
70 977.8 0.404 1.3255 63.8 31.16 64.3
80 971.8 0.355 1.3231 60.9 47.34 62.5
90 965.3 0.315 1.3209 58.2 70.1 60.7
100 958.4 0.282 1.3182 55.6 101.3 58.9
Appendix A 609

A.7.1 Density (Table A.22)

Table A.22 Polynomic equations to calculate the density [5]

Water ρ = 1000.22 + 1.0205 × 10-2ϑ - 5.8149 × 10- in kg.m-3
ϑ + 1.496 × 10-5ϑ3
3 2

Whey ρ = ρWasser + 4.039
dm + 1.273 × 10-2dm2 + 9.62 × 10-5dm3 in kg.m-3

UHT milk ρ = 1040.7 - 0.2665
ϑ - 2.3 × 10-3ϑ2 in kg.m-3
-3 2 -1
Milk and ρ = 2.3 × 10 ϑ - 2.665 × 10 ϑ + 1040.7 - xf(-4.81
in kg.m-3
-5 2 -3
cream 10 ϑ + 9.76 × 10 ϑ + 1.011)
dm dry matter in % (m/m)
ρ density in kg.m-3
ϑ temperature in °C
xf fat content in % (m/m)

Viscosity of water [11]

484:3726
η=mPa
s = 0:0318
eϑþ120:2202

Interfacial Data (Tables A.23, A.24, A.25, A.26, A.27, A.28, A.29, and A.30)
Table A.23 Surface tension data of milk [5]
Surface tension σ / N m-1
Temperature in °C Skim milk Whole milk
0 0.0557 0.0515
10 0.0536 0.0507
20 0.0515 0.0500
30 0.0497 0.0493
40 0.0479 0.0488
50 0.0462 0.0483
60 0.0446 0.0478
70 0.0432 0.0474
80 0.0418 0.0471
90 0.0406 0.0469
100 0.0395 0.0467
610 Appendix A

Table A.24 Surface tension σ of food at 20°C, examples from [5]

Material σ / N m-1
Cream (30 %) 0.0492
Whey (TS 5 %) 0.0566
Egg white 0.060
Skim milk permeate(0.1 % protein) 0.069
Skim milk retentate(5 % protein) 0.0513
Butter milk (bm) 0.0489
bm permeate 0.0630
bm retentate 0.0485
Butter fat 99.9 %, 30 °C 0.0318

Table A.25 Heat conductivity of frozen food, examples from [12]

λ/W K-1 m-1
Product Fresh Frozen
Strawberry 0.49 1.12
Fruit juices 0.56 2.10
Mashed potatoes 0.49 1.10
Cucumbers 0.54 1.26
Salmon 0.50 1.17
Filet of codfish 0.54 1.20
Beef 0.48 1.40
Fat pork 0.37 0.72
Low fat pork 0.50 1.55
Bacon 0.19 0.27

Table A.26 Heat conductivity of food, examples from [12]

Food λ/W K-1 m-1 ϑ /°C
Milk 0.55 20
Mashed banana 0.56 16
Butter (17% H2O) 0.2 20
Codfish 0.44 0
Codfish 1.22 -10
Codfish 1.37 -20
Vegetable 0.3 × 0.6 20
Grain (loose) 0.15 20
Potatoes 0.55 20
Margarine 0.2 20
Fruit 0.35 × 0.55 20
Beef (78.5 % H2O) 0.48 0
Beef (78.5 % H2O) 1.06 -5
Beef (78.5 % H2O) 1.35 -10
Beef (78.5 % H2O) 1.57 -20
Beef fat (7% H2O) 0.2 0
(continued)
Appendix A 611

Table A.26 (continued)

Food λ/W K-1 m-1 ϑ /°C
Beef fat (7% H2O) 0.21 -5
Beef fat (7% H2O) 0.23 -10
Beef fat (7% H2O) 0.26 -20
Pork (76.8 % H2O) 0.2 0
Pork (76.8 % H2O) 0.8 -5
Pork (76.8 % H2O) 0.99 -10
Pork (76.8 % H2O) 1.29 -20
Pork fat 0.19 0
Pork fat 0.29 -20
Starch (loose) 0.15 20
Sodium chloride (solid) 7 20
Sugar (solid) 0.6 20
Sugar (loose) 0.15 . . . 0.35 20
Ricinus oil 0.181 20
Water 0.596 20
Air 0.026 20

Table A.27 Thermal diffusivity of food at 20°C, examples

Food a 106/m2 s-1
Ethanol 0.100
Beer 0.134
Butter 0.086
Fat 0.09
Fish 0.147
Low fat meat 0.15
Vegetable 0.14
Grain (loose) 0.095
Potatoes 0.156
Margarine 0.092
Oranges 0.150
Olive oil 0.115
Mineral salt 3.2
Sugar (loose) 0.2

Table A.28 Heat conductivity λ and thermal diffusivity a of some materials at ambient
temperature
Material λ/W K-1 m-1 a × 106/m2 s-1
Silver 407 176
Copper 384 107
Aluminum 220 94.6
Glass 0.1 × 1.0
Special steel 8 × 16
Bronze 62 18.6
(continued)
612 Appendix A

Table A.28 (continued)

Material λ/W K-1 m-1 a × 106/m2 s-1
Iron 74 16.2
Glass 1 0.61
Quartz glass 1.36
Graphite 169
Constantan 23
Brass 80 × 220
Steel (St37) 45 14.7
V2A-steel 15
Titanium 22
PS 0.15
PVC 0.16
EPS (styrofoam) 0.036
Plexiglass 0.19
PA 0.26
PE 0.4
Concrete 1.3 0.66
Porous concrete 0.3
Earth 0.4 × 1.3
Window glass 0.8 0.6
Gypsum 0.5 0.47
Chalky sandstone 1.1
Marble 2.8
Porcelain 0.9
Sandstone 1.9 1.0 × 1.3
Chamotte 0.5 × 1.2 0.33 × 0.69
Plaster 0.8
Brick 0.5 0.27
Brick wall 0.8 0.55
Asphalt 0.7
Roofing paper 0.2
Oak/beech 0.17 × 0.31 0.11
Fir 0.14 × 0.26
Gum 0.13 - 0.24 0.1
Cork 0.05
Granulated cork 0.036
Mineral faber 0.04
Plant faber 0.05
Asbestos sheet 0.7
Slate 2.1
Appendix A 613

Table A.29 IR emissivity ε of some materials [13]

Material ε
Ideal black body 1
Aluminum, raw 0.06 – 0.07
Aluminum, polished 0.05
Aluminum, oxidized 0.11
Aluminum, anodized 0.7
Brick 0.9
Chromium polished 0.1
Copper, oxidized 0.6 – 0.7
Copper, polished 0.02
Leaves, green 0.88
Ice 0.96
Silicon rubber 0.94
Human skin 0.93 – 0.96
Snow 0.85
Water 0.96
White paper 0.92
Wood 0.93
Table A.30 Refraction index (20 °C, 589 nm) of aqueous sucrose-solution. The concentration in °Bx is noted in left column [14]
614

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

0 1.33299 1.33313 1.33327 1.33342 1.33356 1.33370 1.33385 1.33399 1.33413 1.33428
1 1.33442 1.33456 1.33471 1.33485 1.33500 1.33514 1.33529 1.33543 1.33558 1.33572
2 1.33587 1.33601 1.33616 1.33630 1.33645 1.33659 1.33674 1.33688 1.33703 1.33717
3 1.33732 1.33747 1.33761 1.33776 1.33791 1.33805 1.33820 1.33835 1.33849 1.33864
4 1.33879 1.33893 1.33908 1.33923 1.33938 1.33952 1.33967 1.33982 1.33997 1.34012
5 1.34026 1.34041 1.34056 1.34071 1.34086 1.34101 1.34116 1.34131 1.34146 1.34160
6 1.34175 1.34190 1.34205 1.34220 1.34235 1.34250 1.34265 1.34280 1.34295 1.34310
7 1.34325 1.34341 1.34356 1.34371 1.34386 1.34401 1.34416 1.34431 1.34446 1.34461
8 1.34477 1.34492 1.34507 1.34522 1.34537 1.34553 1.34568 1.34583 1.34598 1.34614
9 1.34629 1.34644 1.34660 1.34675 1.34690 1.34706 1.34721 1.34736 1.34752 1.34767
10 1.34783 1.34798 1.34813 1.34829 1.34844 1.34860 1.34875 1.34891 1.34906 1.34922
11 1.34937 1.34953 1.34968 1.34984 1.34999 1.35051 1.35031 1.35046 1.35062 1.35077
12 1.35093 1.35109 1.35124 1.35140 1.35156 1.35172 1.35187 1.35203 1.35219 1.35234
13 1.35250 1.35266 1.35282 1.35298 1.35313 1.35329 1.35345 1.35361 1.35377 1.35393
14 1.35409 1.35424 1.35440 1.35456 1.35472 1.35488 1.35504 1.35520 1.35536 1.35552
15 1.35568 1.35584 1.35600 1.35616 1.35632 1.35648 1.35664 1.35680 1.35697 1.35713
16 1.35729 1.35745 1.35761 1.35777 1.35793 1.35810 1.35826 1.35842 1.35858 1.35875
17 1.35891 1.35907 1.35923 1.35940 1.35956 1.35972 1.35989 1.36005 1.36021 1.36038
18 1.36054 1.36070 1.36087 1.36103 1.36120 1.36136 1.36153 1.36169 1.36186 1.36202
19 1.36219 1.36235 1.36252 1.36268 1.36285 1.36301 1.36318 1.36334 1.36351 1.36368
20 1.36384 1.36401 1.36418 1.36434 1.36451 1.36468 1.36484 1.36501 1.36518 1.36535
21 1.36551 1.36568 1.36585 1.36602 1.36619 1.36635 1.36652 1.36669 1.36686 1.36703
22 1.36720 1.36737 1.36754 1.36771 1.36788 1.36804 1.36821 1.36838 1.36855 1.36872
23 1.36889 1.36907 1.36924 1.36941 1.36958 1.36975 1.36992 1.37009 1.37026 1.37043
24 1.37060 1.37078 1.37095 1.37112 1.37129 1.37147 1.37164 1.37181 1.37198 1.37216
Appendix A
25 1.37233 1.37250 1.37267 1.37285 1.37302 1.37320 1.37337 1.37354 1.37372 1.37389
26 1.37407 1.37424 1.37441 1.37459 1.37476 1.47494 1.37511 1.37529 1.37546 1.37564
Appendix A

27 1.37582 1.37599 1.37617 1.37634 1.37652 1.37670 1.37687 1.37705 1.37723 1.37740
28 1.37758 1.37776 1.37793 1.37811 1.37829 1.37847 1.37865 1.37882 1.37900 1.37918
29 1.37936 1.37954 1.37972 1.37989 1.38007 1.38025 1.38043 1.38061 1.38079 1.38097
30 1.38115 1.38133 1.38151 1.38169 1.38187 1.38205 1.38223 1.38241 1.38259 1.38277
31 1.38296 1.38314 1.38332 1.38350 1.38368 1.38386 1.38405 1.38423 1.38441 1.38459
32 1.38478 1.38496 1.38514 1.38532 1.38551 1.38569 1.38588 1.38606 1.38524 1.38643
33 1.38661 1.38679 1.38698 1.38716 1.38735 1.38753 1.38772 1.38790 1.38809 1.38827
34 1.38846 1.38865 1.38883 1.38902 1.38920 1.38939 1.38958 1.38976 1.38995 1.39014
35 1.39032 1.39051 1.39070 1.39088 1.39107 1.39126 1.39145 1.39164 1.39182 1.39201
36 1.39220 1.39239 1.39258 1.39277 1.39277 1.39135 1.39333 1.39352 1.39371 1.39390
37 1.39409 1.39428 1.39447 1.39466 1.39485 1.39505 1.39524 1.39543 1.39562 1.39581
38 1.39600 1.39619 1.39638 1.39658 1.39677 1.39696 1.39715 1.39734 1.39754 1.39773
39 1.39792 1.39812 1.39831 1.39850 1.39870 1.39889 1.39908 1.39928 1.39947 1.39967
40 1.39986 1.40006 1.40025 1.40044 1.40064 1.40084 1.40103 1.40123 1.40142 1.40162
41 1.40181 1.40201 1.40221 1.40240 1.40260 1.40280 1.40299 1.40319 1.40339 1.40358
42 1.40378 1.40398 1.40418 1.40437 1.40457 1.40477 1.40497 1.40517 1.40537 1.40557
43 1.40576 1.40596 1.40616 1.40636 1.40656 1.40676 1.40696 1.40716 1.40736 1.40756
44 1.40776 1.40796 1.40817 1.40837 1.40857 1.40877 1.40897 1.40917 1.40937 1.40958
45 1.40978 1.40998 1.41018 1.41039 1.41059 1.41079 1.41099 1.41120 1.41140 1.41160
46 1.41181 1.41201 1.41222 1.41242 1.41262 1.41283 1.41303 1.41324 1.41344 1.41365
47 1.41385 1.41406 1.41427 1.41447 1.41468 1.41488 1.41509 1.41530 1.41550 1.41571
48 1.41592 1.41612 1.41633 1.41654 1.41675 1.41695 1.41716 1.41737 1.41758 1.41779
49 1.41799 1.41820 1.41841 1.41862 1.41883 1.41904 1.41925 1.41946 1.41967 1.41988
50 1.42009 1.42030 1.42051 1.42072 1.42093 1.42114 1.42135 1.42156 1.42177 1.42199
51 1.42220 1.42241 1.42262 1.42283 1.42305 1.42326 1.42347 1.42368 1.42390 1.42411
615

(continued)
Table A.30 (continued)
616

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

52 1.42432 1.42454 1.42475 1.42497 1.42518 1.42539 1.42561 1.42582 1.42604 1.42625
53 1.42647 1.42668 1.42690 1.42711 1.42733 1.42754 1.42776 1.42798 1.42819 1.42841
54 1.42863 1.42884 1.42906 1.42928 1.42949 1.42971 1.42993 1.43015 1.43036 1.43058
55 1.43080 1.43102 1.43124 1.43146 1.43168 1.43190 1.43211 1.43233 1.43255 1.43277
56 1.43299 1.43321 1.43343 1.43365 1.43387 1.43410 1.43432 1.43454 1.43476 1.43498
57 1.43520 1.43542 1.43565 1.43587 1.43609 1.43631 1.43654 1.43676 1.43698 1.43720
58 1.43743 1.43765 1.43787 1.43810 1.43832 1.43855 1.43877 1.43900 1.43922 1.43944
59 1.43967 1.43989 1.44012 1.44035 1.44057 1.44080 1.44102 1.44125 1.44148 1.44170
60 1.44193 1.44216 1.44238 1.44261 1.44284 1.44306 1.44329 1.44352 1.44375 1.44398
61 1.44420 1.44443 1.44466 1.44489 1.44512 1.44535 1.44558 1.44581 1.44604 1.44627
62 1.44650 1.44673 1.44696 1.44719 1.44742 1.44765 1.44788 1.44811 1.44834 1.44858
63 1.44881 1.44904 1.44927 1.44950 1.44974 1.44997 1.45020 1.45043 1.45067 1.45090
64 1.45113 1.45137 1.45160 1.45184 1.45207 1.45230 1.45254 1.45277 1.45301 1.45324
65 1.45348 1.45371 1.45395 1.45419 1.45442 1.45466 1.45489 1.45513 1.45537 1.45560
66 1.45584 1.45608 1.45631 1.45655 1.45679 1.45703 1.45726 1.45750 1.45774 1.45798
67 1.45822 1.45846 1.45870 1.45893 1.45917 1.45941 1.45965 1.45989 1.46013 1.46037
68 1.46061 1.46085 1.46109 1.46134 1.46158 1.46182 1.46206 1.46230 1.46254 1.46278
69 1.46303 1.46327 1.46351 1.46375 1.46400 1.46424 1.46448 1.46473 1.46497 1.46521
70 1.46546 1.46570 1.46594 1.46619 1.46643 1.46668 1.46692 1.46717 1.46741 1.46766
71 1.46790 1.46815 1.46840 1.46864 1.46889 1.46913 1.46938 1.46963 1.46987 1.47012
72 1.47037 1.47062 1.47086 1.47111 1.47136 1.47161 1.47186 1.47210 1.47235 1.47260
73 1.47285 1.47310 1.47335 1.47360 1.47385 1.47410 1.47435 1.47460 1.47485. 1.47510
74 1.47535 1.47560 1.47585 1.47610 1.47653 1.47661 1.47686 1.47711 1.47736 1.47761
75 1.47787 1.47812 1.47837 1.47862 1.47888 1.47913 1.47938 1.47964 1.47989 1.48015
76 1.48040 1.48065 1.48091 1.48116 1.48142 1.48167 1.48193 1.48218 1.48244 1.48270
Appendix A
77 1.48295 1.48321 1.48346 1.48372 1.48398 1.48423 1.48449 1.48475 1.48501 1.48526
78 1.48552 1.48578 1.48604 1.48629 1.48655 1.48681 1.48707 1.48733 1.48759 1.48785
Appendix A

79 1.48811 1.48837 1.48863 1.48889 1.48915 1.48941 1.48967 1.48993 1.49019 1.49045
80 1.49071 1.49097 1.49123 1.49149 1.49175 1.49202 1.49228 1.49254 1.49280 1.49307
81 1.49333 1.49359 1.49386 1.49412 1.49438 1.49465 1.49491 1.49517 1.49544 1.49570
82 1.49597 1.49623 1.49650 1.49676 1.49703 1.49729 1.49756 1.49782 1.49809 1.49835
83 1.49862 1.49889 1.49915 1.49942 1.49969 1.49995 1.50022 1.50049 1.50076 1.50102
84 1.50129 1.50156 1.50183 1.50210 1.50237 1.50263 1.50290 1.50317 1.50344 1.50371
85 1.50398 1.50425 1.50452 1.50479 1.50506 1.50533 1.50560 1.50587 1.50614 1.50641
617
618 Appendix A

A.8 Color Test Solutions

In conformity with the European Pharmacopoeia [15], the five so-called color
standard solutions can be prepared from mixtures of three color stock solutions
(yellow: FeCl3-solution; red: CoCl2-solution; blue: CuSO4-solution). The five
resulting colors are brown, brownish-yellow, yellow, greenish-yellow, and red.
Color test solutions can then be prepared from defined dilution levels carried out
in six steps. The color stock solutions are very stable and keep their quality indefi-
nitely. However, the color standard solutions and the color test solutions are not
stable and must be freshly prepared at the time of use. There are two established
methods by which to compare a sample color with a color test solution that are
described as follows:

Method I: 2 cm3 of each, sample and test solution, are filled in colorless reagent bottles with
12 mm inside diameter and observed under diffuse daylight in a horizontal
orientation to a white background.
Method II: 10 cm3 of each, sample and test solution, are filled in colorless reagent bottles with
16 mm inside diameter, and observed under diffuse daylight in a vertical orientation
to a white background.

The advantage of method II is that results are more reproducible because of the
greater depth of the solutions. It is advisable to use a so-called Nessler-cylinder
instead of the reagent bottles. These are translucent, colorless cylinders with 16 mm
inside diameter and flat, translucent bottom.

A.8.1 Preparation of Color Stock Solutions

Yellow color stock solution:

A quantity of 46 g ferric chloride (FeCl3) is dissolved in ca. 900 ml of a mixture of
25 ml hydrochloric acid and 975 ml water and diluted with this mixture to 1000.0
ml. After analysis to confirm solution composition, the solution is further diluted
with sufficient hydrochloric acid-water mixture until there are 45.0 mg FeCl3
6H2O
in 1 ml of the solution.
Analysis of content
A quantity of 10.0 ml of yellow color stock solution and 15 ml water, 5 ml
hydrochloric acid and 4 g potassium iodide are mixed in a 200-ml Erlenmeyer flask
with glass stopper. Immediately, the flask is closed and placed under darkness for
about 15 min. After the addition of 100 ml water the resulting iodine solution will be
titrated with 0.1 N-sodium thiosulfate solution. At the end of the titration, 10 drops
starch-solution has to be added. 1 ml 0.1 N-sodium thiosulfate solution conforms to
27.03 mg FeCl3
6H2O.
Red color stock solution:
A quantity of 60 g cobalt(II)-chloride (CoCl2) is dissolved in ca. 900 ml of a
mixture of 25 ml hydrochloric acid and 975 ml water and diluted with this mixture to
Appendix A 619

a total of 1000.0 ml. After the analysis of contents, the solution will be diluted with
sufficient hydrochloric acid-water-mixture, until there is 59.5 mg CoCl2
6H2O in
1 ml of the solution.
Analysis of content
A quantity of 5.0 ml of red color stock solution and 5 ml hydrogen-peroxide and
10 ml of a 30 percent (m/V) solution of sodium hydroxide are mixed in a 200-ml
Erlenmeyer flask with glass stopper, and heated to low boiling for about 10 min.
After cooling down to room temperature, the solution is mixed with diluted sulfuric
acid and 2 g potassium iodide, and the flask is immediately closed and shaken slowly
to dissolve the iodide. The resulting brownish colored iodine solution is titrated with
0.1 N-sodium thiosulfate solution until the color is pink. Near to the end of the
titration, 10 drops of starch-solution R are added. 1 ml 0.1 N-sodium thiosulfate
solution conforms 23.79 mg CoCl2
6H2O.
Blue color stock solutions:
A quantity of 63 g cupric sulfate (CuSO4) is dissolved in ca. 900 ml of a mixture
of 25 ml hydrochloric acid and 975 ml water and diluted with this mixture up to
1000.0 ml. After the analysis of contents, the solution is diluted with sufficient
hydrochloric acid-water-mixture, until there is 62.4 mg CuSO45H2O in 1 ml of the
solution.
Analysis of contents:
A quantity of 10.0 ml blue color stock reference solution and 50 ml water, 12 ml
diluted acetic acid, and 3 g potassium iodide are mixed in a 200-ml Erlenmeyer flask
with glass stopper. The resulting iodine solution is titrated with 0.1 N-sodium
thiosulfate solution until the solution is blue colored. Near to the end of the titration,
10 drops of starch-solution R are added. 1 ml 0.1 N-sodium thiosulfate solution
conforms 24.979 mg CuSO4
5H2O.

A.8.2 Color Standard Solutions

The nomenclature for the 5 color standard solutions and the color test solutions are
given in Table A.31.

Table A.31 Color standard solutions [15]

Yellow color Red color Blue color Hydrochloric
Color standard stock solutions stock solutions stock solutions acid 1% (m/V)
solutions ml ml ml ml
B (brown) 3.0 3.0 2.4 1.6
BG (brownish- 2.4 1.0 0.4 6.2
yellow)
G (yellow) 2.4 0.6 0.0 7.0
GG (greenish- 9.6 0.2 0.2 0.0
yellow)
R (red) 1.0 2.0 0.0 7.0
620 Appendix A

With these 5 color standard solutions, the color test solutions are mixed like
shown in Table A.32

Table A.32 Color test solutions [15]

Color test solutions B
Color standard
Color test solution solution B in ml Hydrochloric acid 1 % (m/V) in ml
B1 1.50 0.50
B2 1.00 1.00
B3 0.75 1.25
B4 0.50 1.50
B5 0.25 1.75
B6 0.10 1.90
Color test solutions BG
Color test solution Color standard solution BG in ml Hydrochloric acid 1 % (m/V) in ml
BG1 2.00 0.00
BG2 1.50 0.50
BG3 1.00 1.00
BG4 0.50 1.50
BG5 0.25 1.75
BG6 0.10 1.90
Color test solutions G
Color test solution Color standard solution G in ml Hydrochloric acid 1 % (m/V) in ml
G1 2.0 0.00
G2 1.50 0.50
G3 1.00 1.00
G4 0.50 1.50
G5 0.25 1.75
G6 0.10 1.90
Color test solutions GG
Color test solution Color standard solution GG in ml Hydrochloric acid 1 % (m/V) in ml
GG1 0.5 1.50
GG2 0.30 1.70
GG3 0.17 1.83
GG4 0.10 1.91
GG5 0.06 1.94
GG6 0.03 1.97
Color test solutions R
Color test solution Color standard solution R in ml Hydrochloric acid 1 % (m/V) in ml
R1 2.00 0.00
R2 1.50 0.50
R3 1.00 1.00
R4 0.75 1.25
R5 0.50 1.50
R6 0.25 1.75
Appendix A 621

In the US pharmacopeia color standard solutions are referred to as color matching

fluids and are given letter designations A through T, as shown in Table A.33.

Table A.33 Color Standards according to USP XII [16]

Matching Parts of Cobalt Parts of Ferric Parts of Cupric Parts of
fluid Chloride CS Chloride CS Sulfate CS water
A 0.1 0.4 0.1 4.4
B 0.3 0.9 0.3 8.5
C 0.1 0.6 0.1 4.2
D 0.3 0.6 0.4 3.7
E 0.4 1.2 0.3 3.1
F 0.3 1.2 0.0 3.5
G 0.5 1.2 0.2 3.1
H 0.2 1.5 0.0 3.3
I 0.4 2.2 0.1 2.3
J 0.4 3.5 0.1 1.0
K 0.5 4.5 0.0 0.0
L 0.8 3.8 0.1 0.3
M 0.1 2.0 0.1 2.8
N 0.0 4.9 0.1 0.0
O 0.1 4.8 0.1 0.0
P 0.2 0.4 0.1 4.3
Q 0.2 0.3 0.1 4.4
R 0.3 0.4 0.2 4.1
S 0.2 0.1 0.0 4.7
T 0.5 0.5 0.4 3.6
Literature

1. BIPM - On the revision of the International System of Units (SI) (2018) Bureau International de
Poids et Mesures, Versailles
2. Kurzweil P (2013) Das Vieweg Einheiten-Lexikon: Begriffe, Formeln und Konstanten aus
Naturwissenschaften, Technik und Medizin. Vieweg, Wiesbaden. https://doi.org/10.1007/978-
3-322-92920-4
3. Database of Physical Properties of Food Food Properties Awareness Club. http://www.nelfood.
com
4. Rahman S (2009) Food properties handbook. CRC Press, Boca Raton, FL. https://doi.org/10.
1201/9781420003093
5. Kessler HG (2002) Food and bio process engineering: dairy technology. A. Kessler, München
6. Panagiotou NM, Krokida MK, Maroulis ZB, Saravacos GD (2004) Moisture diffusivity:
literature data compilation for foodstuff. Int J Food Properties 7:273–299. https://doi.org/10.
1081/JFP-120030038
7. Saravacos GD, Maroulis ZB (2014) Transport properties of foods. CRC Press, Boca Raton,
FL. https://doi.org/10.1201/9781482271010
8. Krokida MK, Panagiotou NM, Maroulis ZB, Saravacos GD (2001) Thermal conductivity:
literature data compilation for foodstuffs 2001. Int J Food Prop 4:111–137. https://doi.org/10.
1081/JFP-100002191
9. Berk Z (2018) Physical properties of food materials. Food process engineering and technology,
pp 1–29. https://doi.org/10.1016/B978-0-12-812018-7.00001-4
10. Arana I (2016) Physical properties of foods - novel measurement techniques and applications.
CRC Press, Boca Raton, FL. https://doi.org/10.1201/b11542
11. Gröber E, Grigull U (1963) Die Grundgesetze der Wärmeübertragung. Heidelberg, Berlin.
https://doi.org/10.1007/978-3-662-29015-6
12. Tscheuschner HD (2016) Grundzüge der Lebensmitteltechnik. Behrs, Hamburg
13. Fraden J (2016) Handbook of modern sensors physics, designs, and applications. https://doi.
org/10.1007/978-3-319-19303-8
14. De Whalley HCS (1964) ICUMSA methods of sugar analysis. International Commission for
uniform methods of sugar analysis. Elesvier, Amsterdam
15. Europäsches Arzneibuch. Dt. Apotheker Verlag, Eschborn
16. United States Pharmacopeia USP. United States Pharmacopeial Convention, Rockville, MD

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 623
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8
Index

A Amorphous, 42, 515

Absolute temperature, 312 Ampere, 585
Absorbent, 23 Amphiphilic, 260
Absorption, 495 Analyte, 572
range, 476 Analytical sieving, 134
spectrum, 475 Andrade, 183
Absorptivity, 334 Angle of rotation, 492
Accuracy, 576 Angular velocity, 159
Acoustics, 539 Anion, 403
characterization, 545 Anisotropic, 151, 354
information, 545 electrical conductivity, 415
spectrum, 545 Anisotropy, 407
technology, 546 Anomaly, 71
Activation energy, 183, 299 Antibody, 571, 573
Activity, 2, 322, 525, 532 Anti-ferromagnetism, 434
coefficient, 7 Antigen, 571
Additive color mixing, 497, 499 Antioxidant capacity, 450
Additive resistors, 285, 295, 305, 339, 343, Apparent mass, 62
347, 354 Apparent thermal conductivity, 359
Adhesion, 267, 269 Apparent viscosity, 179
energy, 267 Approximation, 260
Adiabatic, 319 Aptamer, 573
demagnetization, 434 Apta-sensor, 573
mode, 393 Archimedes, 79
Admittance, 404, 419 Areometer, 84
Adsorbate, 23 Arithmetic mean, 104
Adsorption, 7, 12, 18, 271 Arrhenius, 34, 37, 183, 299
equilibrium, 11 At-line, 553
isotherm, 22, 271 Atomic angular momentum, 432
Aerosol, 102 Atomic nucleus, 443
Agglomeration tendency, 424 Atomic number, 443, 517
Air, 63 Attenuation, 470
classification, 140 Atwater, 370, 373
density, 63 Audibility, 544
Alcoholometer, 85 Audible, 223
α-radiation, 524 Auger electron, 516
Alternating DSC (ADSC), 387 Austenitic steel, 433
Alternating electric field, 458 Authentification, 447

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 625
L. O. Figura, A. A. Teixeira, Food Physics,
https://doi.org/10.1007/978-3-031-27398-8
626 Index

Avogradro, 62 Buoyancy correction, 62

Avoidurpois, 591 Burgers, 191
Burgers element, 238

B
Balance, 60, 61, 65 C
Ball drop measurement, 216 Cal, calorie, 318, 366
Ball viscosimeter, 215 Calibration, 61, 62, 562
Basal metabolic rate (BMR), 365, 367 Caloric value, 364–374
Basal metabolism, 364 Calorimeter, 392
Basic unit, 585 Calorimetry, 367, 604–605
Baumé, 85 Candela, 585
Becquerel, 525 Cannon–Fenske, 215
Bending fracture test, 243 Capacitance, 404, 567, 572
Bending oscillator, 560 Capacitive balance, 65
Bending test, 243 Capacitive reactance, 404, 416, 418
Best-before, 43 Capacitive sensor, 567
Best by, 44 Capacitor, 567
β-radiation, 524 Capillary absorption, 14
Bi-disperse, 105 Capillary adsorption, 17
Bilayer, 263 Capillary condensation, 15, 17
Binding heat, 8 Capillary method, 274
Bingham, 167, 173 Capillary pressure, 250, 253
Bio-detection, 572 Capillary viscometer, 214
Biopolymer, 298 Capsule, 263
Bio-recognition, 572, 573 Carnot, 314
Bio-sensitive, 571, 572 Casson, 180, 181
Bio-sensor, 551, 571, 572, 575 Cataphoretic effect, 412
Biot, 491 Cation, 403
Black body, 334 Cavitation, 539, 546
Blaine, 141 Cell disintegration index (CDI), 415
Blood cell, 134 Cell growth, 570
Boltzmann, 312 Celsius, 313
Boltzmann factor, 597 Central value, 103
Bonding, 12, 540 Centrifuge separation, 90
Bostwick, 216 Charge number, 405
Boundary layer, 260, 345 Checkweigher, 560
Bound water, 1, 8, 13 Chemical potential, 2, 322, 328
Bq, 525 Chemical shift, 445
Brewster angle, 495 Chemisorption, 8, 572
Brightness, 500, 502, 503 Chemo-sensitive, 572
Brilliance, 500, 502, 503 Chemo-sensor, 551, 572, 575
Brown, 133 Chewing noise, 545
Brunnauer, Emmet, Teller (BET), 20, 39 Chiral, chirality, 492
equation, 29 Chirality, 494
model, 22 Ci, 527
one-point, 32 CIE, 499, 500
surface, 141 Circularly polarized light, 493
Bubble, 252 Clausius-Clapeyron, 35
Bubble point pensiometer, 279 Clausius-Mosotti-Debye, 460
Bulk module, 152 Closed loop, 558
Bulk modulus, 147 Closed pore, 15
Buoyancy, 62, 77, 79, 82, 86, 88, 273, 379 Closed-pore system, 353
Index 627

Coating, 334 Contact angle, 266, 272, 273, 279

Coaxial cylinder system, 193 Contamination, 565
Cohesion, 267, 269 Continuous phase, 101
Cohesion energy, 267 Continuous wave (CW), 446
Cold plasma, 426 Contour analysis, 276, 277
Colloid, 261, 265 Control, 558
Color, 483, 497, 500, 571 Controlled shear rate (CSR), 195
difference, 503 Controlled shear stress (CSS), 195
match, 504 Convection, 344, 357
perception, 497 Convex, 252, 253
plane, 502 Cookware, 441
space, 500, 503 Coriolis force, 563
standard, 504, 618 Couette, 194
triangle, 499 Coulomb force, 403, 528
vector, 501 Counter tube, 529
Colorimetry, 497 Creep, 189
Combustion, 369 Creep test, 230, 232, 236
calorimeter, 373, 392 Crispness, 539
energy, 364 Critical micelle formation concentration (cmc),
Complementary color, 500 261
Complex conductance, 419 Critical micelle temperature (cmt), 261
Complex elasticity module, 241 Critical temperature, 256
Complex electrical resistance, 404 Curie, 527
Complex heat capacity, 390 Curie temperature, 378, 432, 433, 441
Complex heat flow, 390 Curie transition, 377
Complex interfacial elasticity, 280 Curved interfaces, 250
Complex number, 212, 390, 597, 602
Complex permittivity, 465
Complex physical quantity, 212 D
Complex quantity, 419, 465, 603 Damper, 188
Complex refractive index, 132 Damping, 446, 465
Complex resistance, 419 Deborah, 239
Complex shear modulus, 213 De Brouckere diameter, 128
Complex viscosity, 211 Debye, 460
Compliance, 190, 211, 604 Debye-Falkenhagen effect, 413
Composite film, 295 Debye length, 423
Compressibility, 72, 151 Decay, 450, 525
Compression, 147, 151, 226 Decay constant, 527
modulus, 542 Decibel (dB), 541
test, 231 Decile, 106
Compressive stress, 226 Degradation reaction, 41
Compton effect, 528 De Haven, 176
Computed tomography (CT), 518 Density, 59, 64, 560, 609
Concave, 250, 253 Density gradient column, 89
Condensing, 349 De number, 239
Conductance, 289 Depletion, 263, 271
Conductive heating, 403 Derived unit, 585
Conductive resistance, 354 Desiccator method, 48
Conductivity, 289 Desinfection, 514
Conductometry, 420 Desorption, 8, 12, 17, 374
Cone-plate system, 195, 200 Desorption isotherm, 22
Cones, 499 Detection limit, 576
Consistency coefficient, 177, 178, 181 Diamagnetic material, 433, 437
628 Index

Diamagnetism, 433–435 E
Dielectric, 456 Ear drum, 540
Dielectric heating, 458 Earth, 60, 61, 438
Dietary fibre, 364 Eddy current, 439
Differential, 321 Ehrenfest, 329, 330
Differential scanning calorimeter (DSC), 356, Elastic, 146
379, 382 absorption, 466
Diffraction, 496 body, 153
Diffuse layer, 423 case, 458
Diffuse reflection, 495 oscillator, 465
Diffusion, 285 Elasticity, 72, 148, 150, 228, 239, 539
Diffusion coefficient, 287 Electrical capacitance, 416
Dilatancy, 169 Electrical conductance, 405
Dilatant, 169, 177 Electrical conductivity, 305, 403, 405, 422,
Dipole moment, 459 466, 518
Dispersant, 136 Electrical double layer, 424
Disperse, 101, 255 Electrical nose, 575, 576
Disperse phase, 101 Electrical polarizability, 416, 462, 467
Dispersion, 101, 420, 464, 487, 496 Electrical polarization, 464
Dispersion of permittivity, 420 Electrical potential, 423
Displacement polarization, 463 Electrical properties, 403
Displacement work, 316 Electrical resistance, 134, 403, 405, 559
Dissipation, 465 Electrical resistivity, 404
Dissipation factor, 467 Electrical voltage, 405
Distribution density, 103 Electric bilayer, 423
Distribution density function, 111, 138 Electric cell perforation, 420
Distribution function, 103, 591 Electric current, 403
Distribution sum, 138 Electric field, 403, 455
Distribution sum function, 110, 138 Electric field strength, 403
DNA, 573 Electric polarization, 456
Doppler effect, 546, 567 Electrolyte, 407, 408, 473
Dose, 533 Electromagnetic absorption, 458
Dose equivalent, 533 Electromagnetic detection, 564
Drag force, 139, 140, 412 Electromagnetic induction, 565
Drop, 254, 265 Electromagnetic radiation, 455, 483, 497
shape, 265 Electromagnetic spectrum, 483
viscometer, 216 Electromagnetic wave, 455, 513
volume tensiometer, 278 Electron, 304, 524, 528
volumeter, 279 beam, 536
Droplet, 16, 101, 249, 252, 279 capture, 535
Dry basis (db), 24 conductivity, 352
Drying, 13, 38, 39, 42, 45, 374 spectroscopy, 513
DSC oven, 380 Electronic tongue, 575
DTG, 375 Electron paramagnetic resonance (EPR), 450
Dulong-Petit, 327 Electron spin resonance (ESR), 442, 450–451,
Du Noüy, 272 535
Dynamic dewpoint isotherm (DDI), 49 Electro-optical, 494
Dynamic isopiestic method (DIM), 49 Electro-perforation, 426
Dynamic light scattering (DLS), 132, 133 Electrophoretic analysis, 413
Dynamic mechanical analysis (DMA), 213, 241 Electrophoretic effect, 413, 422
Dynamic vapour sorption (DVS), 49 Electrophoretic mobility, 424
Dynamic viscosity, 158, 164 Electro-resistive sensor, 567
Index 629

Electrozone counter, 134 Film former, 267

Ellipsoid particle, 113 First law of thermodynamics, 318
Ellipsometry, 494 First order transition, 329
Elliptical polarized radiation, 493 Fixed point, 314
Elliptically polarized light, 493 Flavoromics, 577
Ellis, 176 Floating method, 88
Elongation, 226 Flocculation, 424
Emissivity, 334 Flow, 226
E-modulus, 150, 153 behavior index, 177
Emulsion, 101, 263 exponent, 177, 178
Enantiomer, 492 sensor, 563
Encapsulation, 263 Fluid, 9
Energy, 249, 314 Fluidity, 165
Enthalpy, 317, 384 Fluorescence, 516
of adsorption, 34 Flux, 287
change, 349 Foam/foaming, 90, 101, 353
Entropy, 319, 322, 465 Foodomics, 577
Entropy of sorption, 38 Food safety, 311, 573
Enzymatic tanning, 41 Form factor, 118
Eötvös, 256 Fourier, 305, 335, 337, 339, 345, 347, 351, 355,
Equivalent conductivity, 408, 414 360, 390
Equivalent diameter, 113, 133, 134, 136, 140 Fourier transformation, 447
Equivalent number, 408 Fraction, 109
ESR spectroscopy, 535 Fracture, 224
Euclid, 155 strain, 242
Euklid, 183 stress, 242
Euler, 600 test, 242
eV, 515 Fraunhofer, 132
Evaporation, 8, 12, 349 Fraunhofer diffraction, 496
Exergetic, 38 Free energy, 319
Extensional viscosity, 226 Free enthalpy, 262, 319, 322
Extinction, 498 Free induction decay (FID), 446, 449
Free water, 8, 13, 38
Frequency, 455
F Frequency sweep, 213, 446
Fahrenheit, 314 Freundlich, 39
Failure strain, 227 Freundlich-model, 18, 20
Failure stress, 227 Friction, 140
Far infrared radiation (FIR), 474 Frozen fraction, 356
Feces, 364 Fugacity, 3, 7, 322
Feret, 118 Functionalized surface, 574
Ferrimagnetism, 434 Fundamental oscillation, 476
Ferrites, 434
Ferritic steel, 441
Ferromagnetic, 377, 433, 441 G
materials, 433, 437 γ-quant, 528
steel, 441 Gamma radiation (γ-radiation), 450, 524
Ferromagnetism, 433 Gas adsorption method, 22
Ferrous material, 433 Gas chromatography, 535
Ferry, 176 Gas constant, 75
Fibrous, 353 Gates, Gaudin and Schuhmann (GGS), 129,
Fick, 287, 305 130
Field sweep, 446 Gauss, 107
Filling level, 570 Geiger-Müller counter, 529
Film, 291 Gel electrophoresis, 413
630 Index

Gibbs, 37, 329, 331 Herschel-Bulkley, 180

Gibbs energy, 262, 319 Heywood factor, 126
Gibbs enthalpy, 38 Hook, 183, 189
Glass transition, 42, 185, 328, 332, 386, 387, Hooke, 147, 148, 150, 155, 158, 212, 231
391 Hookean, 148, 232, 239
Glass transition temperature, 43 Höppler, 216
Glassy state, 43, 186 Hue, 500, 502, 503
Gloss, 495 Human body, 531, 533
Glossmeter, 495 Human ear, 540, 543, 544
G-modulus, 150, 153 Human eye, 483, 497, 499
Gravitational acceleration, 60 Human viewer, 497
Gravitational force, 60 Humidity, 570
Gravity, 60 Huygens, 485, 496
Gray, 533 Hydration, 263
Guggenheim, Anderson, de Boer (GAB), 23, 40 Hydrogen bond, 11
equation, 32 Hydrometer, 84
model, 23 Hydrophilic, 260
parameter, 33 Hydrophobic, 260
Gy, 533 Hydrostatic balance, 79, 80, 82
Gyromagnetic ratio, 447 Hygroscopicity, 17, 27, 28
Hyper sound, 540
Hyperspectral imaging (HSI), 478
H Hysteresis, 17, 18, 22, 437, 438
Habitus, 118
Hagen-Poiseuille, 214
Half-life, 526 I
Half-power depth, 472 Ideal gas, 3, 74, 325, 597
Half-value depth, 525 Ideal solution, 3
Hall probe, 442 Image analysis, 108, 117
Haptic, 223 Imaginary, 132
Harmonics, 476 number, 597
Hearing, 223 part, 212, 465
Heat, 304, 314, 458, 465 permittivity, 466
capacity, 323, 325, 327, 361, 390, 605 unit, 419, 597
conductance, 354 Impedance, 403, 418, 422, 570, 572, 603
conduction, 335 Impedance-based index, 420
exchanger, 335, 337, 349, 352 Impedance measurement, 415, 420
flow, 347, 604 Inactivation
flow calorimetry, 379 bacteria, 514
radiation, 333 microorganisms, 539
transfer, 311, 332, 345, 470 Incompressible, 155
transfer coefficient, 345, 346 Inductance, 404, 417
transfer resistance, 345 Induction, 439
treatment, 311 cooker, 426
Heating, 470 cooking, 440
Heinz, 180 time, 422
Helmholtz energy, 319 Inductive metal detector, 439
Helmholtz potential, 423 Inductive reactance, 404, 417, 418
Hencky, 238 Inelastic absorption, 466, 468
Henry, 298 In-elastic case, 458
Henry-Gesetz, 20 Infrared-active, 475
Henry´s Law, 20 Infrared radiation (IR), 333, 474
Heraclitus, 145, 240 Infra-red thermometry (IR thermometer), 334
Index 631

Infrasound, 539 Lactometer, 85

In-line sensor, 551, 552 Lambert-Beer law, 495, 516
In-phase, 466 Lambert-Beer’s, 497
Inspection, 475 Laminar, 139, 166
Integral mean, 119–121, 123 Langmuir, 20, 39, 269, 271, 279
Integrating sensor, 558 Langmuir-Blodgett film, 269
Interface, 9, 247 Langmuir model, 19
Interface-active, 260 Laplace, 252, 253, 276, 279
Interfacial elasticity, 280 Larmor frequency, 444, 494
Interfacial energy, 249, 264 Laser, 131
Interfacial phase, 247 diffraction, 108, 131, 496
Interfacial tension, 86, 248, 256, 272, 279 diffractometer, 132
Interference, 496 Latent heat, 314, 349, 350
Intermediate moisture, 47 LED, 514
Internal energy, 314 Lenz´s law, 433
International Office of Cocoa, Chocolate and Lewicki, 39
Sugar Confectionery (IOCCC), 182 Life time, 255, 279
International system de unite, 60 Light, 483, 484, 497
International temperature scale (ITS), 314 Light scattering, 133
Interval limit, 109 Linearly polarized light, 490
Intrinsic viscosity, 187 Liquid crystal display (LCD), 494
Ion, 403, 407 Liter, 591
Ionization, 425 Log-normal-distribution, 130
Ionization detector, 528 Lorentz force, 440, 442, 528, 564
Ionizing radiation, 515, 533 Loss, 465
Irradiation, 534–536 factor, 465, 468, 473
Isentropic, 319 tangent, 465
Isentropic coefficient, 542 Loudness, 543
Isobaric, 317 Loudness level, 543
Isometric, 112 Low angle laser light scattering (LALLS), 131
Isoperibole, 393 Low-resolution nuclear resonance spectroscopy
Isosteric sorption enthalpy, 36 (LR-NMR), 449
Isotope, 443, 447, 449, 524 Lyophilic, 260
Isotropic, 151–153, 353 Lyophobic, 260

K M
Kelvin, 254, 312, 585 Magmeter, 564
Kelvin body, 156 Magnetar, 442
Kelvin element, 188, 189, 237 Magnetic dipole, 431
Kelvin`s equation, 15 Magnetic field, 431, 435, 455, 564
Kelvin/Vogt, 191 Magnetic history, 438
Kilogramm, 585 Magnetic inductive flowmeter, 442
Kinematic viscosity, 165 Magnetic-inductive flow sensor, 564
Kinesthetic, 223 Magnetic moment, 434, 443, 450
Kinetic energy, 312 Magnetic momentum, 432
Kirchhoff, 334 Magnetic permeability, 431, 435, 436
K-modulus, 153 Magnetic polarity, 433
Kohlrausch, 408 Magnetic polarization, 433, 434
Krafft, 262 Magnetic properties, 431
Kurtosis, 107 Magnetic refrigeration, 434–435
Magnetic resonance, 442, 445
Magnetic resonance imaging (MRI), 450
L Magnetic resonance spectroscopy, 444
Lab system, 502, 503 Magnetic stainless steel, 441
632 Index

Magnetic susceptibility, 436, 437, 439, 518 distribution, 105

Magnetization, 377, 431, 436 Model, 23, 40, 258
Magnetocaloric effect, 434 Modulus of elasticty, 151
Magneto-optical effect, 494 Mohr-Westphal balance, 82
Margules, 198, 199 Moisture, 1, 570
Martensitic steel, 441 Moisture content, 23
Martin, 118 Mol, 585
Mass, 59 Molecular imprintable polymer (MIP), 573
Mass density, 59 Mono-disperse, 105
Mathematical model, 40 Monolayer, 12, 13, 22
Maxwell, 483 mass, 22
Maxwell body, 156 moisture content, 30
Maxwell-Boltzmann, 13, 595 sorption enthalpy, 37
Maxwell element, 188, 234 Monomolecular, 260, 269
Maxwell equation, 464 Multilayer, 12, 13, 20, 285, 292, 295, 343, 348
MDSC, 387, 391, 604 Multiple emulsion, 265
Mechanical oscillation, 562 Multi-sensor, 575
Median, 104, 125, 138 Must weight, 86
Mesopore, 15 Mutarotation, 494
Metabolic equivalent (MET), 369
Metabolic rate (MR), 365
Metabolomics, 576 N
Metal detector, 439, 518, 565 Nano-dispersion, 265
Metal separation, 439 Nanoemulsion, 265
Metal solids, 352 Nanoparticle, 134
Meter, 585 Nanopore, 15
Metrology, 575 Natural radionuclide, 530
mho, 406 Near-infrared (NIR), 475, 571
Micelle, 261, 271 Near-infrared spectroscopy (NIRS), 477
Microbalance, 563 Néel temperature, 434
Microbalance sensor, 574 Negative adsorption, 271
Micro-cantilever, 574 Neutron, 443, 523, 535
Microcrystalline cellulose (MCC), 51 Newton, 155, 176, 183, 189, 231, 232
Micro-CT, 518 Newton element, 235
Micro-Electro-Mechanical Systems (MEMS), Newtonian, 232
560, 563, 573 Newtonian fluid, 164
Microemulsion, 264 Nitrogen adsorption, 141
Microorganisms, 570 Noise, 544
Micropore, 15 Non-enzymatic browning, 41
Microscopic, 131 Non-isotropic, 353
Microwave, 467 Non-Newtonian, 173
absorption, 473 Non-Newtonian fluid, 166
heating, 466, 468 Non-reversing heat flow, 390
oven, 458, 469 Non-wetting, 267
packaging, 473 Nuclear decay, 529
MID, 442 Nuclear magnetic resonance (NMR), 356, 442,
Middle fifty, 107 445, 447
Mid infrared radiation (MIR), 474 spectrometer, 445
Mie, 132 spectroscopy, 449
Mie scattering, 496 spectrum, 447
Miscibility, 247 Nuclear spin, 443
Mist, 101 Nuclear spin resonance, 442
Mode, 104, 125, 138 Nuclide, 523
Index 633

Number, 598 size, 102, 111, 131, 140, 539

Number distribution, 103, 121 size analysis, 108
size distribution, 129, 496
sizing, 496
O Pascal, 155, 183
Oe, 86 Peeling test, 227
Oechsle, 86 Pendant drop, 276
Off-line, 553 Penetration depth, 426, 470, 472–474, 525
Ohm, 305, 339, 355, 406 Penetration test, 227
Ohmic heating, 426 Percentage, 25
Ohmic resistance, 404, 418 Percentile, 106
Ohm’s law, 405 Perception of sound, 545
Online sensor, 551, 554 Permanent dipole, 457, 463
Open loop control, 558 Permanent polarization, 459
Open pore, 15 Permeability, 285, 288
Open-pore system, 353 Permeability coefficient, 298
Optical activity, 490 Permeation, 289, 298, 299
Optical properties, 483 Permittivity, 416, 458, 463, 473, 570, 604
Optical rotation, 490 Phase, 247
Organoleptic, 575 change, 349
Orientation polarization, 457, 463 shift, 209, 391
Oscillating air, 540 transition, 328, 332, 357
Oscillating shear stress, 208, 241 Photo effect, 516
Oscillating strain, 240 Photoelectric effect, 529
Oscillating tube, 90 Photometer, 495
Oscillation rheometer, 213 Photometry, 498
Ostwald, 255 Photon, 515
Ostwald-de-Waele, 176, 177, 199 Photon correlation spectroscopy (PCS), 132,
Ostwald-Faktor, 178 133
Out-of-phase, 466 Physical activity, 365, 368
Oven drying, 48 Physical activity level (PAL), 367, 368
Overall heat transfer coefficient, 347, 349 Physical caloric value, 369, 373
Overflowing cylinder, 279 Physical constants, 585
Overpressure, 252 Physical equivalent diameter, 114
Overrun, 90 Physikalisch-Technische-Bundesanstalt (PTB),
Overtone, 476 61
Oxidation, 41 Physiological caloric value, 370, 373
Oxidation stability index (OSI), 422 Physiological energy, 370
O/W emulsion, 263 Physisorption, 8, 572
Pickering, 265
Piezoelectric, 567
P Piezoelectric effect, 545
Packaging, 292, 303 Piezoresistive, 559
Packaging material, 285, 292 Plan sieving, 134
PAR, 368 Planet, 60, 61
Parallel resistors, 290 Plasma, 425, 426
Paramagnetic material, 432, 437 Plasmon resonance, 571
Paramagnetism, 431, 432 Plastic, 148, 155, 173, 180
Partial differential, 321 Plastic flow behavior, 170
Partial integration, 385 Plasto-viscoelastic, 191
Partial pressure, 3 Plate capacitor, 416
Particle, 102 Plate heat exchanger, 335
shape, 118 Plate-plate system, 195, 202
634 Index

Plate-type, 335 Quartile distance, 107

Plato, 113 Quartz crystal microbalance (QCM), 563
Poise, 166 Quasi-elastic light scattering (QELS), 133
Poiseuille, 166
Poisson, 150, 154, 155
Polarimeter, 490, 494 R
Polarimetry, 493 Racemate, 492
Polarizability, 420 Rad, 533
Polarizable component, 416 Radiation, 535
Polarization, 490 Radical, 535
potential, 458, 460, 463 Radioactivity, 523
volume, 459 Radionuclide, 523
Polyacrylamide gel electrophoresis (PAGE), Rainbow, 500
413 Random deviation, 64, 65
Polydisperse, 105 Raoult, 4, 40
Pore, 9, 17 Rate of shear, 160
radius, 15 Rayleigh scattering, 496
shape, 15 Reactance, 403, 418
Porosimetry, 17 Real part, 212, 465
Porosity, 14, 473 Real-time, 553
Positron, 524, 528 Receptor, 572
Potassium, 530 Reference weight, 61, 63
Potential, 286, 320 Reflectance, 495
Potential gradient, 403 Reflection, 494
Pound, 591 Reflective body, 334
Powder, 22, 27, 101 Refraction, 483, 571, 572
Power compensation, 382 Refraction index, 485
Power-law, 129, 177 Refractive index, 464, 488
Precession, 444 Refractometer, 490, 571
Precession frequency, 444 Refractometric sensor, 571
Precision, 575, 576 Reiner, 176
Prefix, 588 Relative density, 66, 77, 82
Preservation, 45, 536 Relative fugacity, 322
Pressure, 147 Relative humidity, 1
Pressure scan, 331 Relaxation, 446, 449
Primary ion, 529 test, 234
Process analytics, 555 time, 189, 234
Process automation, 553 Rem, 533
Process control, 553, 557 Remanent magnetization, 438
Process monitoring, 555 Repeatability, 576
Propagation speed, 455 Reproducibility, 576
Proton, 443, 449, 523 Residue, 136
Pseudoplastic, 167, 168, 177 Resistance, 289, 603
Pulsed NMR, 446 Resistive balance, 65
Pycnometer, 77 Resistivity, 290
Pythagorean sum, 419 Resistor, 290
Resolution, 576
Resonance, 560, 572, 575
Q Resonant frequency, 90
Quality factor, 533 Resonator, 563
Quantile, 106 Retardation time, 191
Quantity, type of, 107, 121 Reversible heat flow, 390
Quartile, 106 Rheology, 145, 224, 604
Index 635

Rheometer, 193 oscillation, 213

Rheopectic, 170 rate, 159
Rheopexy, 170, 173 speed, 160
Rigid, 155 strain, 152
Rigidity, 153 stress, 153, 161
Ring method, 273 test, 227
RNA, 573 velocity, 160
Roasting, 374 Shear-thinning, 168
Rods, 499 Shelf life, 38, 39, 41, 43, 292, 534
Rosin, Rammler, Sperling and Benett (RRSB), Shelf stable, 47
130 SI, 60, 585
Rotational dispersion, 491 Siemens, 406
Rupture, 148 Sieve analysis, 108, 136
strain, 242 Sieve mesh size, 113
test, 242 Sievert, 533
Sieving, 113, 134
Sisko, 176
S Size class, 103, 109, 134, 136
Saccharimeter, 85, 494 Skewness, 107
Safety, 292 Snellius, 487
Sampling, 131 Sodium dodecyl sulfate (SDS), 413
Saturated steam, 351 Solid fat content (SFC), 450
Sauter diameter, 127 Solid foam, 102
Scattering, 496 Solid surface, 7
Scattering, of light, 131 Solvent, 136
Scissoring oscillation, 475 Sorbate, 301
Screening aid, 135 Sorbent, 300
Searle, 194 Sorption, 8
Second, 585 enthalpy, 34, 36, 39
law of thermodynamics, 319 equilibrium, 13
order transition, 329 isoster, 49
Secondary ion, 529 isotherm, 13, 14, 18, 27, 28, 48
Sedimentation, 139 Sorptive, 12
Selective absorption, 497 Sound, 539
Selectivity, 573, 576 intensity, 541
Self-luminous, 498 level, 541
Sensitive heat, 312, 314 power, 540
Sensitive layer, 572, 573, 575 Span, 107
Sensitivity, 573, 576 Specific activity, 527
Sensomics, 577 Specific electrical conductivity, 406
Sensor, 552 Specific electrical resistance, 405
Sensorial analysis, 575 Specific gas constant, 75
Sensory, 223 Specific gravity, 66
Sensory quality, 223 Specific heat, 323
Serial resistors, 290 Specific heat capacity, 323
Sessile drop, 276 Specificity, 575, 576
Shape, 276 Specific polarization potential, 461
Shape of particles, 118 Specific quantity, 69
Shear, 146 Specific rotation, 491
activation energy, 183 Specific sorption enthalpy, 34
angle, 153, 161 Specific surface area, 14, 125
force, 161 Specific viscosity, 187
modulus, 147 Specific volume, 67
636 Index

Spectrophotometry, 498 Systematic deviations, 64, 65

Speed of light, 456, 483 System International, 585
Speed of sound, 539, 541 Szyszkowski, 261
Spherical, 102, 113
Sphericity, 119
Spinning drop, 277 T
Spoilage, 41, 44, 292 Tactile, 223
Spray, 101 Temperature, 312, 324, 606
Spread, 103 coefficient, 421
Spreading, 266 gradient, 336
Spreading pressure, 267 modulated DSC, 387
Spring, 188 scan, 331
Stability, 576 Temporary dipole, 457
Stainless steels, 352, 433 Tensile stress, 226
Stalagmometer, 278 Tensile test, 227, 231
Standard, 51 Tension, 147
climate, 301 Terahertz absorption, 474
deviation, 121 Terahertz radiation, 474
state, 299 Terahertz waves, 474
volume, 288 Tercile, 106
Statistical moment, 119, 121, 125, 141 Texture, 223
Staudinger, 187 analysis, 224, 545
Steady state, 14, 285, 292, 300, 335 test, 232
Stefan-Boltzmann, 333 Texture profile analysis (TPA), 244
Steiner-Steiger-Ory, 176 Thermal analysis (TA), 374
Stern layer, 423, 425 Thermal conductance, 346
Stern´s potential, 423 Thermal conductivity, 304, 337, 351, 354–356,
Sticky, 27, 42 362
Stokes, 139, 140, 166, 215 Thermal degradation, 376
Stokes-Einstein, 133 Thermal diffusivity, 360, 361
Strain, 146, 238 Thermal energy, 312, 324
gauge, 559 Thermal expansion, 69
retardation, 236 Thermal expansion coefficient, 71
Stress, 153, 226 Thermal insulation, 337, 353, 358, 359
relaxation test, 232 Thermal process operations, 311
test, 230 Thermal properties, 311
Stress-strain diagram, 147, 227 Thermal resistivity, 337
Stretch oscillation, 475 Thermodynamic activity, 5
Strong electrolyte, 409 Thermodynamics, 314, 318
St. Venant, 155, 183 Thermodynamic temperature, 312
Submersion technique, 86 Thermogravimetry (TG), 374–379
Subtractive color mixing, 497, 498 Thickness measurement, 535
Sum distribution, 110 Thick-walled tube, 342
Surface, 9, 247 Thin walled tube, 342
Surface-active, 260, 271 Thixotropic, 169
Surface plasmon resonance (SPR), 571 Thixotropy, 173
Surface pressure, 268 Throwing sieving, 134
Surfactant, 260 Time constant, 558
Susceptance, 404 Tomography, 518
Susceptibility, 458 Total resistance, 291
Susceptor, 473 Transient measurement, 363
Suspension, 101 Transient state, 300, 335, 360
Sv, 533 Transition temperature, 329
Index 637

Transmission, 497 Viscous body, 231

Transparent, 495 Vogel, 184
Transport equation, 286, 287, 304, 305
Transport process, 285, 287
Transverse strain, 154 W
Troy, 591 Water activity, 1, 41, 44, 322
True density, 78 Water activity standard, 50
True mass, 62 Water binding, 374
Trueness, 575, 576 Water content, 1, 24, 25
Tubular heat exchanger, 335 Water vapor, 301
Tubular-type, 335 Water vapor permeability, 301
Turnover plot, 386 Water vapour, 3
Two-platen model, 160 Waveguide, 571
Wavelength, 455, 474
Wave number, 474
U Weak electrolyte, 409
Ubbelohde, 215 Weighing, 65, 559
Ultracentrifuge, 90 Weight, 60
Ultrasonic, 135, 141 Weighted mean, 119
flow meter, 566 Weiss, 441
puls, 567 Weiss region, 433
waves, 545 Wet basis (wb), 24
Ultrasonic attenuation spectroscopy (UAS), 141 Wet screening, 135
Ultrasonography, 546 Wetting, 267, 269, 280
Ultrasound, 539, 545 angle, 273
Ultraviolet radiation, 513 tension, 269
Uncertainty, 260 Whipping, 90
Uniaxial, 145, 226, 227 Wiedemann-Franz, 352
Uniformity index, 107 Wilhelmy, 272
Unsteady state, 335, 360 Williams-Landel-Ferry (WLF), 185, 186
Urine, 85, 364 Windhab, 182
UV spectroscopy, 513 W/O emulsion, 264
UV/vis spectrometer, 514

X
V X-ray absorption, 517
Vacuum, 62, 358, 457 X-ray diffractometry, 515
Vacuum UV, 513 X-ray imaging, 517
van der Waals, 11 X-ray inspection, 517
Vapor pressure, 1 X-ray photon, 516
Vapour pressure reduction, 17 X-rays, 513
Variance, 121 X-ray scanner, 517
Vesicle, 263
Vibrating tube, 90
Vibration sieving, 134 Y
Vibration technology, 562 Yield, 148
Virus, 573 point, 170, 171
Viscoelastic, 156, 187, 232, 234, 238 stress, 167, 170, 180
Viscoelasticity, 239 Young, 147, 150, 269
Viscometer, 193
Viscoplastoelastic, 239
Viscosity, 239, 564, 604 Z
Viscosity function, 165 Zeeman effect, 445, 450
Viscous, 155, 157, 226 Zeta potential, 424, 425