High Performance Non-Oxide Ceramics II

102
Structure and Bonding
Managing Editor:
D.M.P. Mingos
Editorial Board:
AJ. Bard • I.G. Dance • P. Day • J.A. Ibers • T. Kunitake
T.J. Meyer-H.W. Roesky-J.-P. Sauvage
A. Simon • F. Wudl
Springer
Berlin
Heidelberg
New York
Barcelona
Hong Kong
London
Milan
Paris
Tokyo
High Performance
Non-Oxide Ceramics
Volume Editor: M. Jansen
With contributions by
R. Haubner, M. Herrmann, B. Lux, G. Petzow,
R. Weissenbacher, M. Wilhelm
Springer
The series Structure and Bonding publishes critical reviews on topics of research
concerned with chemical structure and bonding. The scope of the series spans the
entire Periodic Table. It focuses attention on new and developing areas of modern
structural and theoretical chemistry such as nanostructures, molecular electronics,
designed molecular solids, surfaces, metal clusters and supramolecular structures.
Physical and spectroscopic techniques used to determine, examine and model
structures fall within the purview of Structure and Bonding to the extent that the
focus is on the scientific results obtained and not on specialist information
concerning the techniques themselves. Issues associated with the development of
bonding models and generalizations that illuminate the reactivity pathways and
rates of chemical processes are also relevant.
As a rule, contributions are specially commissioned. The editors and publishers
will, however, always be pleased to receive suggestions and supplementary
information. Papers are accepted for Structure and Bonding in English.
In references Structure and Bonding is abbreviated Struct Bond
and is cited as a journal.
Springer WWW home page: http://www.springer.de

Visit the SB home page at http://link.springer.de/series/sb/ or
http://link.springer-ny.com/series/sb/
ISSN 0081-5993
ISBN 3-540-43132-2
Springer-Verlag Berlin Heidelberg New York
CIP Data applied for
This work is subject to copyright. All rights are reserved, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting,
reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in
any other way, and storage in data banks. Duplication of this publication or parts
thereof is permitted only under the provisions of the German Copyright Law of
September 9, 1965, in its current version, and permission for use must always be
obtained from Springer-Verlag. Violations are liable for prosecution under the
German Copyright Law.
Springer-Verlag Berlin Heidelberg New York a member of BertelsmannSpringer

Science + Business Media GmbH
http://www.springer.de
© Springer-Verlag Berlin Heidelberg 2002
Printed in Germany
The use of registered names, trademarks, 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.
Typesetting: Scientific Publishing Services (P) Ltd, Madras

Production editor: Christiane Messerschmidt, Rheinau
Cover: Medio Technologies AG, Berlin
Printed on acid-free paper SPIN: 108 57239 02/3020 - 5 4 3 2 1 0
Managing Editor Volume Editor
Professor D. Michael P. Mingos Professor Dr. Martin Jansen

Principal Max-Planck-Institut fur Festkorperforschung
St. Edmund Hall Heisenbergstrafie 1
Oxford 0X1 4AR, UK 70569 Stuttgart
E-mail: [email protected] Germany
E-mail: [email protected]
Editorial Board
Prof. Allen J. Bard Prof. Ian G. Dance

Department of Chemistry and Biochemistry Department of Inorganic and Nuclear Chemistry
University of Texas School of Chemistry
24th Street and Speedway University of New South Wales
Austin, Texas 78712, USA Sydney, NSW 2052, Australia
E-mail: [email protected] E-mail: [email protected]
Prof. Peter Day, FRS Prof. James A. Ibers

Director and Fullerian Professor of Chemistry Department of Chemistry
The Royal Institution of Great Britain North Western University
21 Albemarle Street 2145 Sheridan Road
London WIX 4BS, UK Evanston, Illinois 60208-3113, USA
Prof. Toyohi Kunitake Prof. Thomas J. Meyer
Faculty of Engineering: Associate Laboratory Director for Strategic and
Department of Organic Synthesis Supporting Research
Kyushu University Los Alamos National Laboratory
Hakozaki 6-10-1, Higashi-ku PO Box 1663
Fukuoka 812, Japan Mail Stop A 127
E-mail: [email protected] Los Alamos, NM 87545, USA
Prof. Jean-Pierre Sauvage
Faculte de Chimie Laboratoires de Chimie Prof. Herbert W. Roesky
Organo-Minerale Institut fur Anorganische Chemie
Universite Louis Pasteur der Universitat Gottingen
4, rue Blaise Pascal TammannstraCe 4
67070 Strasbourg Cedex, France D-37077 Gottingen, Germany
Prof. Fred Wudl Prof. Arndt Simon
Department of Chemistry Max-Planck-Institut fur
University of California Festkorperforschung
LosAngeles, CA 90024-1569, USA Heisenbergstrafie 1
E-mail: [email protected] 70569 Stuttgart, Germany
Structure and Bonding
Also Available Electronically
For all customers with a standing order for Structure and Bonding we offer the
electronic form via LINK free of charge. Please contact your librarian who can
receive a password for free access to the full articles by registration at:
http://link.springer.de/orders/index.htm
If you do not have a standing order you can nevertheless browse through the
table of contents of the volumes and the abstracts of each article at:
http://link.springer.de/series/sb/
http://link.springer-ny.com/series/sb/
There you will also find information about the
- Editorial Board
- Aims and Scope
- Instructions for Authors
Preface
The nitrides and carbides of boron and silicon are proving to be an excellent
choice when selecting materials for the design of devices that are to be
employed under particularly demanding environmental and thermal condi-
tions. The high degree of cross-linking, due to the preferred coordination
numbers of the predominantly covalently bonded constituents equalling or
exceeding three, lends these non-oxidic ceramics a high kinetic stability, and is
regarded as the microscopic origin of their impressive thermal and mechanical
durability. Thus it does not come as a surprise that the chemistry, the physical
properties and the engineering of the corresponding binary, ternary, and even
quaternary compounds have been the subject of intensive and sustained efforts
in research and development.
In the five reviews presented in the volumes 101 and 102 of "Structure and
Bonding" an attempt has been made to cover both the essential and the most
recent advances achieved in this particular field of materials research. The
scope of the individual contributions is such as to address both graduate
students, specializing in ceramic materials, and all scientists in academia or
industry dealing with materials research and development. Each review
provides, in its introductory part, the chemical, physical and, to some extent,
historical background of the respective material, and then focuses on the most
relevant and the most recent achievements.
Since the degree of maturity reached by the materials considered is rather
varied, the focus of the respective reviews is also quite different. Thus for SiC
and Si3N4, the main emphasis is placed on processing and shaping, while for
BN its transformation to the cubic polymorph is a major concern, and, finally,
the report on the still rather young class of amorphous Si/B/N/C ceramics is
mainly devoted to aspects related to chemical syntheses and basic character-
izations. Inspite of the fact that many phenomena are dominated by kinetic
control, knowing the underlying thermodynamic equilibria is a crucial
prerequisite to any deeper understanding of the nitride and carbide based
materials discussed here. Therefore, a comprehensive and critically assessed
compilation of thermodynamic data and phase equilibria for the quaternary
system Si/B/N/C as well as its ternary and binary sub-systems has been
included as an introductory chapter preceding the reviews devoted to specific
materials and their properties.
Stuttgart, April 2002 Martin Jansen
Contents
Boron Nitrides - Properties, Synthesis and Applications

R. Haubner, M. Wilhelm, R. Weissenbacher, B. Lux 1
Silicon Nitride Ceramics
G. Petzow, M. Herrmann 47
Author Index Volumes 101-102 169
Subject Index 171

Contents of Vol. 101
High Performance Non-Oxide Ceramics I

Volume Editor: M. Jansen
Phase Equilibria in the Si-B-C-N System

H.J. Seifert, F. Aldinger
Silicon Carbide - A Survey of Synthetic Approaches, Properties

and Applications
G. Roewer, U. Herzog, K. Trommer, E. Miiller, S. Friihauf
Amorphous Multinary Ceramics in the Si-B-N-C System

M. Jansen, B. Jaschke, T. Jaschke
Boron Nitrides ± Properties, Synthesis and Applications
R. Haubner1, M. Wilhelm2, R. Weissenbacher2, B. Lux2
1
e-mail: [email protected]
Tel.: +43 1 58801 16128, Fax: +43 1 58801 16199
2
Institute for Chemical Technology of Inorganic Materials, University of Technology
Vienna, Getreidemarkt 9/161, 1060 Vienna, Austria
Boron nitride is a extraordinary topic in the area of materials science. Due to the special
bonding behaviors of boron and nitrogen the BN exists in many different structures. The
well-de®ned crystallographic structures are hexagonal BN (h-BN), rhombohedral BN (r-BN),
wurtzitic BN (w-BN), and cubic BN (c-BN). Additionally, other crystalline and amorphous
structures exist. Exceptional is that there are still discussions about the BN phase diagram.
In the present stage c-BN is the stable phase at standard conditions but exact data about the
phase transition line are not yet available. Synthesis of h-BN powders and coatings is
described as well as applications of BN in ceramic materials and as lubricant. For c-BN the
high-pressure high-temperature synthesis for powder production is discussed, and an
overview about applications in wear resistant ceramics (polycrystalline c-BN) is given. The
low-pressure methods for nano-cBN deposition (PVD and Plasma CVD) are described.
Keywords: Boron nitride, Hexagonal-BN, Cubic-BN, High-pressure high-temperature

synthesis, Chemical vapor deposition
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Properties of the Various BN Phases . . . . . . . . . . . . . . . . . . 5

2.1 The BN Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.1 Hexagonal Boron Nitride (h-BN) . . . . . . . . . . . . . . . . . . . . . . 6
2.1.2 Cubic Boron Nitride (c-BN) and Wurtzite-BN (w-BN) . . . . . . 7
2.1.3 Rhombohedral and Turbostratic Structure . . . . . . . . . . . . . . . 8
2.1.4 Amorphous Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.5 E-BN and i-BN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Conversion of c-BN into h-BN . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1 Reaction at Standard Pressure . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1.1 Solid State Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.1.2 Gas Phase Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.2 Transformation at Increased Pressure . . . . . . . . . . . . . . . . . . 10
2.3 The BN Phase Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Characterization of BN Products . . . . . . . . . . . . . . . . . . . . . . 12
3 Hexagonal Boron Nitride (h-BN) . . . . . . . . . . . . . . . . . . . . . 14

3.1 Synthesis Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Structure and Bonding, Vol. 102

Ó Springer-Verlag Berlin Heidelberg 2002
2 R. Haubner á M. Wilhelm á R. Weissenbacher á B. Lux
3.1.1 Boric Acid with Carrier Substances . . . . . . . . . . . . . . . . . . . . 15

3.1.2 Boric Acid with Organic Nitrogen Compounds . . . . . . . . . . . 15
3.1.3 Various BN Synthesis Methods . . . . . . . . . . . . . . . . . . . . . . . 15
3.1.4 Gas Phase Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2 Applications of h-BN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.1 Lubricants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.1.1 Liquid Lubricants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.1.2 Solid Lubricants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2.2 Ceramics Containing h-BN . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2.2.1 HP-BN and HIP-BN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2.2.2 Nitride Ceramics with BN . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2.2.3 Mixed Nitride ± Oxide Ceramics with BN . . . . . . . . . . . . . . . 19
3.2.2.4 Oxide Ceramics with BN . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.2.5 Borides and Carbides with BN . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.3 Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.4 Pyrolytic BN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.5 Electronic/Electrochemical Applications . . . . . . . . . . . . . . . . . 22
3.2.6 Filler Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.7 BN Fibers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4 c-BN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.1 High-Pressure High-Temperature Synthesis . . . . . . . . . . . . . . 23
4.1.1 Direct Conversion of h-BN into c-BN . . . . . . . . . . . . . . . . . . 23
4.1.2 Catalytic Conversion from h-BN into c-BN . . . . . . . . . . . . . . 24
4.1.2.1 Alkaline and Alkaline Earth Elements . . . . . . . . . . . . . . . . . . 24
4.1.2.2 Alkaline and Alkaline Earth Nitrides . . . . . . . . . . . . . . . . . . . 25
4.1.2.3 Alkaline and Alkaline Earth Fluoronitrides . . . . . . . . . . . . . . 25
4.1.2.4 Water and Ammonium Compounds . . . . . . . . . . . . . . . . . . . 26
4.1.2.5 Hydrazine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.1.3 Temperature Gradient Method . . . . . . . . . . . . . . . . . . . . . . . 26
4.1.4 Dynamic High Pressure Conversion . . . . . . . . . . . . . . . . . . . 27
4.2 Low-Pressure Synthesis of c-BN . . . . . . . . . . . . . . . . . . . . . . 27
4.2.1 Selective Etching of h-BN and c-BN . . . . . . . . . . . . . . . . . . . 28
4.2.2 Mechanism for Ion-Assisted c-BN Deposition . . . . . . . . . . . . 28
4.2.2.1 Compressive Stress-Model . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2.2.2 Sub-Plantation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2.2.3 Sputter-Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2.2.4 Momentum Transfer Model . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2.3 PVD Methods for Nano-cBN Deposition . . . . . . . . . . . . . . . . 30
4.2.3.1 Ion-Beam-Assisted Deposition (IBAD) . . . . . . . . . . . . . . . . . . 30
4.2.3.2 Mass Selected Ion Beam Deposition (IBD) . . . . . . . . . . . . . . . 30
4.2.3.3 Ion Plating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.3.4 Reactive Sputtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.3.5 Laser Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.4 Plasma CVD Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.4.1 ECR Plasma CVD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Boron Nitrides ± Properties, Synthesis and Applications 3
4.2.4.2 ICP CVD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

4.2.4.3 Bias Enhanced Plasma CVD . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.2.5 Properties and Applications of Nano-cBN Coatings . . . . . . . . 33
4.2.6 The Simple Chemical Way? . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.3 Applications for c-BN Products . . . . . . . . . . . . . . . . . . . . . . . 34
4.3.1 Pure Polycrystalline c-BN (PcBN) . . . . . . . . . . . . . . . . . . . . . 34
4.3.2 Polycrystalline c-BN (PcBN) for Wear Applications . . . . . . . . 34
4.3.2.1 Dense PcBN Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.3.2.1.1 Ceramic Binder for c-BN . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.3.2.1.2 Metallic Binders for PcBN . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.3.2.1.3 Hardmetal/PcBN Compacts . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.3.2.2 Porous PcBN Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.3.2.2.1 Resin Bonded Grinding Tools . . . . . . . . . . . . . . . . . . . . . . . . 37
4.3.2.2.2 Ceramic Bonded . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.3.2.2.3 Metal Bonded and Electrodeposited Grindstones . . . . . . . . . . 37
4.3.3 Electronic Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
List of Abbreviations and Symbols
a-BN amorphous BN
BN boron nitride
c-BN cubic BN
CVD chemical vapor deposition
DG free enthalpy
DH enthalpy
DS entropy
d.c. direct current
DTA differential thermo analysis
E-BN BN synthesized by explosion
ECR electron cyclotron resonance
FTIR Fourier transformation infrared spectroscopy
g-BN graphitic BN
h-BN hexagonal BN
HIP hot-isostatic press
HP hot-pressing
HP-HT high-pressure high-temperature
IBAD ion-beam-assisted deposition
IBD ion beam deposition
i-BN BN prepared by energetic ions
ICP inductively coupled plasma
IR infrared spectroscopy
LED light emitting diode
LO longitudinal optical mode
MISFET metal insulator semiconductor ®eld effect transistor

nano-cBN nano-crystalline c-BN
PACVD plasma assisted CVD
PcBN polycrystalline c-BN
PVD physical vapor deposition
r-BN rhombohedral BN
Ref. reference
RF radio-frequency
SEM scanning electron microscope
t-BN turbostratic BN
TO transverse optical mode
w-BN wurtzitic BN
XRD X-ray diffraction
1
Introduction
Boron nitride (BN) ± in all its various structures ± is a synthetic product and
not found in nature. The ®rst synthesis for h-BN was described in 1842 by
Balmain [1], but h-BN became a commercial material about 100 years later.
Boron and nitrogen are neighbors of carbon in the periodic table, and
therefore BN phases are isoelectric to the corresponding carbon phases. The
structure similarities between h-BN and graphite have been well known, and
therefore in 1957 Wentorf [2] successfully tried the high-temperature high-
pressure synthesis of c-BN analogous to the diamond synthesis. The second
hardest material ± after diamond ± was born, and since 1969 c-BN has been
commercially available. Today c-BN is synthesized by various companies like
General Electric (USA), De Beers (South Africa), Sumitomo and Showa Denko
(Japan), and various companies in Russia.
Parts prepared of h-BN as well as c-BN are of great interest for industrial
applications but also for materials science. The thermodynamic data for c-BN
and the BN-phase diagrams found in literature are not in agreement. After the
®rst high pressure experiments the B-N phase diagram was designed, and after
some modi®cations c-BN was described as metastable phase at room
temperature. Contrary to this opinion in 1988 it was reported that c-BN is
the stable phase. Many experiments have con®rmed this result, but exact
thermodynamic data are still not available.
During the last few years the low pressure c-BN synthesis has been the most
important topic in this ®eld. Using PVD methods it has been possible to
nucleate c-BN on various substrates, but only growth of nm-sized c-BN
crystals is possible at the moment.
Due to its excellent properties h-BN is mainly used as ceramic material, as
lubricant and serves also as thin coatings for electronic devices.
Utilization of cubic-BN are wear applications like machining tools and
polishing powders.
2
Properties of the Various BN Phases
In this section we give a short review of various BN phases and their properties.
2.1
The BN Phases
Similar to the carbon system, BN exists in a soft hexagonal (h-BN)

modi®cation, a hard cubic (c-BN) one, and many others which are not very
well crystallized, or amorphous. The properties of h-BN and c-BN are
summarized in Table 1 [2±17], and the crystal structures of c-BN, w-BN
(wurtzitic-BN), and h-BN are illustrated in Fig. 1.
Table 1. Properties of h-BN and c-BN
Property Hexagonal BN Cubic BN
Crystallographic data
Crystal structure Hexagonal Cubic, zinc blende
Space group P63/mmc [3] Fd3 m (8 atoms/unit cell)
Lattice constant Ê [4]
a = 2.504 A a = 3.615 A Ê [2]; a = 3.67 A
Ê [6]
Ê [4]
c = 6.661 A acalc = 3.606 AÊ [7]
B±N ion distance Ê [4]
1.446 A 1.57 AÊ [8]
Density 2.34 g/cm3 [3] 3.4879 0.003 g/cm3 [9]
3.45 g/cm3 [2]
Mechanical properties
Hardness 1.5)1.3 GPa (Vickers) 58)76 GPa (Knoop) [11]
[10] (hot-pressed) 4500 kg/mm2 [12]
Young's modulus 3400±8700 kg/mm2 [13]
Optical and electrical properties
Color White, gray Colorless; B excess changes to
yellow, orange, black [8]
Electrical resistivity a/b axis 3.0 ´ 107 W cm [4] 1010 W á m (289 K) [8]
c axis 3.0 ´ 109 W cm [4] 107 W á m (773 K) [8]
1013 W á m [14]
3.3 ´ 1013 W cm [15]
Bandgap - 3.67 eV [6]
6.4 0.5 eV [16]
5.0 eV [17]
Refraction index - 2.117 (at 589.3 nm) [12]
1.5)1.6 [17]
Maximum 2.295 [15]
Thermal properties
Thermal conductivity a/b axis 0.627 W/cm á K [4] 13 W/cm á K [12]
c axis 0.015 W/cm á K [4]
Debye temperature 1700 K [5]
Linear thermal a/b axis 3.24 ´ 10)1 K)1 [4] 4.80 ´ 10)1 K)1 (700 K) [8]
expansion c axis 81 ´ 10)1 K)1 [4] 5.60 ´ 10)1 K)1 (1170 K) [8]
(pyrolytic BN) 5.80 ´ 10)1 K)1 (1430 K) [8]
Fig. 1. Crystal structures of c-BN, w-BN, and h-BN
2.1.1
Hexagonal Boron Nitride (h-BN)
Hexagonal BN is mostly named h-BN, but also a-BN or g-BN (graphitic-BN)

are used. It crystallizes similar to graphite in a hexagonal sheet layered
structure, and therefore it is often referred to as ``white graphite'' (Fig. 1). The
atomic planes are built by hexagonal rings formed by B and N atoms. The
covalent bonds (r-bonding, sp2-hybridization) between the atoms forming the
rings are very strong. Between the atomic planes the bonding forces are weak,
being van der Waals bonding (p-bonding). Additionally, it has to be pointed
out that the planes are stacked on top of one another, without any horizontal
displacement (boron and nitrogen are alternating along the c-axis). Due to the
higher electronegativity of nitrogen the p-electron is located at the nitrogen
and therefore h-BN is an electrical insulator and its color is white.
The morphologies of typical h-BN powders are shown in Fig. 2.
Fig. 2. Pictures of h-BN powders. (Courtesy of ESK-Kempten, a company of WACKER-

Chemie)
2.1.2
Cubic Boron Nitride (c-BN) and Wurtzite-BN (w-BN)
Cubic boron nitride is commonly called c-BN in literature, but also z-BN (zinc
blende) or b-BN [13] can be found. Wentorf [18] named c-BN ``Borazon'',
which has become the trade name for the products of the General Electric
Corporation. Russian companies call abrasive powders of c-BN `Èlbor'' or
``Cubonite'' [8].
Similar to the diamond lattice the B and N atoms are tetrahedrally
coordinated. Every boron atom is surrounded by four nitrogen atoms and vice
versa. In this arrangement boron and nitrogen atoms have sp3 hybridization.
Because of the special bonding conditions (short bonding length) c-BN and
diamond exhibit high hardness. Both materials are insulators because of
missing p-bonds. The high thermal conductivity is caused by phonons and not
by electrons like in metals.
Typical images of c-BN grown by the high-pressure high-temperature
method are shown in Fig. 3.
Boron nitride can also form a superhard hexagonal phase in wurtzite-type
(w-BN). This modi®cation is a high pressure phase and was described ®rst by
Bundy and Wentorf [19].
Fig. 3. Various commercially available c-BN powders with different grain size and
morphology
Lattice parameters for w-BN are: a = 2.55 0.01 A Ê ; c = 4.20 0.01 A

Ê ; and
3
d = 3.49 0.03 g/cm [19].
Similar data are found in later references [20, 21].
2.1.3
Rhombohedral and Turbostratic Structure
The rhombohedral (r-BN) structure is similar to the h-BN phase but the
atomic layers sequence is ABC ABC. It was reported that r-BN is formed
during conversion of c-BN into h-BN [22] (Fig. 4).
Turbostratic structure is characterized by a layer structure similar to h-BN;
the layers are mostly parallel but not aligned to the c-axis [10] (Fig. 4).
2.1.4
Amorphous Structure
Amorphous boron nitride (a-BN) can be synthesized by decomposition of

B-trichloroborazine and cesium [23]. Such layers are applied during produc-
tion of semiconductor devices [24].
2.1.5
E-BN and i-BN
E-BN (E = explosion) is described as high pressure phase by a few scientists.

For synthesis shock wave methods [25, 26] were used and also reactions at
normal pressure with photon [27] or electron [28, 29] assistance. In a special
three-dimensional phase-diagram (pressure, temperature, electrical ®eld) the
existence of the metastable E-BN was described [30].
An X-ray spectroscopic study of crystalline E-BN in combination with a
detailed literature review described E-BN as an oxygen containing compound
of the type BN1±xOx [31].
The expression i-BN is not for a special modi®cation of BN. It is used by
some authors to describe an h-BN with boron excess and is produced by a
Fig. 4. Crystal structures of h-BN, r-BN, and t-BN

method where the activation energy is supplied by energetic ions [32, 33]. A
broad IR peak at 1400 cm)1 is typically for i-BN. The hardness of i-BN is given
as 26.9 Gpa [33]. Also, boron nitride layers synthesized by ion-bombardment
are called i-BN [e.g. 34±36].
2.2
Conversion of c-BN into h-BN
2.2.1
Reaction at Standard Pressure
The transformation of c-BN into h-BN was investigated during DTA-analysis

[22]. The SEM images show formation of differently textured h-BN at the
surface of the c-BN crystallites, which indicates that there is no uniform
mechanism of phase transition (Fig. 5).
The DTA measurements exhibit the stability of c-BN at standard conditions.
In¯uences of grain size and purity (oxide content) of the cubic boron nitride
crystals on the conversion temperature become obvious. Fine grained samples
containing boron oxide show a signi®cantly lower conversion temperature
than coarse material (conversion at 900 °C for 1.5-lm c-BN containing boron
oxide and 1500 °C for 600-lm pure c-BN).
This investigation also showed that two different routes are possible for the
c-BN ® h-BN transformation at normal pressure and elevated temperatures:
± Phase conversion of the solid bulk material, leading to layered textures of
h-BN on c-BN crystals.
Fig. 5. Mechanisms of the c-BN to h-BN conversion (CVD mechanism and solid state
mechanism) [22]
± A conversion via gas phase transport. The BN vapor pressure raises with
temperature and beyond 1000 °C evaporation of (BN)x species and
deposition of crystalline h-BN and BN whiskers occurs.
2.2.1.1
Solid State Mechanism
The atomic layers from c-BN (ABCABC) have to rearrange into an ABAB
stacking sequence of h-BN during the solid state phase conversion. A possible
mechanism would be the intermediate formation of the rhombohedral BN
phase (r-BN) with ABCABC stacking. The r-BN phase is structurally related to
Ê ; r-BN
the hexagonal phase, but only differs in the d-values (h-BN: d = 6.66 A
d = 10.0 AÊ ) of the layers (Fig. 5b). Subsequently the rhombohedral phase is
transformed into the hexagonal modi®cation at the reaction temperatures [10].
2.2.1.2
Gas Phase Mechanism
For gas phase reactions volatile compounds are necessary. The vapor pressure
of h-BN (1 mbar at 1300 °C, 14 mbar at 1600 °C [37, 38]) is high enough for
the reaction. The evaporation from the c-BN surface can be explained by
formation of distinct sp2-bound BN regions (atomic layers), which may lead to
sublimation of (BN)x species in various degrees of aggregation. Theoretical
investigations show that the c-BN crystals consist of a sp2 terminated surface
which reveals the CVD mechanism [39]. Grain size, crystallinity, and
impurities in¯uence the BN evaporation and thus the wide temperature range
for the phase conversion, (more than 1100 K measured by DTA experiments
[22]) can be explained. At high heating rates the onset of the conversion
temperature seems to be higher, because the phase transition is hindered by
kinetic activation.
As demonstrated in SEM pictures (Fig. 5), the phase conversion starts at the
surface of the c-BN crystals, which leads to the formation of h-BN platelets and
whiskers (Fig. 5). This clearly shows that a CVD mechanism is involved and
the solid state mechanism is not the only reaction pathway [22].
2.2.2
Transformation at Increased Pressure
The transformation of h-BN into c-BN (at 6.5 GPa) and the reverse
transformation of c-BN to h-BN (from 0.6 to 2.1 GPa) were investigated in a
Li3N-BN catalyst system [40]. Synchrotron radiation was used to check the
phases and to examine reactions between the BN-phases and the catalyst.
The reverse transformation was investigated in a temperature range from
800 to 1200 °C. The reaction started at the same time at which a melt was
formed. The experiments described showed that the transformation from c-BN
to h-BN took place within minutes. These experiments revealed that c-BN is
the stable phase at standard conditions in the presence of a Li-catalyst [41].
2.3
The BN Phase Diagram
In 1963 a phase diagram was established by Bundy and Wentorf [19], based on
data of Wentorf [42] and experiments carried out at pressures higher than
4 GPa. This phase diagram described c-BN as the stable phase at standard
temperature and pressure (Fig. 6). In 1975 a new phase diagram was published
by Corrigan and Bundy [11], showing the c-BN/h-BN equilibrium line similar
to the graphite/diamond line in the carbon system.
The phase diagram of Corrigan and Bundy had been considered to be
correct until 1987 when Leonidov et al. [43] published ¯uoro-calorimetric
results for burning c-BN, and additional calculations from Solozhenko and
Leonidov [44] followed in 1988. These papers described c-BN as the stable
phase ± up to 1300 °C. A comparison of the data for burning c-BN and h-BN
con®rm the c-BN stability:
h-BNDHr 298:15 K 884:91 1:15 kJ/mol 45
c-BNDHr 298:15 K 869:2 2:0 kJ/mol 43

Further results were reported by Maki et al. in 1991 [46] and by Solozhenko
in 1993 [47]. However, the thermodynamic data show large discrepancies
(Table 2 [44, 47], Table 3 [43, 45, 46, 48±50]) and therefore the differences in
the phase diagrams are easy to explain.
It can be summarized that c-BN is the stable phase at room temperature but
there are still discrepancies about the phase transformation lines and
thermodynamic data. Several review articles draw conclusions from the
available results (Fig. 7) [11, 41, 51±54] but there are also new results
describing c-BN as metastable at standard conditions [55].
Fig. 6. BN phase diagrams described by Bundy and Wentorf [19] and by Corrigan and
Bundy [11]
Table 2. Thermodynamic data for the conversion of h-BN into c-BN at 298 K (25 °C)
[44] [47]
Enthalpy DHr° (298.15 K) [kJ/mol] )16.3 2.7 )16.5

Entropy DSr° (298.15 K) [J/mol á K] )8.22 0.01 )8.56
Free enthalpy DGr° (298.15 K) [kJ/mol] )13.9 0.17 )3.6
Table 3. Standard free energy of formation for h-BN and c-BN at 298 K (25 °C)
Hf°298 K [kJ/mol]
Reference [48] [45] [49] [46] [50]

h-BN )250.914 1.548 )251.0 1.5 )252.8 )254.0 )254.4
Reference [43] [46]
c-BN )266.8 2.2 )266.1
Fig. 7. Summary of published results about the location of the c-BN/h-BN phase boundary.
A ± Bundy and Wentorf (1963) [19]; B ± Corrigan and Bundy (1975) [11]; C ± Maki et al.
(1991) [46]; D ± Solozhenko (1994) [51]; E ± Will et al. (2000) [41]; F ± Will et al. (2000)
calculated from Solozhenko's data [41]; G ± Fukunage (2000) [55]
2.4
Characterization of BN Products
Due to the complexity of BN-structures and atomic bonding situations, the

characterization of BN-phases by spectroscopic methods (e.g., IR and Raman)
is dif®cult. It is not possible to identify BN phases using only one analytical
method. For example, the X-ray diffraction peaks of c-BN correspond to those
of Cu, Ni, and many other cubic phases. Elemental composition must be
known or measured to be sure that no other phases are present.
Fig. 8. IR- and Raman spectra and X-ray diffractogram of c-BN. Peak positions of other
compounds in the relevant region are marked
Some data for analytical characterization are summarized in the following

sections (Fig. 8).
Infrared spectroscopy (IR) or Fourier transform infrared spectroscopy (FTIR)
are often used to characterize BN products. If pure BN mixtures with B:N ratio
of 1:1 are analyzed, it will be easy to distinguish between h-BN and c-BN.
However, if the chemical composition of the sample is unknown, many
artifacts can occur and a clear statement is often not possible.
To identify c-BN, the characteristic transverse optical mode (TO) at
1065 cm)1 and longitudinal optical mode (LO) at 1340 cm)1 have been
described [56]. When investigating commercial c-BN, commonly only one
IR-peak between 1050 and 1100 cm)1 is observed.
The situation is more complex for h-BN, t-BN, and a-BN, because all of
them show peaks between 780 and 1370 cm)1, which makes it impossible to
distinguish among these phases.
Table 4 summarizes spectroscopic data for BN phases prepared by different
methods [36, 56±65].
Raman spectroscopy is useful to distinguish between h-BN and c-BN, too. As
described above, impurities as well as non-stoichiometric mixtures can result
in misinterpretations.
In the case of c-BN, two characteristic peaks are observed, and only one
peak for h-BN.
The two peaks for c-BN are described by several authors as a TO-mode at
1055±1057 cm)1 and an LO-mode at 1305±1306 cm)1 respectively [66±68].
The crystallinity seems to be important for the Raman peaks of c-BN, because
Table 4. IR-peaks of c-BN and h-BN samples synthesized by different methods
c-BN h-BN Reference Method
TO LO
± 810 1310 [57] CVD

1065 [56] HP-HT synthesis
1050 800 1400 [58] Laser PVD
1050 800 1390 [36] Ion-plating
1060 780±800 1380 [59] ECR PACVD
1060 800 1370 [60] RF sputtering
1060±1080 780 1390 [61] PACVD
1065 770 1380 [62] Sputtering
1100 770 1370 [63] Pulsed laser deposition
1100 800 1380 [64] PACVD
1110 800 1364 [65] ECR PACVD
in the case of very ®ne grained c-BN (found mainly in PVD layers) these peaks
have not been observed [36].
For h-BN the characteristic peak is located at 1367 cm)1 [67].
X-ray diffraction is another possibility to distinguish between c-BN and
h-BN. In that case again the elemental composition of the sample is important,
because the diffractograms of many cubic substances can mimic the one of
c-BN (e.g. Cu, Ni, etc.). The peak positions and their intensities are mainly
in¯uenced by grain size, stress in the layers and impurities.
Summing up, it can be said that for the characterization of c-BN in
unknown samples (mainly PVD and Plasma-CVD deposits), the results of only
one analytical method are not suf®cient for a de®nitive characterization.
Measurements of the elemental composition in combination with IR and/or
Raman and/or X-ray diffraction are necessary to ensure that c-BN is present.
3
Hexagonal Boron Nitride (h-BN)
Hexagonal BN is a high-temperature solid lubricant, good thermal conductor,
and good electric insulator. The speci®c gravity is low, it is stable in air up to
1000 °C, under vacuum up to 1400 °C, and in inert atmosphere it can be used
up to 2800 °C. Hence, the maximum application temperature is higher than
that of Si3N4, Al2O3, or SiC. The temperature resistance of BN can be
compared to MgO, ZrO2, or CaO but BN shows higher thermal shock
resistance than these oxides. BN is chemically inert and not wetted by many
metallic (Al, Cu, Zn, Fe, steel, Ge) and non-metallic (Si, B, glass, cryolite,
halides) melts. The hardness is similar to graphite and therefore BN materials,
produced by hot-pressing, can be machined easily for low costs. Close
tolerances for h-BN components can be reached using conventional shop
tools. The thermal and mechanical properties show an anisotropic behavior
parallel and perpendicular to the hot-press direction. Detailed information
about the chemical and physical properties of h-BN is given in [69, 70].
This combination of excellent properties of hexagonal BN (h-BN) opens a

huge range of technical applications.
3.1
Synthesis Methods
Balmain ®rst synthesized BN in 1842 by reaction of molten boric acid with

potassium cyanide [1]. More than 100 years later the commercial production
of BN was established. Although there are a lot of other general methods for
producing BN, principally two reactions are used on the industrial scale.
3.1.1
Boric Acid with Carrier Substances
Boric acid with ammonia reacts in the presence of carrier substances

(Ca3(PO4)2, CaCO3, CaO, BN, Zn-borate) [71±75]. The carrier substances
prevent the formation of a homogeneous melt of boric acid, which is not
suitable because of its minimal surface: At reaction temperatures exceeding
700 °C a thin ®lm of molten boric acid covers each carrier substance particle.
Because of the large surface a full reaction of the boric compound with
ammonia is possible. After the reaction the carrier is leached with HCl and the
remaining h-BN is washed with water. A second reaction at temperatures
exceeding 1500 °C with ammonia follows, resulting in h-BN powders with 97%
purity. The h-BN crystallites are thin hexagonal platelets with a thickness of
about 0.1±0.5 lm and a diameter up to 5 lm.
3.1.2
Boric Acid with Organic Nitrogen Compounds
The second important way to produce h-BN is the reaction of boric acid or
alkali-borates with organic nitrogen compounds (melamin, urea, dicyanamide,
guanidine) in nitrogen atmosphere [76±80]. These reactions are carried out at
temperatures between 1000 °C and 2100 °C in N2 atmosphere. Before ®nal
thermal treatment, the product can be washed with methanol or diluted acids
in order to remove all non-reacted products. For removing oxygen impurities
a thermal treatment at 1500 °C in inert N2 or Ar atmosphere is used. BN with
turbostratic structure (t-BN) is obtained which is characterized by partial or
complete absence of three-dimensional order in the stacking of its atomic
planes [81].
Annual world production of h-BN powder is approximately 400 tonnes at a
price of $50±150 per kg.
3.1.3
Various BN Synthesis Methods
± In carbothermal reactions BN is synthesized by reduction of boric acid or

borates in nitrogen atmosphere at 1000±1500 °C. One example is the
reduction of boric acid or borates with carbon. The precursors are mixed
intensively and after drying, the mixture is heated up to 1200±1500 °C in
nitrogen atmosphere, remaining at maximum temperature for 60±600 min.
To enhance the yield of BN, catalysts (CaCO3, MnO2) are sometimes added
to the mixture. However, too high carbon content of the mixture leads to the
formation of boron carbide [82±85].
± Reaction of amides, cyanides, cyanamides, and thiocyanates with boron
compounds. NaNH2 forms boron nitride with boric acid or sodium borates
at 300±500 °C in ammonia atmosphere. At the end of the reaction the
temperature is increased to 1000 °C for the decomposition of the remaining
nitrogenous compounds. After washing with water, the BN powder is
stabilized by annealing at 1800 °C in inert atmosphere [86]. Cyanides
(NaCN, KCN) react with boric acid at 800±1500 °C by forming BN. As a by-
product, CO gas and C are produced [87]. The reaction of cyanamides
(CaNCN) with boric acid is carried out at 1400±1750 °C in a 93% N2 ± 7%
H2 atmosphere. After reaction the product is washed with diluted HCl.
Carbon, which is produced as a by-product during the reaction, is burned in
air at 1000 °C [88].
± Preparation of B-N containing precursors with subsequent decomposition is
also a suitable method to produce BN. The adduct BF3-NH3 is formed by
reaction of BF3 with gaseous, liquid, or aqueous ammonia. From aqueous
solutions the BF3-NH3 adduct can be precipitated by adding NaOH. After
drying the precipitate it is heated up to 800 °C in inert atmosphere (N2, H2,
or NH3), and afterwards the BN is formed by pyrolytic reaction. After
washing with hot water the h-BN is obtained [89, 90].
± Fine and ultra-®ne BN is used for lubricants and toners and can be
produced by combustion of boron powder at 5500 °C in a nitrogen-plasma
[91±93].
± Other ways for preparing BN are the reactions of calcium-boride (CaB6)
with additions of boric acid in nitrogen atmosphere at temperatures
exceeding 1500 °C [94, 95], or the synthesis from iron boride (FeB) with
ammonia at 550 °C and subsequent annealing in ammonia at 1000 °C [96].
± Growing of h-BN crystals is possible in molten sodium at temperatures
between 700 and 800 °C. Starting materials are boron and NaN3 powders
which are sealed in a stainless steel tube and heated up [97].
3.1.4
Gas Phase Deposition
Using thermal CVD methods with B2H6±NH3±H2 [98, 99] or BCl3±NH3±H2

[100] gas mixtures, different BN-layers can be deposited; e.g., h-BN, t-BN, or
a-BN. BN with higher boron contents can be deposited at enhanced deposition
temperatures. To deposit crystalline h-BN from the gas phase, temperatures
above 1100 °C and a N/B ratio of 10:1 are necessary.
In the last century a dramatic increase in the number of reports and patents
describing the deposition of h-BN has taken place. Different methods for
coating on a wide range of materials have been developed. The activation of the
gas mixture can be achieved by various methods, e.g., plasma-assisted CVD as
well as laser ablation techniques of BN bulk materials. The plasma is generated
either by microwave discharge or by an RF (radio-frequency) discharge. The
CVD process can also be conducted by laser driven reactions and pulsed laser
deposition. A wide range of different gaseous precursors is used in all the CVD
techniques. The gas mixture may consist of different boron and nitrogen
sources like B2H6, BCl3, BBr3, and NH3, N2, respectively [101, 102].
Most of the reports on h-BN deposition during the last years are in
combination with the low-pressure nano-cBN deposition by PVD methods
because a h-BN interlayer is formed before the c-BN is able to nucleate.
Layers of pure pyrolytic BN are generally produced by thermally induced
CVD processes on graphite substrates. The reaction is performed using BCl3,
NH3, and N2 at low pressures (0.7±70 mbar) and temperatures in a range
between 1500 °C and 1900 °C. For special applications, e.g., coating ceramic
®bers for reinforcing purposes or nuclear fuel pellets [103], the temperature
is kept below 1100 °C. Vacuum evaporation of B-trichloro-borazine and
decomposition on extremely hot surfaces (e.g., graphite or tungsten) gives
pyrolytic BN layers [104].
3.2
Applications of h-BN
Hexagonal boron nitride is an interesting construction material because of its

high heat resistivity and stability against oxygen. Powders of h-BN are used as
lubricants, for producing ceramic parts, and for coatings; all applications are
suitable up to high temperatures.
3.2.1
Lubricants
Graphitic BN (h-BN) is used as lubricant with low friction in numerous

applications. Compared to graphite the h-BN can be used as lubricant in an
oxidizing atmosphere up to 900 °C as well as at extremely low tempera-
tures, e.g., in space because no water inclusions between the atomic sheet
layers are present (graphite always contains small amounts of water
between the layers). Due to its excellent resistance against oxidation, its
extremely low friction coef®cient, and its chemical inertness, h-BN can be
inserted into alloys or ceramics [105]. It can be used as a solid surface
lubricant [106] or added to a liquid to get dispersions with lubricating
properties.
3.2.1.1
Liquid Lubricants
Lubricating dispersions containing h-BN are mainly water based, oil based
(mineral oil, silicone oils, highly viscose organic components), or water/oil
mixtures [107, 108]. Stabilizers, thickeners, oxidation-preventing additives are

added to the dispersions.
3.2.1.2
Solid Lubricants
To produce solid based lubricants, h-BN is added to plastics, rubbers, and

resins in various amounts. BN dip coated surfaces are used for single-use
application.
BN is used for reinforcing ceramics and alloys to reduce wear and friction
and thus obtain self-lubricating parts [109±111].
Lubricant materials containing BN are used as bearing materials for high
temperature applications and as sliding contact materials made of alloys or
porous ceramics ®lled with Cu, Ag, Pb, or graphite [112±114]. BN powder
serves as an additive in paints, and it acts as a lubricant agent for casting and
forming processes.
BN powder is used to produce lubricants for high pressure purposes
[115±117], for polishing materials, as well as for HIP (hot-isostatic pressing)
techniques to prevent reactions between the powder mixture and the mold.
3.2.2
Ceramics Containing h-BN
3.2.2.1
HP-BN and HIP-BN
Dense shapes of h-BN are made exclusively by hot-pressing (HP) or hot-

isostatic pressing (HIP) of BN powders using boric oxide as a sintering
additive [118, 119]. The BN powders should have a ®ne grain size, a free boron
oxide content as sintering aid, and disordered lattice or turbostratic structure.
The resulting shapes are soft and can easily be machined to the desired size but
the big waste of material during shaping makes the products expensive. Hot-
pressed BN ceramics show a signi®cant anisotropy in thermal expansion and
thermal conductivity as well as in strength and Young's modulus, which varies
with the hot-press direction. In this context BN ceramics produced by HIP
technique show isostatic behavior and nearly theoretical density.
HP-BN is an excellent thermal conductor and an electrical insulator, which
makes it ideal for electronic applications. As an insulator it has a dielectric
constant of about 4 and a dielectric strength almost four times higher than that
of alumina. As a thermal conductor BN exceeds almost all other electrical
insulators while maintaining high strength and low thermal expansion. This is
an ideal combination for heat sinks and substrates. Compared to other
favorable materials, e.g., BeO or Al2O3, BN is easier to form, and a smooth
surface is achieved without high cost for ®nishing. Pure BN ceramics are used
as break rings in continuous casting of steel or in the non-iron industry, for
crucibles, tubes, or plates, seal rings for gas sensors, moulds for hot-pressing
of ceramics, neutron absorbers and shields for nuclear reactors and
Fig. 9. Boron nitride ceramics. BN ceramics with SiC addition for break rings in continuous
ferrous metals casting. Hot-pressed and HIPed BN crucibles and electrical insulators.
(Courtesy of ESK-Kempten, a company of WACKER-Chemie)
components for high-temperature electric furnaces (Fig. 9) [120]. Other

applications are insulators and source holders for ion implantation systems,
furnace vents, stacks and ®xtures, welding tips for plasma arcs. Because of its
stability at high temperatures and chemical inertness against carbon and
carbon monoxide up to 1800 °C, BN is used as a refractory ceramic [121].
3.2.2.2
Nitride Ceramics with BN
Due to the good thermal conductivity and electrical insulating properties AlN
ceramics are used in the electronic/electrical industry for integrated circuits
and integrated circuit package materials. BN additions increase the machina-
bility, and they enhance the resistance against erosion by molten metals [122].
The most common way to produce such composites is hot pressing or hot
isostatic pressing at temperatures of 2000 °C and at pressures near 2000 bar
[123]. Sintering aids, which are necessary for densi®cation, can be B2O3, Y2O3,
Al2O3, CaF2, MgO, Si, Al2O3, CaB6, MgBx, and mixtures of them.
An addition of BN powder to Si3N4 ceramics improve the thermal shock
resistance. These composites show excellent stability toward molten steel.
Si3N4-BN composites are produced either by hot-pressing or HIP or via
reaction-sintering, starting with BN powder and elemental silicon and
sintering in nitrogen atmosphere at 1500 °C [124]. These composites are
used to dope silicon wafers, for heat-exchangers, and for nozzles in continuous
steel casting [125, 126].
3.2.2.3
Mixed Nitride ± Oxide Ceramics with BN
SiAlON (Si3N4-Al2O3) ceramics reinforced with BN powder are used as break

rings in continuous casting and for nozzles. The addition of BN provides
better thermal shock resistance [127].
TiN-based composites with additions of Al2O3 and BN show excellent

thermal shock resistance as well as good corrosion resistance. These
composites are stable towards molten metals and are used for abrasives and
cutting tools [128, 129].
3.2.2.4
Oxide Ceramics with BN
Oxide ceramics (Al2O3, ZrO2) can also be reinforced with BN powder. They are
often sintered by hot-pressing with sintering aids like boron oxide and calcium
oxide. The resulting composites show higher thermal shock resistance and
excellent corrosion resistance. Therefore they are used as material for metal
casting or, generally, for materials in contact with molten metals [130]. With
increasing BN content the composites show better machinability, and
therefore the production costs decrease.
Borosilicate glasses, phosphate glasses, glass ceramics, and enamels are
reinforced with BN powder for enhancing the strength of the composites
[131, 132].
3.2.2.5
Borides and Carbides with BN
BN is used for the production of TiB2-BN and TiB2-AlN-BN composites. High

temperature applications are evaporator crucibles and boats used for vacuum
metallization of plastics, paper, textiles, and glass. When the material is soaked
in molten metals it can be used for electrodes. The composites show high
corrosion resistance against molten metals as well as good electrical
conductivity (TiB2 has high electrical conductivity). The crucibles/boats are
heated in direct current, and by varying the BN content the electrical
conductivity of the boats can be controlled (Fig. 10) [133, 134]. However, to
enhance the wettability of the boat-surfaces for molten aluminum, the BN
surface layer is removed by laser sputtering methods. A ®lm of molten
aluminum has lower corrosive potential than molten Al-pearls and therefore
this step is necessary. Today more than 70% of the BN world production is
used for producing TiB2-BN composites.
SiC-BN composites show good thermal conductivity. The electrical
resistivity depends on the BN content and can be adjusted from a few
W/cm to more than 1010 W/cm. When the SiC content of the composites is
high, the composites can be used as cutting tools [135]. BN additions reduce
the friction coef®cient. Thus these composites are used for sliding parts
[136].
3.2.3
Coatings
Coatings of h-BN are used in a wide range of applications due to the

extraordinary properties of BN. BN coatings are used in the steel and iron
Fig. 10. Production sheet for TiB2-BN-AlN evaporation boats. (Courtesy of ESK-Kempten,
a company of WACKER-Chemie)
industry to enhance the corrosion resistance as well as to reduce the wear on

sliding parts like the crankshafts for compressors. Because of the non-
wettability of BN against many metallic and non-metallic melts [137] various
parts are often coated with BN, e.g., silicon in the semiconductor industry, or
BN is used for linings to prevent reactions with Si or Al. Coatings can be
produced by CVD techniques to obtain well-crystallized linings with relatively
high strength. Also BN layers can be achieved by spraying, brushing, dipping,
or pouring [138]. The coatings provide improved high temperature electrical
insulation for semiconductor industry. For mold equipment BN coatings
reduce sticking and reactions between powder and mold. High temperature
lubrication properties are important in steel manufacturing processes like
casting. On BN coated pre-forms, metal layers can be deposited, allowing one
to get the free-standing metal sheets by easily peeling-off [139].
BN coating on SiC ®bers or Al2O3 ®bers reduces the thermal mismatch
between ®ber and matrix and also reduces ®ber-matrix interfacial shear
strength in ceramic matrix composites leading to higher overall strength and
toughness.
3.2.4
Pyrolytic BN
Pyrolytic BN is used in a wide range of applications. Due to its high purity,

pyrolytic BN is used in equipments for semiconductor materials, e.g., for
single-crystal production in the Czochralsky apparatus as well as in horizontal
and vertical Bridgman apparatus. Usually the amount of impurities is low
(typically a few ppm). In the case of growing GaAs single-crystals it is
important to prevent silicon impurities, and therefore instead of using silica
parts, pyrolytic BN coated crucibles are used [140]. Due to the non-reactivity
with the melt and the non-wettability, the pyrolytic BN parts can be used
several times and therefore the costs are low. In the semiconductor industry
pyrolytic BN is used for crucibles and linings, covers of heating elements, in

thermal evaporation of elements to coat semiconductor materials with metals,
e.g., Al2O3, and for coating materials to perform CVD [141].
3.2.5
Electronic/Electrochemical Applications
The excellent insulating and dielectric properties of BN combined with the

high thermal conductivity make this material suitable for a huge variety of
applications in the electronic industry [142]. BN is used as substrate for
semiconductor parts, as windows in microwave apparatus, as insulator layers
for MISFET semiconductors, for optical and magneto-optical recording media,
and for optical disc memories. BN is often used as a boron dopant source for
semiconductors. Electrochemical applications include the use as a carrier
material for catalysts in fuel cells, electrodes in molten salt fuel cells, seals in
batteries, and BN coated membranes in electrolysis cells for manufacture of
rare earth metals [143±145].
3.2.6
Filler Material
Additions of BN powder to epoxies, urethanes, silicones, and other polymers

are ideal for potting compounds. BN increases the thermal conductivity and
reduces thermal expansion and makes the composites electrically insulating
while not abrading delicate electronic parts and interconnections. BN
additions reduce surface and dynamic friction of rubber parts. In epoxy
resins, or generally resins, it is used to adjust the electrical conductivity,
dielectric loss behavior, and thermal conductivity, to create ideal thermal and
electrical behavior of the materials [146].
Ultra-high purity BN powder enhances the performance of cosmetics
products [147]. BN powder is added to pencil lead compositions, paints and
cement in dentistry and medicine.
3.2.7
BN Fibers
BN ®bers can be made by decomposition of hydrated cellulose impregnated

with boric acid or ammonium tetraborate (NH4B4O7) in ammonia and
nitrogen atmosphere at elevated temperatures (>1000 °C) [148, 149].
Another method is to extrude borazine network polymers
([-B(NH2)-N(C6H5)-]3) with addition of B2O3 at T > 300 °C into ®bers and
subsequently ®ring in nitrogen stream at 1800 °C [150].
BN ®bers are used for reinforcing ceramic materials (e.g., Al2O3, Si3N4, SiC)
to enhance mechanical properties as well as to extend the range of possible
applications. They serve as reinforcement of organic polymers (e.g., epoxides,
polyether-polyketones, polyphenylensul®des) which exhibit good thermal
conductivity and low thermal expansion.
4
c-BN
For industrial synthesis of c-BN only the high-pressure high-temperature
(HP-HT) methods are relevant. During the last years the deposition of c-BN
coatings mainly produced by PVD and Plasma-CVD methods has been
investigated. Due to the dif®culties of growing large c-BN crystals and
depositing thick layers, the low-pressure synthesis method is not commercially
used today.
4.1
High-Pressure High-Temperature Synthesis
The high-pressure high-temperature (HP-HT) synthesis was the ®rst method

[2, 42] to grow c-BN, and until now it is the only one available for industrial
production. Investigations in this ®eld have the aim to reduce pressure and
temperature for the process and to ®nd new catalysts to grow larger c-BN
crystals. The fact that c-BN is a stable phase raises hopes that further decrease
in pressure and temperature during synthesis is possible (Figs. 6 and 7).
4.1.1
Direct Conversion of h-BN into c-BN
Bundy and Wentorf [19] showed the direct conversion from h-BN into c-BN
without any catalyst at pressures up to 18 Gpa. At temperatures below 1000 °C
w-BN is formed. Only by increasing the temperature to 1730±3230 °C is c-BN
formed.
Conversion parameters from h-BN to c-BN are strongly in¯uenced by the
properties of the h-BN starting material:
± Pressure and temperature can be decreased using ®ne-grained h-BN
(6.0 Gpa and 1470±1720 K) [151].
± Small-grained h-BN with low crystallinity increases diffusion rates, and
c-BN can grow more easily [152].
± By using amorphous-BN it is possible to reduce the pressure to 7.0 Gpa and
the temperature to 1070 K [153]. Two types of conversion mechanisms have
been described:
a) Direct conversion from a-BN into c-BN
b) Conversion with h-BN as intermediate phase (a-BN ® h-BN ® c-BN)
± With turbostratic-BN the resulting c-BN crystal size is very small, but
conversion conditions of 6.0 GPa and 1250 K are possible [154].
± A further decrease in conversion temperature and pressure is possible by
adding impurities. When small amounts of B2O3 are present in the h-BN
starting material, 4.0±7.0 GPa and 1500 °C are suf®cient to grow c-BN
[9].
For industrial synthesis the direct conversion methods are commonly not
used.
A new approach in this area is the synthesis of high purity polycrystalline

c-BN sintered bodies [155]. To get a full conversion of h-BN into pure c-BN,
pressure and temperature must be relatively high and pure h-BN (<0.03 wt%
B2O3) is necessary. Typical conversion parameters are 7.7 GPa, 1900±2700 °C,
and 15 min reaction time. Above 2400 °C no h-BN was observed by X-ray
diffraction in the sintered bodies. The c-BN grain size increases from 0.5 lm at
2300 °C to 5 lm at 2700 °C [155].
4.1.2
Catalytic Conversion from h-BN into c-BN
Up to now the so-called ``catalytic process'' is the only way to produce c-BN on
an industrial scale. However, ``catalytic'' is not the correct scienti®c term,
because the activation energy for transformation is not decreased by these
substances. The substances which are used have the function of a solvent, and
are responsible for the formation of c-BN. This method is successful because of
the different solubilities of c-BN and h-BN in the ¯ux. The precursor
substances form a eutectic melt with the h-BN [152]. If the reaction conditions
are in the domain of stable c-BN, spontaneous crystallization takes place and
the c-BN growth rate is relatively high.
The domain for c-BN formation and the solubility conditions are strongly
in¯uenced by the ¯ux composition. In the catalytic process h-BN is mixed with
the catalyst or just only brought into contact. In some experiments it has also
been shown that c-BN can nucleate on substrates placed in the reaction
chamber [156] (Fig. 11).
In the text below an overview of different solvents, and their properties, for
the production of c-BN is given.
4.1.2.1
Alkaline and Alkaline Earth Elements
Lithium, magnesium, and potassium were already used by Wentorf in 1957

for the ®rst reported synthesis of c-BN [2]. First the melted metals react
with h-BN to the corresponding nitrides (Li3N, Mg3N2, Ca3N2) and
Fig. 11. High-pressure high-temperature reaction chambers for the synthesis of c-BN
(substrates were added to study the c-BN nucleation) [156]
elemental boron (3Li + BN ® Li3N + B). In a second step the nitrides and
h-BN react to a eutectic mixture. When a condition is reached, at which c-
BN is the stable phase, the rest of the h-BN can be dissolved in the ¯ux and
c-BN crystallizes. The color of the obtained powder is black or brown,
which can be explained by the incorporation of excess boron in the c-BN
lattice [2, 157].
4.1.2.2
Alkaline and Alkaline Earth Nitrides
Starting with nitrides of the alkaline and alkaline earth group elements no
elemental boron is formed and the c-BN produced is yellow and transparent.
Both the quality and the yield of c-BN are increased.
In 1961 Wentorf [157] assumed the existence of a lithiumboronitride-phase
in the Li3N-BN system, which was identi®ed in 1972 by DeVries and Fleischer
[158] as Li3BN2. At high pressure Li3N and h-BN react to Li3BN2, forming a
melt at HP-HT conditions and acting as solvent for h-BN. Because the
solubility of c-BN in the melt is lower than that of h-BN, nucleation and growth
of c-BN takes place. For the Li3N-BN melt, the BN diffusion coef®cient in the
melt was calculated as 3 1 ´ 10)7 cm2 s)1, at the synthesis conditions of
6.6 GPa and 1770 K [159]. The Li3N-BN system is often used to study the BN
phase diagram [40, 41, 160] and to produce c-BN in industrial scale.
By using Mg3N2 for the synthesis of c-BN a similar mechanism is observed.
Endo et al. [161] found Mg3B2N4 as active ¯ux in the system Mg-BN. In the
system Mg3B2N4-h-BN on eutectic point exists at 1568 K and 2.5 GPa. At
5.5 GPa the eutectic temperature decreases to 1550 K [162]. In several papers
the c-BN growth is described in the Mg3N2-BN system at various conversion
parameters [163±165]. For the direct production of c-BN compacts by catalytic
conversion sintering only 1 vol.% Mg3BN3 was used. The compacts contain
3±10 lm c-BN grains and the Mg3BN3 is located at the grain boundaries [166].
Similar to the magnesium boronitrides, the calcium compound Ca3N2 was
used as catalyst by DeVries and Fleischer in 1972 [158]. Later in 1981, Ca3B2N4
was used, which melts at 1685 K at a pressure of 2.5 GPa. Ca3B2N4 and h-BN
built a eutectic mixture at 1589 K and 2.5 GPa [167].
4.1.2.3
Alkaline and Alkaline Earth Fluoronitrides
The application of alkali and alkaline earth ¯uoronitrides has been investigated
by Demazeau et al. [168, 169]. The synthesis of c-BN was successful by using
Mg2NF, Mg3NF3, and Ca2NF under similar pressure and temperature condi-
tions as they were found for Mg3N2, but the yield was much higher when using
the ¯uoronitrides (90% instead of 30% at 7 GPa and 1470 K after 10 min).
Kinetic dependences of the conversion of h-BN into c-BN in the Li-B-N-F
system have been studied by introducing about 6% additional nitrogen with
addition of nitrogen compounds [170]. The stoichiometry of the c-BN was
increased by adding nitrogen (synthesis pressure 4.3 GPa, 1750±1850 K).
4.1.2.4
Water and Ammonium Compounds
When water is added to the h-BN starting material, c-BN can be produced at
5 GPa and 1770 K or at 6 GPa and 870±970 K [171, 172]. These are rather low
temperature and low pressure conditions compared to other syntheses. The
c-BN obtained is rather ®ne-grained and suitable for sintering. NH4B5O8 was
identi®ed as the ¯ux in this process [173]. Essential for the conversion from
h-BN to c-BN is the ratio NH 4 /B2O3.
Unusual ¯ux precursors like urea, ammonium nitrate, boric acid, or
ammonium boron-hydride have also been studied [173]. In all these cases
derivatives of amino borate form the liquid phase.
The limits for the synthesis parameters are: lower temperature limit 1070±
1270 K; lower pressure limit 4.8 GPa for H2O/4.3 GPa for urea/4.6 GPa for
boric acid.
A rather interesting discovery was made by Kobayashi et al. [174]. At 1470±
1870 K and 4.2 GPa, c-BN was transformed into h-BN. According to the earlier
BN phase diagrams this should not have happened, because these conditions
are located in the region where c-BN should be stable.
Synthesis in the B-B2O3-hBN system at temperatures >2300 K and in the
pressure range of 4±6.5 GPa resulted in w-BN and c-BN phases [175].
4.1.2.5
Hydrazine
Recently Demazeau reported the use of hydrazine and lithium nitride as ¯ux,
which allowed the transformation from h-BN to c-BN at a rather low pressure
of 2±3 GPa and a temperature of 1000 K [176]. Similar experiments by
Solozhenko et al. [177] con®rmed these results but there are still some
questions about measuring the process parameters. One explanation for the
effects of hydrazine is that it explodes during heating the sample, which result
in higher pressures and temperatures than described.
4.1.3
Temperature Gradient Method
In this high pressure process large c-BN crystals should be grown. The key to
obtain larger c-BN crystals of high quality is to control the super-saturation of
BN in the solvent. A temperature gradient method with exact regulation of the
temperature has been developed (Fig. 11) [156, 178].
There are two possibilities for the regulation of the temperature gradient
[179]. On the one hand a vertical reactor can be used, in which a temperature
gradient exists. On the other hand a molybdenum disc with a hole in the
middle can be put between the h-BN and the ¯ux, and by varying the hole's
diameter the temperature gradient can be regulated.
It was possible to grow c-BN single crystals up to 2 mm in size, without
using seeding crystals [178]. Process parameters were 1500±1750 °C and
5.5 Gpa, with a h-BN source and Li3BN2 or Ba3B2N4 as solvents. Crystal quality
of the c-BN was best in the lower temperature range where c-BN nucleation is
rare and crystal growth slow. The reaction time for the larger c-BN crystals was
up to 80 h.
Using the temperature gradient method it was also possible to deposit
c-BN on various substrates, e.g., on diamond crystals or CVD-diamond sheets
[156, 180].
4.1.4
Dynamic High Pressure Conversion
For the dynamic high pressure conversion, also known as shock wave synthesis,
an explosive shock wave is used for the conversion of h-BN into c-BN. High
pressure and high temperature are reached for a short period of time
(milliseconds). To guarantee a fast temperature decrease a metal powder, like
copper, is added to h-BN in an amount of 5% [181].
This method was ®rst applied by Du Pont to prepare ultra®ne diamond
powders on an industrial scale.
Today BN synthesis by shock-wave methods is mainly used to produce
superhard boron-carbon-nitrogen mixtures called ``heterodiamond'' [182,
183].
4.2
Low-Pressure Synthesis of c-BN
Low-pressure deposition of diamond is a commonly used industrial process

[184]. Because HP-HT synthesis for c-BN and diamond works under similar
conditions, it was assumed that the low-pressure synthesis of c-BN should be
possible analogous to that of diamond. However, there are a lot of differences
between c-BN and diamond that make a low-pressure deposition of c-BN
rather dif®cult:
± c-BN is the stable phase ± diamond is metastable (at standard conditions).
± c-BN consists of two elements ± diamond contains only carbon.
± h-BN selective etching is complicated ± graphite can easily be etched by
atomic hydrogen.
± Stabilization of the c-BN surface is dif®cult ± the diamond surface is
stabilized by hydrogen.
± Complex precursors are needed to deposit c-BN ± for diamond methane is
convenient.
± Analytical characterization of c-BN is complex ± diamond is identi®ed by
Raman measurements.
Deposition of nano-cBN ®lms is possible with ion-assisted CVD and ion-
assisted PVD techniques. Since the c-BN crystal size in such layers is in the
range of nanometers, such layers should be called nano-cBN. The amount of
grain boundaries in such materials is rather high and this fact should not be
negated by using the notation ``pure c-BN''.
4.2.1
Selective Etching of h-BN and c-BN
For a successful CVD-synthesis of c-BN a reaction system has to be found,

where c-BN is deposited and h-BN and amorphous BN formation can be
prevented or where the undesirable phases can be removed by selective
etching during the deposition.
For the carbon system the atomic hydrogen acts as medium for selective
etching and stabilization of the diamond surface.
In order to ®nd a substance which allows selective etching of h-BN, several
compounds (e.g. atomic hydrogen, chlorine, mixtures of chlorine with
hydrogen, BF3) have been tested. Of these substances BF3 shows the best
selectivity for etching h-BN compared to c-BN (Fig. 12) [185, 186]. The
disadvantage of BF3 ± in contrast to atomic hydrogen in the carbon system ± is
the smaller dissolution rate of h-BN (about ten times lower than in the carbon/
atomic hydrogen system). The consequence is that selective etching becomes
relevant only at very low BN growth rates. CVD deposition experiments with
BF3 addition always resulted in h-BN or a-BN layers [185]. Last but not least,
BF3 is a very reactive and corrosive substance which also reacts with the
substrate materials. This results in additional problems during deposition of
coatings on conventional tools.
4.2.2
Mechanism for Ion-Assisted c-BN Deposition
In most cases nano-cBN deposition is supported by the generation of ions.

Parameters for the substrate bias (ion energy, ion mass, etc. [187]) are similar
to that in PVD as well as in CVD methods. Therefore a equal growth
mechanism in both methods can be considered. When starting the deposition
process, commonly an oriented h-BN layer is deposited on the substrate. On
Fig. 12. Selective etching of h-BN and c-BN with various gas phases and comparison with the
carbon system [185]
this interlayer the c-BN nucleates and forms a layer [188]. Because of the
deposition conditions and the ion-bombardment ± which is necessary for the
c-BN deposition ± creation of crystal defects and secondary nucleation occur,
and the single-crystalline areas are very small (nm range) [189].
To describe the nano-cBN deposition four models have been proposed: the
compressive stress model [190, 191], the sub-plantation model [192, 193], the
selective sputter model [194], and the momentum transfer model [195].
4.2.2.1
Compressive Stress-Model
For the c-BN formation a stress threshold was observed in the deposited
layers. The h-BN intermediate layer shows a preferred orientation, where
the c-axis of the h-BN is parallel to the substrate. Both effects are explained
by the compressive biaxial stress induced by the ion bombardment. The
mechanism for the conversion of h-BN into c-BN is explained by rather
high temperatures originated during thermal spikes (direct h-BN ® c-BN
transformation). The stress caused by the bombardment with high energetic
ions is considered to be a reason for the growth of the c-BN crystals [190,
191]. A stress within the layer of up to 10 GPa has been observed. This
biaxial stress causes a hydrostatic pressure up to the values usual in HP-HT
synthesis.
4.2.2.2
Sub-Plantation Model
Another possibility to explain the ion assisted c-BN deposition is the sub-
plantation model. The nucleation of c-BN crystals takes place under the
surface of the substrate caused by sub-plantation of the ions and stress. The
sub-plantation and high nucleation rates result in the nano-cBN coatings.
4.2.2.3
Sputter-Model
The essential mechanism in the sputter-model is that h-BN can be removed

more easily by selective sputtering than c-BN (if the BN mixtures are deposited
simultaneously and the h-BN is selectively etched, the c-BN layer remains)
[187, 196].
4.2.2.4
Momentum Transfer Model
A correlation between the total momentum of impinging ions per deposited

boron atom and the c-BN deposition has been observed. In this model, c-BN
formation is correlated with the momentum-drive process, such as the formation
of point defects in conjunction with the stress-induced phase transformation.
4.2.3
PVD Methods for Nano-cBN Deposition
The basic concept of the c-BN nucleation by ion-beam-deposition has been

described by Weissmantel et al. in 1980 [197, 198]. Adhesion problems and
dif®culties in analytical characterization have caused large problems. A
breakthrough of nano-cBN coatings and their various applications seems to be
possible.
The methods mostly used for nano-cBN deposition are: ion-beam-assisted
deposition (IBAD) [199]; mass selected ion beam deposition (IBD) [200]; ion
plating [201]; RF- or magnetron sputtering [202] and laser deposition [203]
(Fig. 13).
4.2.3.1
Ion-Beam-Assisted Deposition (IBAD)
Ion beam deposition employs an ion bombardment onto the substrate with
high energetic nitrogen ions. Boron is vaporized as ions (mostly by an electron
beam) and shot onto the substrate together with the nitrogen ions.
This method was used in 1983 by Satou and Fujimoto [204]. Boron was
evaporated by an electron beam, and nitrogen ions (40 keV) were shot onto
the substrate (Ta, NaCl). The deposited layers had a thickness of 6800 A Ê and
Ê
the growth rate was up to 100 A/min.
IBAD allows one to control all decisive parameters for the c-BN deposition
(e.g. ion energy, ion ¯ux, and ion/boron ratio). Typical deposition parameters
are: 300±1200 eV ion energy, 0.4±0.8 mA/cm2 ion current, 2 ´ 10)4 mbar
pressure, 300 °C substrate temperature, ratio ions/boron atoms 0.5±3 [196, 199,
205±207].
4.2.3.2
Mass Selected Ion Beam Deposition (IBD)
The B+ and N+ ions are extracted from an ion source with high voltage (e.g.
30 kV), and after mass selection and deceleration (10 eV to 1 keV) the ions are
Fig. 13. Dependence of the nano-cBN deposition on deposition parameters by IBAD

experiments. The results of literature data are summarized for the parameters ion energy
and substrate temperature
deposited onto the substrate. Substrate temperatures between room temper-

ature and 400 °C are possible. Ion energy, ion ¯ux, and substrate temperature
in¯uence the quality of the nano-cBN layers [200].
4.2.3.3
Ion Plating
Ion plating and ion beam deposition differ in the use of additional ions (e.g.
Ar) for deposition. The process can be divided into three steps [201]:
± Elementary boron is vaporized by an electron beam.
± Support gases (NH3, N2, and Ar) are activated by ionization.
± The ions are accelerated by d.c. or RF bias and deposited onto the substrate
surface resulting in the growth of c-BN.
The ionization of the gas can be achieved by electron beam vaporization
[208], hollow cathode discharge [209, 210], glow discharge [34, 192], arc
discharge [36], or glow discharge in a parallel magnetic ®eld [33].
4.2.3.4
Reactive Sputtering
The material is sputtered off an h-BN target and deposited on the substrate in
a nitrogen/argon atmosphere. For deposition a high negative substrate bias is
applied [60, 202, 211].
As sputtering sources, h-BN and B4C compacts can be used [212, 213], and
different reactive sputtering techniques can be applied by variation of the
power supply on the target and the substrate (e.g. RF sputtering with h-BN
targets and RF-powered substrate; RF sputtering with B4C targets and RF- or
d.c.-powered substrate; or d.c. magnetron sputtering with B4C targets and
RF-powered substrate) [214±216].
4.2.3.5
Laser Deposition
This method uses a laser for evaporating the boron compounds from a target
(e.g., BN target). To regulate the ¯ux of BN to the substrate the laser can be
pulsed. Additionally, a nitrogen/argon ion-beam is generated and directed
onto the substrate's surface. Various layers have been deposited, and the BN
transition from substrate to the h-BN interlayer and ®nally to the nano-cBN
has been studied in detail [203].
4.2.4
Plasma CVD Methods
Using conventional thermal CVD various BN modi®cations but no c-BN or

nano-cBN are formed. Therefore, to synthesize c-BN a plasma is applied to
activate the gas phase. Typical deposition methods are ECR plasma CVD [217,
218], ICP CVD [207], and bias enhanced plasma CVD [219].
4.2.4.1
ECR Plasma CVD
Electron cyclotron resonance (ECR) is used to deposit c-BN layers. Yokohama

et al. [220] considered a negative substrate bias as an essential parameter for
the deposition in ECR plasma to accelerate the ions formed in the plasma. A
linear coherence was found between the bias and the etching rate of c-BN and
h-BN, which was higher for h-BN. Three points are important for the crystal
growth:
± The deposition of h-BN is the result of precursor substances that are created
in the plasma by the reaction with electrons.
± The deposition of the c-BN phase is the result of precursor substances that
are created by the reaction with Ar+ ions.
± The growth of both phases depends on the nucleation rates, caused by the
precursor substances and by the etching rate.
Typical parameters are 0.5 Pa, plasma gas Ar/N2 mixtures, microwave
power 1000 W, substrate temperature 250 °C, and boron precursor B2H6 [221].
4.2.4.2
ICP CVD
A very intense inductively coupled plasma (ICP) could be created, using a

13.56-MHz RF-source and coupling the power through a quartz tube into the
plasma. Working pressure was 2 ´ 10)2 mbar and the substrate could
be heated up to 800 °C [222]. The precursor was trimethylborazine
((HBN-CH3)3), which was transported in a nitrogen/argon carrier gas. Similar
to other deposition processes, prior to the c-BN nucleation an oriented h-BN
layer was formed [223].
4.2.4.3
Bias Enhanced Plasma CVD
Plasma CVD has been used since the middle of the 1970s. For the creation of
the plasma, DC glow discharge [224], RF glow discharge [219, 225±227],
microwave plasma [228, 229], or plasma jets [230] are used.
Under conditions similar to those used for the synthesis of diamond layers,
only layers with a small amount of c-BN can be deposited. A lot of publications
can be found about the deposition of c-BN with CVD methods, reporting
different parameters that are necessary for the growth of c-BN.
Mendez et al. [231] showed in deposition experiments with RF plasma and
without substrate bias that there is a coherence of the nano-cBN amount with
the substrate temperature and the plasma power. Additionally a dependence
on the substrate material used (e.g. Si or NaCl) could be found. Further
investigations used Ar ions, where the nano-cBN amount depends on the

process parameters of ion bombardment [232].
Nano-cBN in layers has been deposited by activating the gas phase with RF
plasma and hot-®lament. The highest amount was reached by introducing a
gas mixture consisting of BH3NH3-H2 into the reactor at conditions where the
amount of atomic hydrogen in the gas phase reached a maximum. As a reason
for the high nano-cBN amount, the selective etching of h-BN by elemental
hydrogen was proposed by Saitoh and Yarbrough [229].
Matsumoto et al. [233] tried a bias assisted plasma jet with the reactive gas
mixture Ar-N2-BF3-H2 for selective etching of h-BN [234]. Using a d.c. bias
voltage, the nano-cBN grain size in the layers could be increased from 7 nm
()150 V) to 12 nm ()80 V) [235]. By optimizing the process, nano-cBN
coatings thicker than 20 lm with c-BN grain size up to 100 nm could be
deposited [230].
4.2.5
Properties and Applications of Nano-cBN Coatings
One goal in the ®eld of c-BN deposition is a coating which can be used for
machining of iron-based materials. Such coatings offer the possibility to use a
superhard material instead of the conventional carbide and oxide layers.
Matsumoto and co-workers showed that the deposition of thick nano-cBN
layers is possible and the adhesion on Si substrates is of acceptable quality
[230]. The one major question is whether this process is also suitable for
hardmetal (WC-Co) substrates or not, and whether there might be any
problems with the Co binder phase and the BF3 in the gas phase?
The stress in c-BN layers caused by the high energetic ions is a problem,
because many of the nano-cBN coatings delaminate during or immediately
after deposition. Several investigations have showed ways to reduce the stress
in the layers (e.g. buffer layers [205, 217, 236], regulation of the ion energy
[207], or ion-induced stress relaxation [206]).
The outlook for industrial applications of nano-cBN coatings looks quite
good, but some problems with the deposition process and the substrate
materials still have to be solved.
4.2.6
The Simple Chemical Way?
Several attempts to grow c-BN at standard pressure have been performed.

Similar to the high-pressure high-temperature synthesis, melts were used in
the temperature range between 600 and 800 °C at standard pressure. Various
compounds were mixed with ®ne grained c-BN powder (for seeding) and
heated up to allow grain growth. The stability of c-BN in chemically active
media depends on the properties of the reacting agent. Reductive media like
lithium metal or lithium-boride generally lead to a strong degradation of c-BN.
In this case there is no uniform reaction with c-BN, and different phases result.
The reductive dissolution of c-BN ®rst led to the formation of boron or boron-
rich boron nitride B50N2, which further react with excess of lithium to lithium-
borides.
It could also be shown that the system Li3BN2/c-BN interacts at elevated
temperatures, leading to an obvious change in the morphology of the c-BN
seed crystals. In all cases a formation of well faceted surfaces could be
detected. From this system the growth of c-BN is most likely.
Actually, a spontaneous nucleation of c-BN from a degradation of Li3BN2
will not occur due to the fact that the less dense phase will nucleate ®rst
according to the Ostwald-Volmer rule, which is a general rule of thumb for the
kinetic behavior of reactions [237]. Thus, seeding with c-BN can overcome
this problem. Hence, this approach seems to be well suited to develop a
low-pressure synthesis of c-BN.
4.3
Applications for c-BN Products
Because of its excellent mechanical and electrical properties c-BN is of great

interest for a number of applications (e.g. grinding powders, wear parts,
electronic parts, etc.). Searching the literature, the impression arises that the
applications of c-BN are kept a bit secret. Most of the relevant references
are patents (up to 90% depending on the topic) giving less exact data about
the process. Papers published in journals giving detailed information are
rare.
4.3.1
Pure Polycrystalline c-BN (PcBN)
Pure PcBN can be produced by direct conversion without catalyst [155] or

with small additions of catalyst (<1 vol.% Mg3BN3) [166]. Compared to PcBN
sintered with a binder phase, the pure PcBN materials show increased
hardness (5000±5500 kg/mm2 instead of 3000±4000 kg/mm2). Thermal stabil-
ity in air is high for the pure PcBN (1350 °C) but lower for the type containing
Mg3BN3 (700 °C) [155]. Pure PcBN is an interesting material for wear
applications and shows high thermal conductivity. Because of this instance
many data are available [238±241].
4.3.2
Polycrystalline c-BN (PcBN) for Wear Applications
In the ®eld of high temperature wear applications c-BN is superior to diamond

because of its higher stability against temperature, its better oxidation
resistance, and lesser reactivity with iron. The resistance of c-BN against
oxidation in air up to 1200 °C is much higher than that of diamond because a
protecting layer of B2O3 is formed by reaction with the oxygen in the air,
protecting c-BN from further oxidation. Therefore, c-BN is mostly used for the
processing of steel, whereas diamond is preferred for the processing of stone,
ceramics, and Al-alloys. Conventional cutting and milling tools made of

hardmetal, high speed steel, Al2O3, SiC, BC, etc. can be replaced by c-BN with
all its advantages, like higher feed rate or smaller abrasion (e.g. factor of 7±50
higher compared to WC tools [152]).
The c-BN products can be divided up into dense products (e.g. for cutting,
milling operations) and in porous products (e.g. for grinding).
4.3.2.1
Dense PcBN Products
A polycrystalline compact ± containing c-BN powder and a second phase ± can

be sintered at parameters where c-BN is the stable phase (to prevent h-BN
formation). Therefore, high-pressure high-temperature sintering is necessary,
and the maximal diameter for the produced parts is limited by the dimension
of the high pressure apparatus.
The binder phase of such materials can be a ceramic, a metal, or a resin.
4.3.2.1.1
Ceramic Binder for c-BN
Superhard BN parts are used as advanced grinding and cutting materials for
processing metals. Typical trade names are Amborite, BZN HTZ, Elbor-RM,
Kiborit, and Hexanit-R.
Amborite, containing AlN/AlB2 as interphase, has a high thermal stability
and is used for machining Ni-Cr containing martensitic iron alloys [242].
Elbor-RM, Kiborit and Hexanit-R contain intermetallic phases based on
Al, Cr, and Zr. Additionally, Hexanit-R contains higher amounts of w-BN
[243].
For the production of bodies containing c-BN and w-BN, the w-BN powder
is mixed with various carbides from Group 4 or 5 (IVB, VB) elements, and
afterwards the mixture is hot-pressed under high-temperature and high-
pressure conditions. With this method a sintered workpiece containing w-BN
and c-BN in a continuous binding phase is obtained [244].
Many different compositions of the PcBN binder phase are described in the
literature (e.g. AlN [245], AlN/AlB2 [246], Al2O3 [247], SiC [248, 249], Si3N4
[250], TiB2, ZrB2, HfB2 [251, 252], TiC [253], TiN [254], borosilicate glass
compositions [255]).
With a high-pressure hot-pressing method c-BN-TiC/TiN composites are
prepared by sintering and subsequently heat treatment between 1000 and
1400 °C. The samples exhibit a dense polycrystalline structure of c-BN-TiN/
TiC, and a thin layer of ®ne TiB2 is visible at the c-BN-binder interface.
Hardness decreases signi®cantly after heat treatment [256].
A special type of such compacts is a mixture where diamond is the second
phase. The products contain ³90 vol.% c-BN and 2±10 vol.% diamond
(particle size 0.5±2 lm). After mixing the powders, HP-HT sintering without
any binder phases follows (<100 kbar and <1800 °C). The sintered products
show high density [257].
4.3.2.1.2
Metallic Binders for PcBN
The bonding between c-BN and metallic binder phases is mostly quite good,
due to the formation of interlayers of metal nitrides or borides. For the
sintering process itself, the composition as well as the melting point of the
alloys used are important. In the case of applying low sintering temperature
(below the conversion temperature of c-BN into h-BN) the process can be
conducted at standard pressure; when higher temperature is necessary, high-
pressure must also be applied. Some examples for metallic binder are: Al, Al
alloy and/or Ni, Fe, Co, Cr, and/or Mn alloys [258]; Al alloys containing Co, Cr,
Fe, Mo, Ni, Si, Ti, V, W, and/or Zr [259]; Ni- or Co-based heat resistant alloys
[260].
Compacts of c-BN and Al are prepared by high-pressure hot-pressing and
subsequent annealing at 950 °C under pressure (3 ´ 10)3 Pa) for 1 h. After
annealing a multilayer is formed on c-BN, starting with AlB10 and AlB12
phases, followed by a layer of columnar AlN. The thermal treatment results in
an increase of mechanical strength of the sintered BN-Al system [261].
When sintering c-BN with Ti and subsequent heat treatment at 950 °C, all
the Ti reacts with the BN matrix forming TiN and TiB2 at the BN interface, and
voids are observed at the BN/TiB2 interface [262]. Similar to the Ti compacts,
Zr-containing compacts are investigated in which ZrN and ZrB2 formation is
observed. Again the mechanical properties can be increased by heat treatment
[263].
Another way of production is the coating of c-BN by electro-less plating
with Ni-P, Ni-B, Ni-Fe-P, Ni-Cr-P, Ni-Cu-P, or Ni-W-P alloys, and mixing
these powders with ³1% of various carbides, borides, nitrides, silicides, and/or
oxides. These powders are compacted and pre-sintered at 700±900 °C. Finally,
hot-isostatic pressing at 1000±1400 °C and 1000±2000 bar is performed to
reduce porosity [264].
In the case of a metallic binder phase (e.g. Co, Ni, Fe, and its alloys),
ceramic whiskers can also be added to the c-BN [265]. The process has been
described in detail for Si3N4 whiskers [266].
4.3.2.1.3
Hardmetal/PcBN Compacts
PcBN compacts are often bonded to a hardmetal or hard steel alloy for
manufacturing a tool. This composite can be produced by direct sintering of
WC-Co-PcBN mixtures or by brazing.
Onto the surface of the sintered part (e.g. hardmetal) the mixture of c-BN
and/or diamond powder is bonded during high-pressure high-temperature
sintering [267].
Sintered cutting tips or inserts containing c-BN can be brazed to a cemented
carbide or other substrates. Brazing bond strength is increased by interlayers
of various carbides, nitrides, or carbonitrides [268].
The various PcBN compacts are mainly used for machining of a wide variety
of hard and/or abrasive ferrous workpiece materials up to high temperatures
exceeding 1000 °C. For example: pearlitic gray cast iron; hardened ferrous
metals (>45 HRc); high speed steel; case hardened steel, superalloys; sintered
tungsten carbide >16% Co are machined by PcBN compacts.
4.3.2.2
Porous PcBN Products
Porous composites containing hard materials are mainly used for grinding
operations. For such applications several parameters are important: e.g., the
distance between the grains for good chip formation; good adhesion between
the hard grains and the binder phase; good balance between the erosion of the
binder phase and the hard material; porosity to allow better cooling [269].
The grinding tools can be classi®ed by their binder phase as resin bonded,
ceramic bonded, or metal bonded (sintered or electrodeposited).
Wear resistance of such tools strongly depends on the tool composition, the
machined material, and the machining conditions, which makes this topic very
complex [270±272].
4.3.2.2.1
Resin Bonded Grinding Tools
Grindstones are manufactured by the following steps: (1) mixing abrasive grit
with epoxy resin and polymeric foaming agent, (2) casting the resulting
mixture in a mould, (3) hardening the epoxy resin, and (4) heating to form
designed pores which also can be formed during steps 3 and 4 [273, 274].
4.3.2.2.2
Ceramic Bonded
Grinding tools of c-BN with ceramic binders are used for grinding of hardened
steel parts, e.g. roller bearing rings, tooth gears, or engine injection valves
[275]. Ceramic bonded grinding wheels can be produced with porosity in the
range 50±80%. Grain size of the hard material and the composition of the
ceramic binder phase can be varied over a wide range [276].
Mixtures of c-BN particles with other hard materials (e.g. Al2O3, SiC) can be
bonded by vitreous material. The binder is sintered at 850±900 °C giving a
crystalline phase which increases the mechanical strength and the hardness of
the tools. A typical binder consists of: SiO2 10±20 wt%, ZnO 40±55 wt%, B2O3
15±30 wt%, Al2O3 2±10 wt%, MgO 2±10 wt%, alkali metal oxides 0.1±2 wt%,
and TiO2 0.1±2 wt% [277, 278].
A c-BN ± silicon nitride ceramic composite can be produced directly by
sintering a mixture of c-BN powder and Si powder in N2 atmosphere. The
composites have high resistance against heat, oxidation, and thermal shock
[279].
4.3.2.2.3
Metal Bonded and Electrodeposited Grindstones
Metal bonded grindstones can be prepared either by powder metallurgy or by
electrodeposition. Fabrication of metal-bonded grinding tools by green tape
laser sintering is a rather new method. The porosity of Cu sintered parts is

utilized for manufacturing tools containing various hard materials (e.g. c-BN
or Al2O3) [280].
An advantage of metal-bonded grindstone (e.g. brass-bond cubic BN) is the
possibility of electrolytic dressing (surface preparation). By applying voltage
between the grindstone and an opposite electrode the metal bond can be
selectively etched [281].
To produce grinding wheels, a foamed substrate (e.g. polyurethane foam) is
coated with an electrically conducting material (e.g. graphite) and then coated
with a metal layer (e.g. Ni [282], Cu, Cr, Cd, Sn [283]) by electrodeposition.
Finally, a coating with abrasive particles (e.g. c-BN, diamond) is deposited to
produce the grinding wheels [284]. The electrodeposition can be performed by
dipping the substrates in a plating solution containing the ultra®ne super-
abrasives. The superabrasive surface is treated to generate a difference between
its isoelectric point and the pH of the plating solution. Uniform and dense
electrodeposited layers are formed [285].
4.3.3
Electronic Applications
Because of its large band gap c-BN is a good electrical insulator. Due to its
extraordinarily high thermal conductivity it is used as heat sink for
semiconductor laser or microwave applications.
Semiconducting c-BN was synthesized by Wentorf [286] using HP-HT
methods. For p-material Be doping was used, the n-type was produced by
doping the material with S, Si, or KCN.
With HP-HT methods the production of p-n modules, diodes, or LEDs [287,
288] is possible. Diodes of c-BN can be used up to 600 °C because of the large
band gap and its heat resistance. LEDs emitting light from red to ultra violet
can be produced. Such electronic applications are at the beginning of their
development. The main problem today is the fabrication of single crystals of a
suf®cient size, which is not even possible with HP-HT methods. Thus, an
industrial application does not exist up to now.
Sintered heat sink materials were produced of c-BN and AlN (£1% O) with a
composition between 40±90 vol.% c-BN and the balance AlN. Sintering
conditions were temperatures exceeding 1200 °C and pressures exceeding
40 kbar. The products have a thermal conductivity exceeding 0.6 cal/cm.s.K at
25 °C. A mixture consisting of 80 vol.% c-BN (grain size 15 lm) and 20 vol.%
AlN (5 lm) gave a thermal conductivity of 1.7 cal/cm.s.K at 25 °C [239, 289].
5
References
1. Balmain WH (1842) J Prakt Chem 27: 422
2. Wentorf RH Jr (1957) J Chem Phys 26: 956
3. Mishima O, Era K (2000) In: Kumashiro Y (ed) Electric Refractory Materials. Marcel
Dekker, New York Basel, pp 495±549
4. Pierson HO (1975) J Compos Mater 9: 228±240

5. Landolt ± BoÈrnstein (1978) Serie III, Band 7, Springer, Berlin Heidelberg, New York
6. Chopra KL, Agarwal V, Vankar VD, Deshpandey CV, Bunshaw RF (1984) Thin Solid
Films 307±312
7. Wentzcovitch R, Chang KJ, Cohen ML (1986) Phys Rev B34(2): 1071±1079
8. DeVries RC (1972) Cubic boron nitride: handbook of properties. General Electric
Company
9. Soma T, Sawaoka A, Saitoh S (1974) Mater Res Bull 9: 755
10. Gmelin (1974) Handbuch der anorganischen Chemie, Band 13: Borverbindungen,
Springer, Berlin Heidelberg, New York
11. Corrigan FR, Bundy FP (1975) J Chem Phys 63: 3812±3820
12. Demazeau G (1993) Diamond Related Mater 2: 197±200
13. Taylor KM (1955) Ind Eng Chem 47: 2506±9
14. Sokolowski M, Sokolowska A, Rusek A, Romanowski Z, Gokieli B, Gajewska M (1981)
J Cryst Growth 52: 165±167
15. Karim MZ, Cameron DC, Hashmi MSJ (1993) Surf Coat Technol 60: 502±505
16. Chrenko RM (1974) Solid State Commun 14: 511±515
17. Sokolowski M, Sokolowska A, Michalski A, Romanowski Z, Rusek-Manzurek A,
Wronikowski M (1981) Thin Solid Films 80: 249±254
18. Hibbs LE Jr, Wentorf RH Jr (1974) Eighth Plansee Seminar 2±42, p 5
19. Bundy FP, Wentorf RH (1963) J Chem Phys 38: 1144±1149
20. Xu YN, Ching WY (1993) Phys Rev B48: 4335±4351
21. Kurdymov AV, Solozhenko VL, Zelyavsky WB, Petrusha IA (1993) J Phys Chem Solids
54: 1051±1053
22. Sachdev H, Haubner R, NoÈth H, Lux B (1997) Diamond Related Mater 6: 286±292
23. Hamilton EJM, Dolan SE, Mann CM, Colijn HO, McDonald CA, Shore SG (1993)
Science 260: 659±661
24. Schmolla W, Hartnagel HL (1983) Solid State Electron 26: 931±939
25. Batzanov SS, Blokhina GE, Deribas AA (1965) J Struct Chem 6: 209
26. Akashi T, Sawaoka A, Saitoh S, Arahi H (1976) Jpn J Appl Phys 15: 891
27. Sokolowska A, Olszyna A (1992) J Cryst Growth 121: 733±736
28. Sokolowska A, Olszyna A (1992) J Cryst Growth 116: 507±710
29. Michalski A, Olszyna A (1993) Surf Coat Technol 60: 498±501
30. Sokolowska A, Wronikowski M (1986) J Cryst Growth 76: 511±513
31. Batsanov SS, Kopaneva LJ, Lazareva EV, Kulikova IM, Barinsky RL (1993) Propel
Explos Pyrotech 18: 352±355
32. Weissmantel C, Bewilogua K, Breuer K, Dietrich D, Ebersbach U, Erler HJ, Rau B,
Reisse G (1982) Thin Solid Films 91: 31
33. Murakawa M, Watanabe S (1990) Surf Coat Technol 43/44: 128±136
34. Rother B, Zscheile HD, Weissmantel C, Heiser C, HolzhuÈter G, Leonhardt G, Reich P
(1986) Thin Solid Films 142 83±99
35. Halverson W, Quinto DT (1985) J Vac Sci Technol A3: 2141±2146
36. Ikeda T, Kawate Y, Hirai Y (1990) J Vac Sci Technol A8: 3168±3174
37. Lorenz R, Woolcock J (1928) Z Anorg Chemie 176: 283±304
38. Campbell J (1949) J Electrochem Soc 96: 318±333
39. Widany J, Frauenheim T, Lambrecht WRL (1996) J Mater Chem 6: 899
40. von der GoÈnna J, Meurer HJ, Nover G, Peun T, SchoÈnbohm D, Will G (1998) Mater Lett
33: 321±326
41. Will G, Nover G, von der GoÈnna S (2000) J Solid State Chem 154: 280±285
42. Wentorf RH (1959) J Phys Chem 63: 1934±1941
43. Leonidov VY, Timofeev IV, Solozhenko VL, Rodionov IV (1987) Russ J Phys Chem 61:
1503±1504
44. Solozhenko VL, Leonidov VY (1988) Russ J Phys Chem 62: 1646±1647
45. Wise SS, Margrave JL, Feder HM, Hubbard WN (1966) J Phys Chem 70: 7±10
46. Maki J, Ikawa H, Fukunaga O (1991) New Diamond Sci Technol 1051±1055
47. Solozhenko VL (1993) Thermochim Acta 218: 221±227

48. Svensk EnergiData Agersta (1986) Datenbank EKVICALC 1.21, Svensk EnergiData
Agersta, BaÈlinge, Sweden
49. Phillips JC, Van Vechten JA (1970) Phys Rev B2: 2147±2160
50. Lide DR (1991/2) CRC handbook of chemistry and physics, 72nd edn, CRC Press,
pp 5±21
51. Solozhenko VL (1995) High Pressure Res 13: 199
52. Solozhenko VL (1995) J Hard Mater 6: 51±65
53. Solozhenko VL, Turkevich VZ, Holzapfel WB (1999) J Phys Chem B 103: 2903±2905
54. Will G, Nover G, von der GoÈnna J (1998) SoRev High Pressure Sci Technol 7: 975±979
55. Fukunaga O (2000) Diamond Related Mater 9: 7±12
56. Gielisse PJ, Mitra SS, Plendl JJN, Grif®s RD, Mansur LC, Marshall R, Pascoe EA (1967)
Phys Rev 155: 1039±1046
57. Li PC, Lepie MP (1965) Am Ceram Soc 48: 277±278
58. Mineta S, Kohata M, Yasunaga N, Kikuta Y (1990) Thin Solid Films 189: 125±138
59. Weber A, Bringmann U, Nikulski R, Klages CP (1993) Surf Coat Technol 60: 493±497
60. Mieno M, Yoshida T (1990) Jpn J Appl Phys 29: L1175±L1177
61. Dworschak W, Jung K, Ehrhardt H (1994) Diamond Related Mater 3: 337±340
62. Kidner S, Taylor CA II, Clarke R (1994) Appl Phys Lett 64: 1859±1861
63. Doll GL, Sell JA, Taylor CA II, Clarke R (1991) Applications of diamond ®lms and
related materials. Mater Sci Monogr 73: 653±660
64. Becht JGM, Bath A, Hengst E, van der Put PJ, Schoonman J (1991) J Phys (Paris) C2:
617±624
65. Shapoval SY, Petrashov VT, Popov OA, Westner AO, Joder MD Jr (1990) CKC Lok Appl
Phys Lett 57: 1885±1886
66. Brafman O, Lengyel G, Mitra SS, Gielisse PJ, Plendl JN, Mansur LC (1968) Solid State
Commun 6: 523±526
67. Huong PV (1991) Diamond Related Mater 1: 33±41
68. Herchen H, Cappelli MA (1993) Phys Rev B47: 14,193±14,199
69. Gmelin (1991) Handbook of inorganic and organometallic chemistry, 8th edn, Boron
compounds, 4th suppl, Vol 3a, Springer Berlin, Heidelberg, New York
70. Schwetz KA, Lipp A (1979) Ber Dt Keram Ges 56: 1
71. Moeser L, Eidmann W (1902) Ber Dt Chem Ges 35: 535
72. Taylor KM (1959) US Pat 2 888 325
73. Ingles TA, Popper P (1960) The preparation and properties of boron nitride. In: Popper
P (ed) Special Ceramics. Heywood and Comp, London, p 144
74. Lipp A (1964) (1971)DP 1 153 DOS 1 792 741
75. Yokoi K, Ito K, Tokuda M (1995) Mat Transact JIM 36: 645
76. King EM (1967) Canadian Pat 675,630
77. Babl A, Geng H (1977) DP 1 667 538
78. Hagio T, Nonaka K, Sato T (1997) J Mat Sci Lett 16: 795
79. Chakrabartty S, Kumar S (1995) Trans Indian Ceram Soc 54: 48
80. Kawasaki T, Kuroda Y, Nishikawa H, Hara H (1997) DE 97 19,701,771
81. Kawasaki T, Nagahama K, Kuroda Y (1996) Jpn Kokai Tokkyo Koho JP 95±22830
82. Samsonov GV (1960) USSR-Pat 129 647
83. Slepstov VM, Samsonov GV (1960) Bornitrid Zhur Prikl Khim 34: 501
84. Wood AAR, Shears E C (1959) DAS 1 193 821
85. Su JY, Jha A (1996) J Mat Sci 31: 2265
86. Tagawa H, Itouji O (1962) Bull Chem Soc Japan 35: 1536
87. Conant LA, Hittle EF (1958) US Pat 2 834 650
88. Luberoff BJ (1966) US Pat 3 261 666
89. Mexer F, Zappner R (1921) Ber Dt Keram Ges 54: 560
90. Romanov VD, Samsonov GV, Nikitin DI (1959) USSR-Pat 120 509
91. Komatsu W, Morikawa T, Sedaka R (1988) Jpn Kokai Tokkyo Koho JP 63±62544
92. Canon KK Japan (1985) Jpn Kokai Tokkyo Koho JP 60±00826
93. Chen GQ, He XD, Han JC, Wood JV (2000) J Mat Sci Let 19: 81
94. Schwetz KA, Vogt G, Lipp A (1978) DOS 2 461 821 and US Pat 4 107 276
95. Zhang YM, Wang H, He XD, Han J, Du S (2001) Trans Nonferrous Met Soc China 11: 76
96. Warner TE, Fray DJ (2000) J Mater Sci 35: 5341
97. Yano M, Yap Y, Okamoto M, Onda M, Yoshimura M, Mori Y, Sasaki T (2000) Jpn J
Appl Phys 39: 300±302
98. Hirayama M, Shohno K (1975) J Electrochem Soc 122: 1671±1676
99. Adams AC, Capio CD (1980) J Electrochem Soc 127: 399±405
100. Gafri O, Grill A, Itzhak D, Inspector A, Avni R (1980) Thin Solid Films 72: 523±527
101. Choi BJ, Park DW, Kim DR (1995) J Mater Sci Lett 14: 452
102. David P, Mathiot A, Lulewicz JD, Narcy B (1994) Mater Res Soc Symp Proc 327: 233
103. DurmazucËar H, GuÈnduÈz G (2000) J Nucl Mater 282: 239
104. Francis R, Flint EP, Little AD (1961) WAL-766 41: 28
105. Akaishi M, Yamaoka N, Ueda F, Sasano M (1992) Jpn Kokai Tokkyo Koho JP
90±216,361
106. Torre M (1971) Swiss Pat Ch 19,700,227
107. Matje P, Reck F, Roehlinger HU (1995) DE 94±4,405,864
108. Faustinus F, Tani M (1998) Jpn Kokai Tokkyo Koho JP 96±301,051
109. Saito T, Honda F (2000) Wear 244: 132
110. Saito T, Honda F (2000) Wear 237: 253
111. Saito T, Imada Y, Honda F (1999) Wear 236: 153
112. Takahashi K, Georuku U (1995) Jpn Kokai Tokkyo Koho JP 93±170,390
113. Haraguchi F, Kanayama H, Kamiya S, Kawakami S, Michioka H, Fuwa Y (1997) Jpn
Kokai Tokkyo Koho JP 95±237,018
114. Ozaki K, Yamamoto K, Shibayama T (1997) DE 97±19,708,197
115. Ditter J, Allen R, Dorchester T, Harold T, Gerstein M, Christian JB (1967) Lubr Eng 23: 330
116. Kaldonski T (1997) Tribologia 28: 647
117. Kinoshita H, Nomura S, Mishima M (1992) EP 508,115
118. Hubacek M, Rehak B, Prnka T (1990) Hot pressing of hexagonal boron nitride. In:
Exner HE, Schumacher V (eds) Adv Mater Processes Proc Eur Conf 1st, p 653
119. Hara H, Ootsu K, Yoshino N, Nakamura Y (1997) Japan Jpn Kokai Tokkyo Koho JP
09,052,771
120. Schwetz KA (1999) Nichtoxidische Werkstoffe In: Taschenbuch Feuerfeste Werkstoffe
Vulkan Verlag Essen, p 130
121. Fister D (1985) Ceram Eng Sci Proc 6: 1305
122. Kurita S, Nakashima M (1999) Yoyuen oyobi Koon Kagaku 42: 99
123. Kanai T, Tanemoto H, Kubo H (1990) Jpn Kokai Tokkyo Koho JP 89±8548
124. Kusunose T, Choa Y-H, Sekino T, Niihara K (1999) Key Eng Mater 161/163: 475
125. Fukuda M, Sato Y, Ueki M (1993) Jpn Kokai Tokkyo Koho JP 92±137,596
126. Herrmann M, Schubert C (1996) DE 94±4,435,182
127. Takenouchi T, Sasaki H, Niihara K, Kususe H (1999) Jpn Kokai Tokkyo Koho JP
97±315,006
128. Rolander U, Weinl G (2000) EP 1,043,410
129. Kida O (1989) Jpn Kokai Tokkyo Koho JP 63±270,359
130. Sindlhauser P (1993) DE 91±4,143,344
131. Endo M, Ishikawa S (1997) Jpn Kokai Tokkyo Koho JP 09,278,457
132. Unuma H, Kikuchi A, Nakagawa T, Yamamoto T (1994) Hokkaido Kogyo Gijutsu
Kenkyusho Hokoku 60: 1
133. Lipp A (1969) DAS 1 289 712
134. Fubacheku M, Ueki M, Sakamoto K, Sato Y (1998) Jpn Kokai Tokkyo Koho JP 97±103
283
135. Yao X, Rai G (1995) GB 94±2514
136. Moessner B, Knoch H (1996) EP 95±116,831
137. Izumitani T, Shoda K, Nakyjima M, Kurita S (1993) Jpn Kokai Tokkyo Koho JP
92±12,958
138. Advanced Ceramic Corporation (2000) Product information

139. Iwamoto Y, Ebe A, Nishama S, Ogata K (1996) Jpn Kokai Tokkyo Koho JP 94±229,124
140. NuoFu C, Xingru Z, Lanying L, Xie X, Mian Z (2000) Mater Sci Eng B 75: 134
141. Moore AW (1990) J Cryst Growth 106: 6
142. Schwetz KA (1999) In: Ullmann's encyclopedia of industrial chemistry. Wiley-VCH,
Weinheim, p 1
143. Cameron DC (1996) J Chem Vap Depos 3: 133
144. Okano T, Yamashita H (1998) Jpn Kokai Tokkyo Koho JP 10,134,815
145. Iwaki T, Niikura J, Gyoten H, Hosoi A (1986) Jpn Kokai Tokkyo Koho JP 61±248,363
146. Shaffer G, Hill RF (1999) Advanced Ceramics Corporation USA US 5,898,009
147. Nakao K (2000) Jpn Kokai Tokkyo Koho JP 2,000,086,443
148. Bartnitskaya TS, Ostrovskaya TM, Fenochka BV (1999) Powder Metall Met Ceram 38:
152
149. Bartnitskaya TS, Ostrovskaya NF, Vereshchaka VM, Kurdyumov AV (1999) Powder
Metall Met Ceram 38: 240
150. Zhang G, Chen Z (1998) Gaofenzi Cailiao Kexue Yu Gongcheng 14: 94
151. Wakatsuki M, Ichinose K, Aoki T (1972) Mater Res Bull 7: 999
152. Vel L, Demazeau G, Etourneau J (1991) Mat Sci Eng B10: 149±164
153. Sumiya H, Iseki T, Onodera A (1983) Mater Res Bull 18: 1203
154. Gladskaya IS, Kremkova CN, Slesarev VN (1986) J Less-Common Met 117: 241
155. Sumiya H, Uesaka S, Satoh S (2000) J Materials Science 35: 1181±1186
156. Lux B, Kalss W, Haubner R, Taniguchi T (1999) Diamond Related Mater 8: 415±422
157. Wentorf RH Jr (1961) J Chem Phys 34: 809
158. DeVries RC, Fleischer JF (1972) J Cryst Growth 13/14: 88
159. Solozhenko VL, Turkevich VZ (1998) Diamond Related Mater 7: 43±46
160. Solozhenko VL, Turkevich VZ (1997) Mater Lett 32: 179±184
161. Endo T, Fukunaga O, Iwata M (1979) J Mater Sci 14: 1676
162. Lorenz H, Orgzall I (1995) Diamond Related Mater 4: 1046±1049
163. Lorenz H, Peun T, Orgzall I (1997) Appl Phys A 65: 487±497
164. Solozhenko VL, Turkevich VZ, Holzapfel WB (1999) J Phys Chem B 103: 8137±8140
165. Lorenz H, Orgzall I, Hinze E (1995) Diamond Related Mater 4: 1050±1055
166. Sumiya H, Tsuji K, Yazu S (1989) J Jpn Soc Powder Powder Metall 36: 752
167. Endo T, Fukunaga O, Iwata M (1981) J Mater Sci 16: 2227
168. Demazeau G, Biardeau G, Vel L (1990) CR Acad Sci II 310: 897
169. Demazeau G, Biardeau G, Vel L (1990) Mater Lett 10: 139
170. Gameza LM (2000) Poroshk Metall (Minsk) 22: 21±23
171. Susa K, Kobayashi T, Taniguchi S (1974) Mater Res Bull 9: 1443
172. Susa K, Kobayashi T, Taniguchi S (1976) High Temp High Press 8: 631
173. Kobayashi T, Susa K, Taniguchi S (1975) Mater Res Bull 10: 1231
174. Kobayashi T, Susa K, Taniguchi S (1977) Mater Res Bull 12: 847
175. Shulzhenko AA, Sokolov AN, Dub SN, Belyavina NN (2000) Sverkhtverd Mater 2:
30±35
176. Demazeau G, Gonnet V, Solozhenko V, Tanguy B, Montignaud H (1995) CR Acad Sci
Ser II 320: 419±422
177. Solozhenko VL, Le Godec Y, Mezouar M, Besson J-M, Syfosse G (1999) ESRF Highlights
92±93
178. Taniguchi T, Yamaoka S (2001) J Cryst Growth 222: 549±557
179. Mishima O, Yamaoka S, Fukunaga O (1987) J Appl Phys 61: 2822
180. Taniguchi T, Yamaoka S (2000) New Diamond Front Carbon Technol 10: 291±299
181. Sawaoka A, Soma T, Saito S (1974) Jpn J Appl Phys 13: 891±892
182. Komatsu T, Nomura M, Kakudate Y, Fujiwara S (1998) J Chem Soc Faraday Trans 94:
1649±1655
183. Komatsu T (1998) J Chem Soc Faraday Trans 94: 101±104
184. Dischler B, Wild C (1998) Low pressure synthetic diamond, Springer, Berlin,
Heidelberg, New York
185. Kalss W (1998) Doctoral Thesis, Vienna University of Technology

186. Sachdev H, Strauss M (2000) Diamond Related Mater 9: 614±619
187. Reinke S, Kuhr M, Kulisch W (1994) Diamond Related Mater 3: 341±345
188. Widany J, Weich F, KoÈhler T, Porezag D, Frauenheim T (1996) Diamond Related Mater
5: 1031±1041
189. Shtansky DV, Tsuda O, Ikuhara Y, Yoshida T (2000) Acta Mater 48: 3745±3759
190. McKenzie DR, McFall WD, Sainty WG, Davis CA, Collins RE (1993) Diamond Related
Mater 2: 970±976
191. McKenzie DR (1993) J Vac Sci Technol B11: 1928±1935
192. Dworschak W, Jung K, Ehrhardt H (1995) Thin Solid Films 254: 65±74
193. Uhlmann S, Frauenheim T, Stephan U (1995) Phys Rev B51: 4541
194. Reinke S, Kuhr M, Kulisch W, Kassing R (1995) Diamond Related Mater 4: 272
195. Kester DJ, Messier R (1992) J Appl Phys 72: 504
196. Reinke S, Kuhr M, Kulisch W (1995) Surf Coat Technol 74/75: 723±728
197. Weissmantel C, Bewilogua K, Dietrich D, Erler HJ, Hinneberg HJ, Klose S, Nowick K,
Reisse G (1980) Thin Solid Films 72: 19
198. Weissmantel C (1981) J Vac Sci Technol 18: 179±185
199. Kulisch W, Reinke S (1997) Diamond Films Technol 7: 105±138
200. HofsaÈss H, Ronning C, Griesmeier U, Gross M, Reinke S, Kuhr M, Zweck J, Fischer R
(1995) Nucl Instrum Methods Phys Res B 106: 153±158
201. Saitoh H, Yarbrough WA (1992) Diamond Related Mater 1: 137±146
202. Seidel KH, Reichelt K, Schaal W, Dimigen H (1987) Thin Solid Films 151: 243±249
203. Reisse G, Weissmantel S (1999) Thin Solid Films 355±356: 105±111
204. Satou M, Fujimoto F (1983) Jpn J Appl Phys 22: L171±L172
205. Setsuhara Y, Kumagai M, Suzuki M, Suzuki T, Miyake S (1999) Surf Coat Technol
116/119: 100±107
206. Boyen H-G, Widmayer P, Schwertberger D, Deyneka N, Ziemann P (2000) Appl Phys
Lett 76: 709±711
207. Kulisch W, Freudenstein R, Klett A, Plass MF (2000) Thin Solid Films 377/378: 170±176
208. McKenzie DR, McFall WD, Smith H, Higgins B, Boswell RW, Durandet A, James BW,
Falconer IS (1995) Nucl Instrum Methods Phys Res B 106: 90±95
209. Barth K-L, Neuffer A, Ulmer J, Lunk A (1996) Diamond Related Mater 5: 1270±1274
210. Barth K-L, Lunk A, Ulmer J (1997) SurfCoat Technol 92: 96±103
211. Goranchev B, Schmidt K, Reichelt K (1987) Thin Solid Films 149: L77±L80
212. SchuÈtze A, Bewilouga K, LuÈthje H, Kouptsidis S, JaÈger S (1995) Surface Coatings
Technol 74/75: 717±722
213. LuÈthje H, Bewilouga K, Daaud S, Johansson M, Hultman L (1995) Thin Solid Films 257:
40±45
214. Deng J, Wang B, Tan L, Yan H, Chen G (2000) Thin Solid Films 368: 312±314
215. Kulikovsky V, Shaginyan LR, Vereschaka VM, Hatynenko NG (1995) Diamond Related
Mater 4: 113±119
216. Ulrich S, Theel T, Schwan J, Ehrhardt H (1997) Surf Coat Technol 97: 45±59
217. Okamoto M, Yokoyama H, Osaka Y (1990) Jpn J Appl Phys 29: 930±933
218. Weber A, Bringmann U, Nikulski R, Klages CP (1993) Diamond Related Mater 2: 201
219. Chayahara A, Yokoyama H, Imura T, Osaka Y (1987) Jpn J Appl Phys 26: L1435±L1436
220. Yokoyama H, Okamoto M, Osaka Y (1991) Jpn J Appl Phys 30: 344±348
221. Ye M, Delplancke-Ogletree MP (2000) Diamond Related Mater 9: 1336±1341
222. Kuhr M, Reinke S, Kulisch W (1995) Surf Coat Technol 74/75: 806±812
223. Kuhr M, Reinke S, Kulisch W (1995) Diamond Related Mater 4: 375±380
224. Kouvetakis J, Patel VV, Miller CW, Beach DB (1990) J Vac Sci Technol A8: 3929±3933
225. Yuzuriha TH, Hess DW (1986) Thin Solid Films 140: 199±207
226. Karim MZ, Cameron DC, Hashmi MSJ (1991) Surf Coat Technol 49: 416±421
227. Karim MZ, Cameron DC, Hashmi MSJ (1994) Diamond Related Mater 3: 551±554
228. Saitoh H, Yarbrough WA (1991) Appl Phys Lett 58: 2228±2230
229. Saitoh H, Yarbrough WA (1991) Appl Phys Lett 58: 2482±2484
230. Matsumoto S, Zhang W (2001) New Diamond Front Carbon Technol 11: 1±10
231. Mendez JM, Muhl S, FarõÂas M, Soto G, Cota-Araiza L (1991) Surf Coat Technol 41:
422±426
232. Mendez JM, Muhl S, Andrade E, Cota-Araiza L, FarõÂas M, Soto G (1994) Diamond
Related Mater 3: 831±835
233. Matsumoto S, Nishida N, Akashi K, Sugai K (1996) J Mater Sci 31: 713±720
234. Zhang WJ, Matsumoto S (2000) Chem Phys Lett 330: 243±248
235. Zhang WJ, Matsumoto S (2000) Appl Phys A 71: 469±472
236. Yap YK, Aoyama T, Wada Y, Yoshimura M, Mori Y, Sasaki T (2000) Diamond Related
Mater 9: 592±595
237. Hollemann AF, Wiberg E (1995) Lehrbuch der anorganischen Chemie, 105th edn.,
Walter de Gruyter, Berlin
238. D'Evelyn MP, Taniguchi T (1999) Diamond Related Mater 8: 1522±1526
239. Sumiya H, Sato S, Yazu S (1994) US Pat US 5,332,629
240. Uesaka S, Sumiya H, Itozaki H, Shiraishi J, Tomita K, Nakai T (2000) SEI Tekunikaru
Rebyu 156: 18±23
241. Sumiya H, Uesaka S (2000) New Diamond Front Carbon Technol 10: 40
242. Anonymous (1988) Int Steel Met Mag 26: 516±8
243. Novikov NV, Shulzhenko AA, Petrusha IA (1987) Sverkhtverd Mater 6: 3/8
244. Hara A, Yatsu S (1986) Jpn Kokai Tokkyo Koho JP 61,209,958
245. Hara A, Yatsu S (1987) Jpn Tokkyo Koho JP 77±32,600
246. Hooper RM, Brookes CA (1986) Inst Phys Conf Ser 75: 907±917
247. Hara A, Yatsu S (1987) Japan Kokai Tokkyo Koho 62±07,151
248. Kurokawa Y (1985) Jpn Kokai Tokkyo Koho JP 60,210,573
249. Cerceau JM, Hall HT (1986) Eur Pat Appl EP 181,258
250. Hasegawa H (1989) Jpn Kokai Tokkyo Koho JP 01,131,067
251. Nishigaki K, Kikuchi N, Miwa K (1980) Jpn Kokai Tokkyo Koho JP 55,015,963
252. Collier MW, Yao X, Bowers BG (2000) US Pat US 6,140,262
253. Hooper RM, Guillou MO, Henshall JL (1991) J Hard Mater 2: 223±31
254. Yazu S, Kohno Y, Sato S, Hara A (1981) Mod Dev Powder Metall 14: 363±71
255. Hasegawa H (1988) Jpn Kokai Tokkyo Koho JP 63,035,455
256. Benko E, Stanislaw JS, Krolicka B, Wyczesany A, Barr TL (1999) Diamond Related
Mater 8: 1838±1846
257. Cerceau J-M, Boyat Y (1995) Fr Demande FR 2,713,223
258. Mitsusaka K, Yatsu S (1985) Jpn Kokai Tokkyo Koho JP 60,184,650
259. Kuratomi T (1985) Jpn Kokai Tokkyo Koho JP 60,230,956
261. Benko E, Morgiel J, Czeppe T (1997) Ceram Int 23: 89±91
262. Morgiel J, Benko E (1995) Mater Lett 25: 49±52
263. Benko E, Morgiel J, Czeppe T, Barr T (1998) J Eur Ceram Soc 18: 389±393
267. Kono Y, Hara A (1985) Jpn Kokai Tokkyo Koho JP 60,264,371
268. Nakai T, Hara A, Goto M (1987) Eur Pat Appl EP 223,585
269. Journal for grinding and cutting technologies (1997) Issue 123
270. Yokogawa M, Yokogawa K (1986) Werkstatt Betr 119: 788±794
271. Arnot RN, Fischbacher MJ (1995) VDI-Z 137(7/8): 36±38
272. Yokogawa M, Yokogawa K (1992) Int J Jpn Soc Precis Eng 26: 108±114
273. Nagata A (2001) Jpn Kokai Tokkyo Koho JP 2,001,062,732
274. Choe S, Sonn S (1997) Korea Patent KR 9,709,218
275. Viernekes N (1986) Sprechsaal 119: 1105±1106
276. Sato K, Nagata A (1990) Jpn Kokai Tokkyo Koho JP 02,055,272
277. Juricek V (1989) Czech Pat CS 257,551
278. Procyk B, Staniewicz-Brudnik B, Majewska-Albin K, Zawada A, Bieniarz P, Hohne D

(2000) Interceram 49: 308±314
279. Ichikawa K, Ando D (1989) Jpn Kokai Tokkyo Koho JP 01,179,763
280. Maekawa K, Yokoyama Y, Ohshima I (2001) Key Eng Mater 196: 133±140
281. Omori O, Furukawa H (2001) Jpn Kokai Tokkyo Koho JP 2,001,062,721
282. Yoshioka H, Kodama Y, Shigemura S, Funada T, Morimoto S (1989) US Pat US
4,855,019
283. Hagiuda Y, Takaya M, Matsunaga M, Yasunaga N (1990) Hyomen Gijutsu 41: 852±853
284. Osaka Diamond (1983) Jpn Tokkyo Koho JP 58,002,034
285. Takeuchi K, Senba T (2000) Jpn Kokai Tokkyo Koho JP 2,000,254,866
286. Wentorf RH (1962) J Chem Phys 36: 1990±1991
287. Mishima O, Tanaka J, Yamaoka S (1987) Science 238: 181±183
288. Mishima O, Era K, Tanaka J, Yamaoka S (1988) Appl Phys Lett 53: 962±964
289. Hara A, Yatsu S (1987) Jpn Tokkyo Koho JP 62,007,151
G. Petzow1, M. Herrmann2
1
Max-Planck-Institut fuÈr Metallforschung, Heisenbergstr. 5, 70569 Stuttgart, Germany
2
Fraunhofer Institut fuÈr Keramische Technologien und Sinterwerkstoffe,
Winterbergstrasse 28, 01277 Dresden, Germany
Silicon nitride ceramics is a generic term for a variety of alloys of Si3N4 with additional
compounds necessary for a complete densi®cation of the Si3N4 starting powder. They are
heterogeneous, multicomponent materials characterised by the inherent properties of the
crystalline modi®cations a and b of Si3N4 and the signi®cant in¯uence of the densi®cation
additives. With a view to ability of the a and b modi®cation to form solid solutions a-Si3N4
(ass) and b-Si3N4 (bss) solid solutions can be distinguished. Each group contains engineered
materials with interesting properties for special applications. Phase relations and micro-
structures determine the properties decisively. Composition of the phases, the distribution
of the grains, their aspect ratio and the grain boundary phase are pronounced
microstructural features. The formation of the microstructure strongly depends on the
one hand on the quality of the Si3N4 starting powders, which closely is related to
the chemistry of the production process, and on the other on the liquid phase sintering as
the most important step in the densi®cation route. The interrelation between pure Si3N4, the
densi®cation of the powder including the role of sintering additives, microstructural
engineering, physicochemical properties of the sintered Si3N4 ceramics (SSN, GPSN, HPSN,
HIP-SSN, HIP-SN) are described in more detail and compared to reaction bonded Si3N4
ceramics (RBSN), which are produced by nitridation of silicon powders.
Keywords: Silicon nitride ceramics, Phase relations, Processing, Microstructure, Properties
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2 Crystalline Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.1 Crystal Structures and Inherent Properties . . . . . . . . . . . . . . . . 53
2.2 a/b Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3 Phase Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.1 Thermodynamics and Phase Diagrams . . . . . . . . . . . . . . . . . . . 60
3.1.1 Si-N-O System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.1.2 Si-N-O-Additives Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.2 b-Si3N4 Solid Solutions (bss) . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.2.1 Extended b-Si3N4 Solid Solutions . . . . . . . . . . . . . . . . . . . . . . . 64
3.2.2 Low b-Si3N4 Solid Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.3 a-Si3N4 Solid Solutions (ass) . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Structure and Bonding, Vol. 102

Ó Springer-Verlag Berlin Heidelberg 2002
48 G. Petzow á M. Herrmann
4 Si3N4 Powders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.1 Powder Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.2 Powder Characterisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5 Consolidation of Si3N4 Powders . . . . . . . . . . . . . . . . . . . . . . . . 78

5.1 Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.1.1 Processing Additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.1.2 Sintering Additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.2 Shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.3 Densification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.3.1 Densification Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.3.2 Densification Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.3.3 Influence of Densification Parameters . . . . . . . . . . . . . . . . . . . 87
5.3.3.1 Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.3.3.2 Influence of Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.3.3.3 Influence of Powder Bed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.3.4 Densified Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.4 Finishing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.5 Precursor-Derived Si3N4 Ceramics . . . . . . . . . . . . . . . . . . . . . . 91
6 Microstructures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.1 Development of Microstructures in bss Ceramics . . . . . . . . . . . 91
6.1.1 Influence of Starting Powder . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.1.2 Influence of Sintering Additives . . . . . . . . . . . . . . . . . . . . . . . . 95
6.1.3 Influence of Sintering Conditions . . . . . . . . . . . . . . . . . . . . . . . 95
6.1.4 Microstructural Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.1.4.1 bss Grains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.1.4.2 Secondary Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.1.4.3 Amorphous Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.2 Development of Microstructures in ass Ceramics . . . . . . . . . . . 101
6.3 Development of Microstructures in ass/bss Ceramics . . . . . . . . 102
6.4 Characterisation of Microstructure . . . . . . . . . . . . . . . . . . . . . . 103
6.5 Microstructure/Property Relation . . . . . . . . . . . . . . . . . . . . . . . 104
7 Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.1 Physical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.2 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
7.2.1 Room Temperature Properties . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.2.2 High Temperature Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 113
7.3 Chemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
7.3.1 Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
7.3.2 Interaction with Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.3.3 Corrosion in Liquids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.4 Colours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Silicon Nitride Ceramics 49
8 Reaction Bonded Silicon Nitride (RBSN) . . . . . . . . . . . . . . . . . 133

8.1 Production and Nitridation Process . . . . . . . . . . . . . . . . . . . . . 133
8.2 Microstructure and Properties . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.3 Sintered Reaction Bonded Si3N4 (SRBSN) . . . . . . . . . . . . . . . . . 136
9 Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
9.1 Si3N4/TiCN Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
9.2 Si3N4/SiC Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
9.3 Si3N4/Si3N4 Composites (Seeded Materials) . . . . . . . . . . . . . . . . 142
10 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
11 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
12 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Abbreviations
GPSN gas-pressure sintered Si3N4 ceramics

HPSN hot pressed Si3N4 ceramics
HIP-SN encapsulated hot isostatically pressed Si3N4 ceramics
HIP-SSN presintered hot isostatically pressed Si3N4 ceramics
(also Sinter-HP-SN)
Sinter-HIP-SN presintered hot isostatically pressed Si3N4 ceramics
SSN sintered Si3N4 ceramics
RBSN reaction bonded Si3N4 ceramics
SRBSN sintered reaction bonded Si3N4 ceramics
HIP-SRBSN hot isostatically pressed SRBSN ceramics
AAS atomic absorbtion ¯ame spectroscopy
AES atomic emission spectroscopy
CAPLUS data bank of Chemical Abstracts
CVD chemical vapour deposition
E Young's modulus
HRTM high resolution transmission electron microscopy
HT high temperature
HIP hot isostatic pressing
KIC fracture toughness (critical stress intensity factor)
M interstitial metal in ass (Eq. 6)
i
MiV cations with the charge Vi
Me metals
MS mass spectroscopy
OES optical emission spectroscopy
RE rare earth metals
RT room temperature
SEM scanning electron microscopy
SHS self-propagating high temperature synthesis

Tg transition temperature of the glassy phase
TEM transmission electron microscopy
XiZi anion with the charge Zi (usually O2) and N3))
XRF X-ray ¯uorescence analysis
XRD X-ray diffraction
XPS X-ray photoelectron spectroscopy
Z atomic number
a ¯aw size (Sect. 7.2)
a thermal diffusivity (Sect.

7.1)
at% (Mi) atomic % of cation MiVi with the charge Vi
at% (Xi) atomic % of anion XiZi with the charge Zi
equ.% equivalent %
m x á V+
n number of components (Sect. 3)
n degree of nitrogen substitution in ass (Eq. 6; Sect. 3.3; 6.2; 7.1)
n growth rate exponent (Sect. 7.2)
V+ charge of cation M
x amount of stabilising cation M
z degree of substitution in bss (Si6±zAlzN8±zOz)
a thermal expansion coef®cient (Sect. 2; 7.1)
ass solid solutions based on the a-Si3N4 modi®cation
bss solid solutions based on the b-Si3N4 modi®cation
e_ creep rate
e deformation during creep
m Poisson ratio
r bending or tensile strength
f zeta-potential
1
Introduction
Silicon nitride has the composition Si3N4 and its chemical bonding is
predominantly covalent. Si3N4 represents the backbone of silicon nitride
ceramics, a class of ceramic materials which, because of their exceptional pro®le
of properties, are gaining increasing acceptance in engineering applications.
Natural sources of Si3N4 are extremely rare and are a mineralogical curiosity
that has no signi®cance as a raw material. The mineral Nierite is named after
the pioneer of mass spectroscopy, A.O.C. Nier, and is observed as an inclusion
in meteorites. It consists chie¯y of the a modi®cation which is interspersed
with inclusions of b-Si3N4 whiskers. The ratio of a to b in the different ®nds
can show considerable variance [1, 2].
All silicon nitride ceramics are derived from synthetic materials, exclusive-
ly. The ®rst report on the synthesis of Si3N4 was in 1859 by Sainte-Claire
Deville and WoÈhler [3]. Among the problems of greatest concern to chemists
in those days was the utilisation of atmospheric nitrogen for agricultural and
industrial purposes. In particular, there was a need for a highly effective
fertiliser to improve agricultural yields. In this connection, the bonding of

nitrogen to silicon became of great interest since it was discovered that silicon
is able to take up nitrogen at high temperatures and to release it in the form of
ammonia. This led to many investigations into the synthesis of nitrogen-
silicon compounds [3±7]. It might be mentioned that all the synthesis methods
used today in industry were developed in principle in the 19th century.
Unfortunately the dream of a new effective fertiliser from silicon nitride was
not ful®lled. Because of its high chemical and thermal stability it does not
decompose easily, which is an important precondition for fertilisers. The goal of
synthesising nitrogen-silicon compounds ®nally lost its appeal when simpler and
cheaper processes for the synthesis of ammonia were discovered [8]. On the other
hand, the high stability ± undesirable in those days ± is a signi®cant feature of
silicon nitride as a basis for advanced ceramics which, because of their peculiar
combination of properties, ®ll a gap in the spectrum of engineering materials.
Soon after 1950 silicon nitride was being employed as refractories and nozzles in
jet engines and rockets [9, 10] which are produced by the reaction bonding route,
i.e., by direct nitridation of compacted silicon powder parts. At the same time,
comprehensive studies were conducted on the dependence of the properties of
these particular silicon nitride materials on process variables [11, 12].
The reaction bonded materials are not completely dense. For optimal
utilisation of the inherent good properties of silicon nitride, the powders must
be fully densi®ed to compacts. This was not possible in those days because
suitable densi®cation techniques were lacking. Since Si3N4 under ambient
conditions does not melt but decomposes into the elements at about 2120 K a
densi®cation by melting was not feasible. The production of high-density
silicon nitride materials became possible only after the techniques of
densi®cation had matured. This happened in the early 1960s with the
introduction of hot pressing, originally developed in powder metallurgy for
densi®cation of high-melting metallic powders under the simultaneous
application of high pressure and high temperature. The hot pressing of
silicon nitride powders led to further progress. Deeley, Herbert and Moore
reported on sintering studies of silicon nitride powders with small amounts of
various oxide and nitride additives and achieved bodies of almost theoretical
density by hot pressing [13]. So it is not surprising that the ®rst publication on
fully densi®ed silicon nitride qualities did not appear in a chemical or a
ceramic periodical but in the journal ``Powder Metallurgy''.
Whilst today the densi®cation of silicon nitride powders is possible by
normal sintering without additional pressure, the additives in general are still a
must. Only with high pressures and temperatures is densi®cation without
additives possible. As a consequence the resulting materials are not pure
silicon nitride qualities but mixtures or alloys of silicon nitride with additional
compounds necessary for a complete densi®cation. These materials are
generally named SILICON NITRIDE CERAMICS, which is a generic term used
for a variety of types having different compositions. In other words: silicon
nitride ceramics are multicomponent mixtures or alloys of Si3N4.
About one hundred years after Si3N4 was synthesised for the ®rst time,
dense compacts could ®nally be prepared from its powders. Thus did an
obscure chemical compound become an engineering material. The signi®cant

stages of its development are mentioned in Fig. 1.
The spectrum of properties provided by dense Si3N4 ceramics has proved to
be of exceptional value in numerous applications and improvements that were
previously not considered feasible for lack of a suitable material. This has
Fig. 1. Some historical data for the development of Si3N4 ceramics; a chemical compound
becomes an advanced ceramic
generated a large body of systematic scienti®c and technological investigations

that have led to a deeper understanding of this class of materials. The results
and insights thereby gained are re¯ected in numerous reports and publications
whose number, based on the CAPLUS data bank of Chemical Abstracts, is
approaching ®fty-®ve thousand. The continuing rise in the number of relevant
publications on the subject of Si3N4 ceramics is illustrated in Fig. 2. But as yet,
no comprehensive compilation of the information on this interesting and
diverse class of materials is available in text book form, although several
helpful review articles, summaries of detailed aspects and data collections can
be found in the literature [14±21].
2
Crystalline Modifications
Three crystalline modi®cations of Si3N4 are known (Fig. 3a±c) [22±28]. The
corresponding data are listed in Table 1 together with some characteristic
properties. Whilst a- and b-Si3N4 are to be produced under normal nitrogen
pressure, the recently observed c modi®cation is originated at high pressure
and temperature [23, 29].
2.1
Crystal Structures and Inherent Properties
The a modi®cation dominates in the commonly produced Si3N4 powders

(Sect. 4).
The lattice parameters of the a phase depend on the oxygen content
dissolved in the structure. With increasing amounts of oxygen the a-parameter
Fig. 2. The number of publications on Si3N4 ceramics since 1965 (source: Chemical
Abstracts; CAPLUS)
Fig. 3a, b. Projection of the crystal structures of a, b, c (a, b, c)
decreases to 0.775 nm and the c-parameter increases to 0.5625 nm [22, 45].

From earlier investigations it is assumed that a is an oxide nitride with an
oxygen content of 0.9±1.48 wt% [45]. (It should be taken into account that in
ceramic literature `òxide nitride'' is generally called `òxynitride''.) However,
more detailed investigations clearly demonstrate the existence of a-Si3N4 with
oxygen solubility lower than 1 wt% in the absence of other impurities; in
monocrystals up to 0.3 wt% have been measured [22]. The melting point of a-
Si3N4 was detected at 2560 K under a nitrogen pressure of 120 MPa [46]. This
value is more reasonable than that determined as 2200 K and 3.58 GPa [47].
Table 1. Crystal structures and properties of Si3N4 modi®cations

a-Si3N4 b-Si3N4 c-Si3N4
Space group P31c; No. 159 [22] P63; No. 173 [22, 25] Fd3m [23, 26]
Lattice parameter
a, nm 0.7818(3) [22] 0.7595(1) [25] 0.7738
c, nm 0.5591(4) 0.29023(6)
Density, g cm)3 [22] 3.18 3.20 4,0 [26], 4.12 [29]
Coordinates of atoms [24], [30] [25], [45] Table 2
Hardness (110) plane: 2250 (HV 0.3) (100) plane: 2100 (HV 0.025) >30 GPa [23]
(001) plane: 2200 (HV 0.3) [31] (001) plane: 1326 (HV 0.025) [34]
(100) plane: 3660 (HV 0.3)
(001) plane: 2782 (HV 0.3) [32] (100) plane: 2610 (HV 0.3)
(100) plane: 2830 (HV 0.3) (001) plane: 1642 (HV 0.3) [32]
(001) plane: 2890 (HV 0.3) [33]
Fracture toughness, MPa m1/2 1.9±2.8 [ 31]
Elastic constants [calculated after 35] [36]
Ex, GPa 341 280
Ez, GPa 343 540
mxy 0.30 0.35
mzx 0.29 0.25
Gxz, GPa 132 124
Standard molar enthalpies of formation )828.9 3.4 [37] )827.8 2.5 [37]

D f Hm KJ mol)1 )850.9 22.4 [38] )852.0 8.7 [38]
Coef®cient of thermal expansion 0±1000 °C, 10)6/K 3.64 [39] 3.39 [39]
aa (0±500 °C) 2.72 2.01
ac-(0±500 °C) 3.14 [18a] 2.84 [18a]
aa (0±1000 °C) 3.61 3.23
ac-(0±1000 °C) 3.70 3.72
Thermal conductivity, W(mK))1 110±150 [40] [001] Direction 180
[100] Direction 68 [41]
Si-diffusion, cm2 s)1 [42] 1673 K: 0.45 á 10)15
1773 K: 1 á 10)15
Activation energy = 197 KJ mol)1
N-diffusion coef®cient, cm2 s)1 [43] DNa = 1.2 á 10)12 DNb = 5.8 á 106
Exp()233 KJ mol)1/RT) Exp()777 KJ mol)1/RT)
Refractive index [44] no = 2.03 no = 2.02
55
ne = 2.02 (optically negative) ne = 2.04 (optically positive)
These data seem unlikely because Si3N4 can be sintered at 1950±2000 °C

(2225±2275 K) and 1±10 MPa nitrogen pressure without melting.
The b-modi®cation is the main constituent of the majority of the Si3N4
ceramics. Different space groups were observed: the centrosymmetric P63/m
[45] and the corresponding non-centrosymmetric P63 [25, 35]. Detailed
investigations on single crystals reveal P63. The atomic coordinates in the unit
cell remain nearly constant up to 1633 K [48]. Deviations from the observed
cell dimensions are presumably caused by aluminium and oxygen impurities
(Sect. 3). The solubility of oxygen in the b structure is up to 0.258 wt% in the
absence of other elements [49].
In perfect monocrystals of b-Si3N4 thermal conductivities up to
320 W(mK))1 are possible as shown by calculations [40]. This intrinsic
thermal conductivity of Si3N4 is very near that of AlN [319 W(mK))1]. The
deviations from these optimal values to the considerably lower values
[<150 W(mK))1] obtained for real sintered Si3N4 ceramics are caused by the
grain boundary phases and their distribution [40]. It is supposed that the
thermal conductivity of Si3N4 can be improved by reduction of the oxygen
dissolved in the lattice, as known for AlN ceramics. Compared to AlN the
lower diffusion coef®cient of oxygen in b-Si3N4 causes a much longer heat
treatment to attain oxygen-free materials [40].
The cubic c-modi®cation has been recently observed under a pressure of
15 GPa and temperatures above 2000 K by the laser heating technique in a
diamond cell [23] and in shock-wave compression experiments with pressures
>33 GPa at 1800 K and >50 GPa at 2400 K [29]. This modi®cation is often
designated as the c-modi®cation in the literature in analogy to the cubic boron
nitride (c-BN). It has a spinel-type structure in which two silicon atoms are
octahedrally coordinated by six nitrogen atoms, one silicon atom is coordi-
nated tetrahedrally by four nitrogen atoms (Fig. 3c). The atomic coordinates
for the cubic modi®cation are given in Table 2. From calculations it is shown
that this structure should have a high hardness similar to that of diamond and
c-BN [23].
2.2
a/b Relations
Both modi®cations are based on SiN4 tetrahedra connected at the corners. The
dimension of the SiN4 tetrahedron is very similar to that of SiO4. Therefore
Table 2. Atomic co-ordinates of c-Si3N4 with space group Fd3m; a = 0.77381(2) nm [Int.
Tables No. 227) [26]
Position x/a y/a z/a
Sitet in 8(a) 1/8 1/8 1/8

Sioct in 16(d) 1/2 1/2 1/2
N in 32(e) 0.25968(1) 0.25968(1) 0.25968(1)
some oxide nitrides exist containing both, SiO4 and SiN4 tetrahedra [50]. In
the a- as well as in the b-modi®cation each nitrogen belongs to 3 tetrahedra,
i.e., the three-dimensional arrangement is a network of tetrahedra. Whilst in b
only one layer of SiN4 tetrahedra exists, a-Si3N4 has two layers shifted with
respect to each other. This leads to a doubling of the c-parameter in the
a-Si3N4 unit cell (Figs. 3a and b). The defects in the a- as well as in the
b-structure are similar; in both cases the most commonly observed disloca-
tions have a <0001> type Burgers vector [22, 51, 52].
The a-phase is metastable during sintering (e.g. at 1673±2273 K and 0.1 to
100 MPa N2 pressure) and transforms irreversibly to b-Si3N4. Extensive
studies have been made to determine the kinetics and mechanisms of the
transformation [22, 39, 53]. A heat treatment of sintering additive-free a-Si3N4
powder at 1900°C results in a slow phase transformation and an epitaxial
orientation of the a- and b-modi®cations in the c-direction [53a]. A wide range

of data about the difference of the standard molar enthalpy of formation Df Hm
of the two modi®cations are found in the literature [25, 54], earlier data
indicating that a-Si3N4 has a lower standard molar enthalpy of formation
by several KJ in comparison to b [25] and therefore a-Si3N4 is the stable
modi®cation at room temperature [54]. These quite large deviations are very
likely caused by the low qualities of the available Si3N4 powders containing
different amounts of oxygen, carbon and other impurities, and in all cases the
investigations were done on a/b mixtures. In recent investigations on high
purity powders also different assumptions have been made on the chemical
bonding of the impurities, e.g., carbon impurities were considered as SiC,

oxygen as SiO2 [37, 38]. These investigations have shown Df Hm of the pure a
)1
and b to differ by only 1 KJ mol , which is lower than the error of

determination of Df Hm (Table 1).

The most recent measurements of Df Hm show that dissolved oxygen in the
a-structure leads to a slight destabilisation [38]. The observed destabilisation
contradicted the previous work reporting that the a-modi®cation is stabilised

by dissolved oxygen [45]. The recently observed values for Df Hm are more
negative than those given in compilations of chemical thermodynamic
functions for inorganic substances, e.g., the JANAF tables which recommend
Df Hm
744:75 29:3 KJ mol)1 [55].
The b-Si3N4 structure exhibits channels parallel to the c-axis which are
about 0.5 nm in diameter causing the relative high diffusion coef®cient
compared to a which do not show such channels (Table 1). These channels are
changed into voids with seven nearest nitrogen neighbours in the a structure.
The hardness values of the crystals are given in Table 1. These data show
that b has a much lower hardness than a. Also a high anisotropy of hardness
and elastic properties are found for b, which is less pronounced in a.
Ab-initio calculations of the a- and b-Si3N4 structures show that the elastic
constants of the a-structure are less anisotropic than those of the b-structure
[35]. For both modi®cations the infrared absorption Raman spectra are
reviewed [18b, 35]. Both a and b exhibit a more or less pronounced solid
solubility for other elements. This phenomenon is treated in Sect. 3.
Table 3. Comprehensive overview of investigated Si3N4 ceramic systems and quasi-systems
Systems and quasi-systems References
2-component systems
Si-N [60±62]
[Structure and Bonding,
Vol. 101 (2002) p 1±58]
Si-O [61, 63±67]
3-component system and quasi-binary
Si-N-O [7, 60, 61, 66±69]
Si3N4-3(SiO2)
4-component systems and quasi-ternaries
Si-N-O-Al [70±88]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)
Si-N-O-Be [70, 82, 83, 89±91]
Si3N4-4(Be3N2)-6(BeO)-3(SiO2)
Si-N-O-C [60, 61, 67, 92±96]
Si-N-O-Ca [82]
Si3N4-2(Ca3N2)-6(CaO)-3(SiO2)
Si-N-O-Ce [97, 98]
Si3N4-4(CeN)-2(Ce2O3)-3(SiO2)
Si-N-O-Dy [99]
Si3N4-4(DyN)-2(Dy2O3)-3(SiO2)
Si-N-O-Er [99]
Si3N4-4(ErN)-2(Er2O3)-3(SiO2)
Si-N-O-Gd [99±101]
Si3N4-4(GdN)-2(Gd2O3)-3(SiO2)
Si-N-O-La [99, 101, 102]
Si3N4-4(LaN)-2(La2O3)-3(SiO2)
Si-N-O-Mg [68, 69, 82, 84, 91, 98, 103±105]
Si3N4-2(Mg3N2)-6(MgO)-3(SiO2)
Si-N-O-Nd [99±101, 106, 107]
Si3N4-4(NdN)-2(Nd2O3)-3(SiO2)
Si-N-O-Sm [99]
Si3N4-4(SmN)-2(Sm2O3)-3(SiO2)
Si-N-O-Th [82, 108]
Si3N4-Th3N4-3(ThO2)-3(SiO2)
Si-N-O-Y [82, 99, 109±112]
Si3N4-4(YN)-2(Y2O3)-3(SiO2)
Si-N-O-Yb [99, 113±116]
Si3N4-4(YbN)-2(Yb2O3)-3(SiO2)
Si-N-O-Zr [82, 109, 114]
5-component systems and quasi-quaternaries
Si-N-O-Al-Be [70, 82]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-6(BeO)-2(Be3N2)
Si-N-O-Al-Ca [82, 117±128]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-6(CaO)-2(Ca3N2)
Si-N-O-Al-Ce [129±131]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Ce2O3)-4(CeN)
Si-N-O-Al-Dy [100, 132±138]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Dy2O3)-4(DyN)
Si-N-O-Al-Er [100, 136]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Er2O3)-4(ErN)
Systems and quasi-systems References
Si-N-O-Al-Gd [100, 133, 136, 139]

Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Gd2O3)-4(GdN)
Si-N-O-Al-La [130, 131, 140]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(La2O3)-4(LaN)
Si-N-O-Al-Li [71, 91, 141]
Si-N-O-Al-Mg [71, 82, 83, 98, 109, 142, 143]
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-6(MgO)-2(Mg3N2)
Si-N-O-Al-Nd [88, 99±101, 107, 126, 132, 133,
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Nd2O3)-4(NdN) 135, 136, 140, 144±152]
Si-N-O-Al-Sm [100, 126, 132±134, 136, 138,
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Sm2O3)-4(SmN) 145, 150, 153, 154]
Si-N-O-Al-Y [82, 83,88, 98, 106, 110, 111,
Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Y2O3)-4(YN) 127, 132, 133, 135, 147, 148,
150, 152, 154±170]
Si-N-O-Al-Yb [100, 116, 135, 136, 138, 147,

Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Yb2O3)-4(YbN) 150, 171]
Si-N-O-Al-Zr [71, 82, 172]
Si-N-O-C-Ti [173]
Si-N-O-Ca-Mg [174]
Si3N4-2(Ca3N2)-2(Mg3N2)-3(SiO2)-6(CaO)-6(MgO)
Si-N-O-La-Y [175]
Si3N4-4(LaN)-4(YN)-2(La2O3)-2(Y2O3)-3(SiO2)
Si-N-O-Mg-Y [176]
Si3N4-2(Mg3N2)-4(YN)-2(Y2O3)-6(MgO)-3(SiO2)
6-component systems
Si-N-O-Al-B-C [177]
Si-N-O-Al-Ca-La [178]
Si-N-O-Al-Ca-Mg [178]
Si-N-O-Al-Sm-Y [154]
Si-N-O-Ar-C-H [179]
7-component systems
Si-N-O-Al-C-Mg-Y [180]
3
Phase Diagrams
Phase diagrams are concise plots of equilibrium relationships for understanding
heterogeneous materials. By their nature they only represent thermodynamic
equilibrium conditions, relating the physical state of a mixture to the number of
substances of which it is composed and the environmental conditions imposed
on it. In other words: phase diagrams are comprehensive descriptions of the
constitution of matter as far as relations between different phases are concerned.
The principles of phase equilibria are central to an understanding of many
scienti®c and technological disciplines and provide important guidelines in the
production, processing, and applications of materials in general and especially
of Si3N4 ceramics, which are related to heterogeneous systems of higher order,
i.e., containing more than two components n.
In Table 3 all the systems are compiled on which the known Si3N4 ceramics
are based, arranged according to the number of elements (components)
involved. From the binary and ternary subsystems (n £ 3) only Si-N, Si-O and
Si-N-O are listed, because they are the only ones of all the subsystems which
are existent in the higher order systems of all Si3N4 ceramics. All other side
systems (subsystems), whose components may affect the phase diagrams and
properties of Si3N4 ceramics to varying degrees are not depicted in the interest
of space. The pertinent metal oxide (Si-Me-O), metal nitride (Si-Me-N), and
metal oxide nitride (Me-N-O) systems are generally more or less well-known
and described elsewhere [56±59].
3.1
Thermodynamics and Phase Diagrams
In general, truly thorough investigations of systems are limited to a small

number of areas of special scienti®c or technological interest. Often, there are
only a few experimental results and thermodynamic estimates, and in some
cases, the treatment and assumptions are purely speculative. This applies in
particular for the vapour phase which is of great importance with respect to the
operating conditions on the materials, such as sintering behaviour, vaporisa-
tion, oxidation, corrosion, etc. Almost all the thermodynamically calculated
phase diagrams include the vapour phase which gives signi®cant additional
information to that gained by experiments, usually carried out under reducing
conditions in excess nitrogen atmosphere. But in some cases the vapour phase
is not or only partially considered which to some extent explains the different
data and inconsistencies in the literature. In Table 3 the literature is compiled
which includes descriptions derived mainly from controlled nitrogen (N2)
conditions or is a calculated result at a ®xed nitrogen potential.
3.1.1
Si-N-O System
Silicon and Si3N4 have a high af®nity for oxygen, which it contains both in
solution and in adsorbed form. Analytically it is possible to distinguish between
the two oxygen species, bulk or surface oxygen, respectively (Sect. 4.2). To
understand the role of nitrogen and oxygen in the phase relations of Si3N4
ceramics, the binary systems Si-N and Si-O as well as the ternary system Si-N-O
are of essential value. There exist both experimental data and solid thermo-
dynamic calculations, which are based on critical evaluations of several of those
experimental investigations and predictions using thermodynamic modelling.
For the side system N-O the gas phase only is considered. The phase relations in
the system Si-N-O depend sensitively on the partial pressures of N2, O2, SiO and
temperature as shown in Fig. 4 for example. Therefore, the knowledge of the
partial pressures of the vapour species N2, O2 and SiO is a valuable tool.
In the system Si-N-O a quasi-binary relation exists between the two stable
compounds Si3N4 and SiO2 of the two binary systems Si-N and Si-O.
In respect to Si3N4 ceramics the quasi-binary system Si3N4-SiO2 is of
greatest importance (Fig. 5). However, if the vapour phase is included in the
Fig. 4. Potential Si-N-O phase diagram at 1500 K including SiO isobares [66]; the pressures
are given in bar
phase equilibria, this system can no longer be treated as a quasi-binary system

since the composition of the vapour phase and the metallic melt is not
restricted to compositions within the quasi-binary as is taken into account
above 2100 K in Fig. 5. The decomposition of the ternary phase Si2N2O has
been calculated to occur at 2137 K, according to the equation:
Si2 N2 O , N2 SiO Sil 1
As can be concluded from Eq. (1), the decomposition will be signi®cantly re-
duced with increasing N2 pressure. Therefore at high N2 pressure, the condensed
phases can be treated as a quasi-binary system up to higher temperatures.
3.1.2
Si-N-O-Additives Systems
As mentioned before, the systems on which (technologically useful) Si3N4

ceramics are based, are without exception multicomponent, having n > 3
because in the production process (Sect. 5) sintering additives are necessary,
which remain in the ®nal product. The phase relations of such systems and
their dependence on temperature and pressure can be represented in
polyhedrons of three- or multi-dimensional order by introducing restricting
conditions. Therefore phase diagrams of four-component systems become
extremely cumbersome and those of more than four components are complex
to the point of becoming intractable.
Fortunately the graphical representation of many phase relations in the
Si3N4 ceramics can be simpli®ed by conditions which lower the degree of
freedom. Such a condition exists for reciprocal systems in which a double
exchange reaction occurs
AX BY , AY BX 2
Fig. 5. Temperature-concentration section (isopleth) through the Si-N-O phase diagram

from SiO2 to Si3N4 [69]. Below 2114 K it is a quasi-binary system. G = gas phase; LS = oxide
nitride liquid; LM = metallic liquid
where A and B are cations and X and Y anions. This reaction reduces the
number of independent components to three and, thus, a two-dimensional
representation of the concentrations is feasible. The requirement of electro-
neutrality and the constant valence state of the components, labelled in
equivalent percent (equ.%), result in a representation of the system on a
square plane (isobaric and isothermal) with the four compounds at the
corners, and the system can be treated as a quasi-ternary system with the same
topological elements as all ternary systems. Five-component systems (quinary
systems) can be treated in an analogous fashion if conditions between the
involved compounds are quaternary and valency states of the elements are
®xed. The resulting reciprocal salt system is a triangular prism, the JaÈnecke
prism, named after E. JaÈnecke, a pioneer in the application of phase diagrams
[181]. The conversion of at% into the less familiar equ.% is given by:
V
!
V Vi at%Mi i
equ:%Mi i 100 P V
n
i1 Vi at%Mi
i
Z
!
Z Zi at%Xi I
and equ:%Xi i 100 Pm Z
3
i1 Zi at%Xi i
V
Mi i are cations with the charge Vi ; and at% (Mi) are the atomic % of the
Vi Zi
cation Mi : Xi are anions the charge zi (usually O2) and N3)); and at% (Xi)
is the atomic % of the anion XiZi .
Fig. 6a, b. Derivation of quasi-subsystems in higher order systems (schematic). a) The quasi-
ternary subsystem Si3N4-AlN-Al2O3-SiO2 in the 4-component system Si-N-O-Al. b) The
quasi-quaternary subsystem Si3N4-AlN-Al2O3-SiO2-Y2O3-YN (JaÈnecke prism) of the
5-component system Si-N-O-Al-Y
In Fig. 6a the derivation of the square diagram Si3N4-AlN-Al2O3-SiO2 in the

quaternary system Si-N-O-Al is schematically explained and in Fig. 6b
the JaÈnecke prism Si3N4-AlN-Al2O3-SiO2-Y2O3-YN in the quinary system
Si-N-O-Al-Y is derived.
The Si3N4 ceramics which are arranged in Table 3 according to the number
of the involved components show remarkably different characteristics in their
solution behaviour for additional elements. With a view to their ability to form
solid solutions two groups can be distinguished:
1) b-Si3N4 solid solutions (bss) and
2) a-Si3N4 solid solutions (ass).
Each group contains engineered materials for special applications. Out of
each group examples will be discussed in the following.
3.2
b-Si3N4 Solid Solutions (bss)
The solid solution capacity of the b-phase differs drastically depending on the
dissolved elements; remarkable differences in the processing behaviour and
properties are the consequence. Therefore it seems useful to subdivide the

solubility of b into extended and low solid solutions.
3.2.1
Extended b-Si3N4 Solid Solutions
Aluminium and beryllium can be dissolved extensively in b-Si3N4. In these

cases the Si4 and N3) ions may be exchanged by the other metals and oxygen,
respectively. The exchange appears as a coupled substitution for reason of the
valence electron balance occasioned by the stoichiometry. As can be concluded
by the high number of citations (Table 3) the system Si-N-O-Al, usually named
SiAlON, has found considerable interest and has been more intensively studied
than other Si3N4 ceramics. It has a certain technological and scienti®c
signi®cance and therefore is chosen as an example for the extended b-solid
solution group. In the many publications on these materials, several
abbreviations are used to designate the different phases. Here we follow the
recommendations given by IUPAC [182]. Incidentally, the expression SiAlON
is often used as a generic name for Si3N4 ceramics, even when they do not
contain Al. To avoid confusion, the relevant expressions a- and b-solid
solutions are to be preferred.
All compositions in the system can be represented by a regular tetrahedron
(Fig. 6a). Since the elements have their normal valencies (Si4+, Al3+, O2), N3))
and the compositions are in equivalents all possible solid compounds lie on a
square plane (shaded in Fig. 6a) in which any point represents a combination
of 12 positive and 12 negative valencies. Therefore the square plane with
Si3N4, AlN, Al2O3, and SiO2 at its corners can be treated as a reciprocal salt
system with the double exchange reaction:
Si3 N4 2Al2 O3 , 4AlN 3SiO2 4
The corresponding phase relations can be represented in a quasi-ternary
system with the four compounds as components [181].
Many versions of phase diagrams of the subsystem Si3N4-4(AlN)-2(Al2O3)-3
(SiO2) have been published to date, based on experimental studies
(Table 3). In the early periods the diagrams were reported as ternary systems
Si3N4-AlN-Al2O3 [183] and Si3N4-SiO2-Al2O3 [184]. But as mentioned previ-
ously, the treatment as a quasi-ternary reciprocal salt system is by far more
appropriate. Based on more recent experimental and thermodynamic data in the
side systems Al2O3-AlN [185] and improved modelling [186], it became possible
to perform thermodynamic calculations which have secured the knowledge of
the quasi-ternary system. In Fig. 7 one example is selected from many others to
illustrate a calculated isothermal section at 1873 K and compare it with the
experimental data. Other important intermediate phases in the system are
Si2N2O (O), Si12Al18O39N8 (X) and six phases near the AlN corner [83]. These are
interpreted in terms of AlN polytypoids and are not indicated in Fig. 7.
In b-Si3N4 the Si and N ions can be replaced by Al and O ions to form an
extended solid solution according to the formula
Fig. 7. Calculated isothermal section at 1873 K of the subsystem Si3N4-4(AlN)-2(Al2O3)-

3(SiO2) [77], compared with experimental data (j). LS = oxide nitride liquid; O = Si2N2O;
X = Si12Al18O39N8; Mullite = Al2Si3O12; Spinel = Al3O3N
Si6 z Alz Oz N8 z 5

where z represents the number of replaced Al and O ions, respectively. The values
of z reaches a maximum between 4 and 5 at about 2000 K but decreases with
decreasing temperature. Full solubility occurs only with a constant cation: anion
ratio of 3:4. This phase is strictly semistoichiometric, and compositional
variations within this ratio do not require vacancies or interstitial atoms in the b-
Si3N4 structure. In Table 4 the dependence of the a and b lattice parameters of the
substitution of Si and N ions by Al and O ions is given.
Under high pressures cubic css (Sect. 2.1) with z values (Eq. 5) up to 2.8
have been observed [186a].
3.2.2
Low b-Si3N4 Solid Solutions
The mentioned SiAlONs (and SiBeONs) are notable exceptions to the extended
solubility of the b phase, whereas the solubility of all other alloying elements
listed in Table 3 is quite limited. The related systems are important for an
understanding of the effect of sintering additives.
Yttria is an often used additive to improve the sintering behaviour
of Si3N4. In the quaternary system Si-N-O-Y the quasi-ternary system
Si3N4-4(YN)-2(Y2O3)-3(SiO2) exists. In Fig. 8a and b an isothermal section
at 2000 K is shown as an example. Four oxide nitrides are existent: Y2Si3O3N4,
YSiO2N, Y4Si2O7N2, and Y10(SiO4)6N2, which are isotypic with the silicates
mellilite, wollastonite, woehlerite and the phosphate apatite, respectively.
Fig. 8a is the latest version of the diagram and differs slightly from earlier
ones. For the purpose of sintering, only compositions in the compatibility
Table 4. Dependence of lattice parameters in a- and b-solid solutions on composition
Phase a, nm c, nm Lit.
bss 0.7601 + 0.00304z 0.2906 + 0.002554z 87

Si6)zAlzN8)zOz 0.7603 + 0.002967z 0.2907 + 0.002554z 88
0.76069 + 0.00279z 0.29068 + 0.00263z 84
ass
MSi12)n)mAlm+NN16)nOn 0.7752 + 0.0036m + 0.002n 0.5620 + 0.0031m + 0.004n 127
M=Y 0.7752 + 0.0045m + 0.0009n 0.5620 + 0.0048m + 0.0009n 155
ass
MSi12)n)mAlm+nN16)nOn 0.7749 + 0.00673m + 0.00023n 0.5632 + 0.0055m + 0.00054n 118
M = Ca
triangles on either side of the Si3N4-Y2Si2O7 join are essential. In this region
there are no differences among the diagrams published.
For compositions on the Si3N4-Y2Si2O7 tie line, a liquid forms at the binary
eutectic temperature below 1550 °C and crystallises at 1500 °C. Compositions
with higher Y2O3 contents than found on the join will fall into the
compatibility triangle Si3N4-Y2Si2O7-Y10(SiO4)6N2, which has a lower eutectic
temperature than the triangle outlined by Si3N4-Si2N2O-Y2Si2O7. Compositions
with lower Y2O3 contents fall into the compatibility triangle outlined by
Si3N4-Y2Si2O7-Si2N2O.
The marked area of the Si3N4 corner in Fig. 8a indicates the composition
range of the useful Si3N4 ceramics and is enlarged in Fig. 8b. The narrow
arrangement of phase regions that fans out of the Si3N4 corner shows clearly
that small variations in composition, as occur readily in the fabrication and
processing of Si3N4 ceramics, may have very serious consequences for the
composition of phases and the ensuing properties. In this connection it
becomes obvious that the type of crucibles, powder beds and furnace
atmospheres chosen for sintering are crucial for the physico-chemical and
mechanical properties of the materials produced. These problems are
exacerbated by the kinetic retardation of the equilibria, which is especially
true of the intermediate oxide nitride Y10(SiO4)6N2.
The systems with low bss have been studied far less than Si-Al-O-N, except
those without Al2O3 as a second additive. However, most of the commercial
Si3N4 ceramics do contain Al2O3 as sintering aid, usually as a mixture with
others. But the resulting bss have only small values of z. It is considered that
Si3N4 ceramics with z-values ³0.5 belong to the group of extended bss and all
others with z-values <0.5 belong to the group of low bss [15]. This separation
appears arbitrary, but is justi®ed by the marked differences in the compo-
sition, corresponding properties, and by differences in the processing. To
produce Si3N4 ceramics with z ³ 0.5 usually AlN is added to the presintered
powder mixture. Therefore this distinction will be retained herein.
Usually the sintering aids are low melting mixtures of two or more
oxides. As a consequence the related systems become more complex. As
can be seen from Table 3, often mixtures of Al2O3 and another oxide are
common, as for instance Al2O3 with Y2O3. They form together the ®ve-
Fig. 8a, b. The subsystem Si3N4-4(YN)-2(Y2O3)-3(SiO2) [111]. a) Isothermal section at

1773 K. b) Si3N4-rich area of a
component system Si-N-O-Al-Y, which can be reduced to a quasi-

quaternary system according to the schematic Fig. 6b, the resulting JaÈnecke
prism. It consists of three quasi-ternary reciprocal salt systems on the
square plans Si3N4-4(AlN)-2(Al2O3)-3(SiO2), Si3N4-4(YN)-2(Y2O3)-3(SiO2),
4(AlN)-2(Al2O3)-2(Y2O3)-4(YN) and two ternary systems on the triangular
faces Si3N4-4(AlN)-4(YN), 2(Al2O3)-2(Y2O3)-3(SiO2). These relations make
the complex diagrams of the Si3N4 ceramics appear simpler and easier to
depict, but without necessarily making them more accessible. Detailed
information from the three-dimensional ®gure can be obtained from two-
dimensional plots of compatibility triangles. This is demonstrated with
the system Si3N4-4(AlN)-2(Al2O3)-3(SiO2)-2(Y2O3)-4(YN) in Fig. 9. In
this JaÈnecke prism for reasons of clarity only phase relations of
the Si3N4-4(AlN)-2(Al2O3)-3(SiO2) plane and the subsystem
Si3N4-YNá3AlN-4/3(AlNáAl2O3) (shaded) are represented.
Fig. 9. The JaÈnecke prism Si3N4-4(AlN)-4(YN)-2(Y2O3)-2(Al2O3)-3(SiO2) with the ass plane

(schematic)
3.3
a-Si3N4 Solid Solutions (ass)
In the subsystem Si3N4-YN á 3AlN-4/3(AlN á Al2O3) a stable a-Si3N4 solid

solution exists, representing the ass group of Si3N4 ceramics. In Fig. 10 an
enlarged section of the ass plane illustrates the homogeneity ranges for Nd, Y
and Yb.
The a-structure exhibits void positions with seven N atoms as nearest
neighbours (Sect. 2). These positions can be partially occupied by ions with an
ionic radius of about 0.1 nm, causing a stabilisation of the metastable
a-modi®cation (Fig. 11). The cations M which can be incorporated in the
a-structure are Li+, Mg2+, Ca2+, Y3+ and lanthanoides with Z ³ 60 [123, 169,
187]. Additionally Si4+ and N3) must be replaced by Al3+ and O2) to obtain
electroneutrality. The formula for the resulting a-solid solutions is
Fig. 10. ass for different rare earth (RE) ions [147]; only the shaded area has been studied.
Compare the shaded plan Si3N4-YN á 3AlN-4/3(AlN á Al2O3) in Fig. 9
Fig. 11. Projection of the ass crystal structure
Mx Si12 mn Almn On N16 n 6
x is the amount of the stabilising cation Mv+ with the charge v+, n the amount of
oxygen replacing nitrogen. To obtain electroneutrality the amount of Al3+
replacing Si4+ must be x á v+ + n in which x á v+ is normally expressed as m.
In addition to Fig. 10 in Fig. 12 some data are given regarding phase
relations and stability areas with different rare earth cations. The lowest
x-value for trivalent cations is 0.33; for Ca2+ the lowest value is 0.3 [123±127,
169, 187]. It should be emphasised that all ass do not include the pure a-Si3N4.
The atomic positions for different cations M are listed in Table 5, showing that
the Si- and N-positions do not differ signi®cantly from that in pure a. This is
Fig. 12. Range of m values in ass as function of the ionic radii [100]
Table 5. Crystal structure data of different a-solid solutions
70
Composition NdxSi12)n)mAlm+nN16)nOn YxSi12)n)mAlm+nN16)nOn YxSi12)n)mAlm+nN16)nOn
Parameter x = 0.33; n = 1.26; m = 1 x = 0.33; n = 1.26; m = 1 x = 0.5 ; n = 0.75; m = 1.5

Literature 152 152 170
Space group P31c (No. 159) P31c (No. 159) P31c (No. 159)
a, nm 0.78255 0.78167 0.782946
c, nm 0.57008 0.56948 0.570765
RE (Nd, Y) Occupation factor 0.17 0.170 0.274
(position 2b) z 0.695 0.691 0.233
(Si, Al)1 x 0.511 0.511 0.5090
(position 6c) y 0.428 0.427 0.0830
z 0.660 0.657 0.2338
(Si, Al)2 x 0.170 0.170 0.1717
(position 6c) y 0.917 0.916 0.2537
z 0.4498 0.448 )0.0038
(N, O)1 x 0.612 0.611 0.3468
(position 6c) y 0.956 0.952 )0.0490
z 0.443 0.446 )0.0092
(N, O)2 x 0.320 0.322 0.3227
(position 6c) y 0.00634 0.00691 0.3199
z 0.707 0.704 0.2460
(N, O)3 z 0.107 0.103 0.6462
(position 2b)
(N, O)4 z 0.452 0.437 0
G. Petzow á M. Herrmann
(position 2a)
caused by the high rigidity of the (Si,Al)3-(O,N)4 tetrahedron net-

work. Equation (6) demands the homogeneity range of ass on the one plane
in the relevant M-Si-N-O-Al systems and no defects in the (Si,Al)3(O,N)4
tetrahedron framework. This is in full agreement with the experimental results
[132, 187].
The maximum cation RE and oxygen solubility in a-solid solutions
depends on the ionic radii and increases with decreasing radius of the
lanthanoides (Fig. 10 and 12). Solubility limits of the Y-ass phase,
Ym/3Si12±(m+n)Al(m+n)OnN16±n have been determined for the 1.0 £ m £ 2.4
and n £ 1.7 by [155] and as 1.0 £ m £ 2.5 and 0.5 £ n £ 1.24 by [169].
These data differ slightly from more recent ones (Fig. 9b) [147]. Especially
the oxide solubilities in the a-solid solutions, e.g., the maximum n values
strongly depend on the temperature [147]. Recent investigations based on
Rietveld measurements show that, in the Si-N-O-Al-Y system, compositions
of ass up to an x value of 0.22 for low n values can be observed at a
sintering temperature of 1825 °C [152]. The differences in the experimental
results may be caused by dif®culties to reach equilibrium (absence of a
liquid phase and low diffusion coef®cients) and by the interactions with the
sintering atmosphere, leading to a shift of the composition. Weight losses
of 1±2 at% are quite typical (Sect. 5.3.3). Therefore, the a- and b-lattice
parameters calculated by the several formulae are different [127, 155, 169,
171] (Table 4). The reason for the deviations could be incompletely
equilibrated samples.
The stability areas of the a-solid solutions depend both on the size of the
rare earth ion and the temperature [147]. Increasing the size of the rare earth
and decreasing temperature leads to a reduced solubility of Al and O in the
ass, e.g., to lower maximum n values [188]. On the other hand, recent results
show a more extended ass with more than one stabilising cation, as can
expected with only one stabilising cation (different rare earths and Sr or Ca
[128]). Even elements like La or Ce, which alone do not form an ass, can be
effective as stabiliser together with Ca or Yb [131, 189]. Those multi-cation
a-solid solutions offer additional possibilities of variations in processing
and properties [154].
The temperature dependence of the stability of the ass could explain why
the ass ® bss transformation takes place at 1350±1700 K [123, 128, 138, 188]
in samples with compositions on the boundary of the ass region or in two-
phase ass/bss mixtures. The transformation is a solution-reprecipitation
process which often is kinetically hindered, and the presence of a liquid or low
viscosity glassy phase is necessary [188].
4
Si3N4 Powders
The majority of Si3N4 ceramics are made from powders and therefore depend
to a large extent on the quality of the starting powders. Si3N4 ceramics of
superior quality require well characterized and often extremely pure powders.
The powders determine the processing, the sintering behaviour and the
subsequent formation of the microstructure which strongly in¯uences many

properties of the densi®ed materials.
4.1
Powder Synthesis
Many different synthesis routes for obtaining Si3N4 powders are available:
± direct nitridation of silicon [21, 190±196],
± diimide synthesis [191, 197],
± carbothermal nitridation of silica [198±200], alumosilicates [201, 202], or
rice husk [203, 204],
± vapour phase synthesis [191, 205, 206],
± plasmachemical synthesis [21, 207±211],
± laser induced reactions [191, 212],
± pyrolyses of silicon organic compounds [213].
All of these synthesis routes result in mixtures of a and b, mainly with a
high a to b ratio and are based on four different chemical processes (Table 6).
Presently only the direct nitridation and the diimide synthesis are of
commercial importance; the industrial production by gas phase reaction was
given up recently.
The direct nitridation of Si is carried out at temperatures higher than
1100 °C but below the melting point of Si (1412 °C). This reaction is strongly
exothermic and must be precisely controlled to prevent the melting of Si and
the subsequent formation of a higher amount of coarse b-crystallites from the
melt [191, 216].
Commonly the powders are agglomerated and must be milled [191]. The
purity depends on the purity of the raw material and the milling procedure
after nitridation. Chemically puri®ed powders usually contain impurities of
chlorine and ¯uorine. Especially the low-cost refractory-grade Si3N4
powders produced by fast nitridation have a higher impurity content and
are coarser.
An improved nitridation process of Si in a rotary tube furnace allows a
better control of the exothermic reaction, avoids overheating and agglomer-
ation [192], is faster and results in smaller crystallites. Those powders have an
improved sintering behaviour and after sintering a more favourable micro-
structure compared to powders produced by a batch process [192]. Recently
successful fast nitridation in a ¯uidised bed reactor was realised in laboratory
experiments [194].
The self-propagating high temperature synthesis (SHS) uses the heat
formation due to the strong exothermic nitridation reaction for self propa-
gation [214, 217, 218]. Due to uncontrolled thermal peaks above the melting
point of Si high amounts of coarse b-grains are formed. This is the main
reason that the elegant SHS method has not succeeded to a technological
extent for production of high quality Si3N4 powders.
The direct nitridation of Si powder by a plasma chemical reaction takes
place at temperatures much higher than the melting point of Si up to 6000 K
[208]. The mean reaction time is very short, therefore only very small a- and
b-crystallites (about 20 nm) and amorphous Si3N4 are formed. By an
additional heat treatment ®ne a- or b-rich powders can be produced, which
allow the production of Si3N4 ceramics with mean grain sizes of about 100 nm
[219].
The diimide synthesis takes place at the interface between liquid ammonia
and the organic solvent in which the SiCl4 is dissolved. The product must be
washed and calcined to remove the NH4Cl. At high temperatures (1300±
1500 °C) the crystallisation to an a-rich powder takes place [220]. Temper-
atures above 1500 °C cause an increase of b and grain size [221]. A very ®ne
b-rich powder can be obtained, when the crystallisation of an amorphous
Si3N4 powder takes place at 1300±1450 °C in presence of an oxide nitride
liquid in which the amorphous phase can be dissolved and reprecipitated
mainly as b [219].
There are three more possibilities for the powder production by the diimide
synthesis [206]. Beside the reaction at interfaces of two liquids the formation
of the diimide can be realised by
± bubbling of gaseous SiCl4/N2 through liquid ammonia at about 0 °C,
± reaction of gaseous ammonia with liquid SiCl4 or SiCl4 solutions in organic
solvents at about 0 °C,
± reaction of gaseous NH3 and SiCl4 [196, 197, 206].
Very similar to the diimide synthesis process is the vapour phase synthesis.
Different gaseous components {CCl4 [21, 191], hexamethyldisilazane (HMDS)
or other silazanes [210, 222]} react at 800±1400 °C (Table 6) directly to form
amorphous Si3N4, which has to be crystallised by a further heat treatment. The
main starting materials in the gas phase process are SiCl4 and ammonia which
react to amorphous Si3N4 at 800 °C. Removal of the by-product NH4Cl,
crystallisation of the amorphous powder (1300±1500 °C), and deagglomera-
tion must follow.
Beside the production of pure Si3N4, the gas phase processes are used for
the preparation of Si3N4/SiC composite powders used for nanocomposite
materials [210, 222, 223]. For the production of such complex powders
different kinds of evaporable Si organic compounds are used, e.g., hexameth-
yldisilazane.
The reaction in the gas phase can also be achieved in a plasma or laser
induced. The result of this process is an amorphous or only partially
crystallised powder with a high speci®c surface area.
The gas phase processes and the diimide process offer a very ®ne, high
purity powder with high sinterability. The main disadvantage of these powders
are the poorer shaping behaviour and the higher price in comparison to direct
nitrided powders.
The production of Si3N4 powders by carbothermal reduction has not yet
been commercialised. The attempts to produce ®ne, cheap powders by this
method have not been successful [198]. For refractory-grade powders a
carbothermal reduction of different minerals to form bss and/or ass (SiAlONs)
may play an important role in the future [215].
74
Table 6. Synthesis of Si3N4 powders
Method Process Technology/raw materials Powder features Literature
1. Direct 1100±1400 °C Si-powder; N2 or N2/H2 a content: <97; variable [21, 191, 192,
nitridation 3Si + 2N2 ® Si3N4 or N2/NH3 gas purity (puri®cation); 195, 196]
solid + gaseous variable particle size
(milling); good shaping
behaviour; without
puri®cation cheep process
>1400 °C High temperature self Coarse powder with high [214]

3Si + 2N2 ® Si3N4 propagating synthesis b-Si3N4 content
Solid + gaseous (SHS)
<6000 °C Plasmachemical reaction Fine particle size, high purity [208]

3Si + 4NH3 ® Si3N4 + 6H2 powder; variable degree of
Solid + gaseous crystallinity, variable a/b
content
2. Diimide )40±0 °C Reaction at phase High purity powder, high [21, 197, 206]
synthesis SiCl4 + 6NH3 ® Si(NH)2 + 4NH4Cl boundaries liquid/liquid a content ®ne particle
900±1200 °C or liquid/gas or in the size, special treatment: ®ne
3Si(NH)2 ® Si3N4 + 2NH3 gas phase b powders; processing/
1300±1500 °C shaping more sensitive
Si3N4 (amorph) ® Si3N4 (cryst.) than for direct nitrided
powders
3. Vapour phase 300±1600 °C Chemical vapour reaction High purity, high a content, [21, 205, 213]

synthesis 3SiCl4 + 4NH3 ® Si3N4 + 12HCl (CVR) ®ne particle size, special
1300±1500 °C HMDS or other evaporable treatment: ®ne b-powders,
Si3N4 (amorph) ® Si3N4 silicon organic processing/shaping more
compounds sensitive than direct
nitrided powders
800±1400 °C
3(CH3)6Si2NH + 5NH3 ® 2Si3N4 + 18CH4
Plasma chemical synthesis Fine, high purity powder, [21, 209, 210,
high a content with 212]
Laser induced reaction variable degree of
crystallinity;
special treatment: ®ne b
powders, dif®culties
during processing/shaping
4. Carbothermal 1450±1600 °C SiO2 or kaolin, different Variable purity [198, 199, 202,
nitridation 3SiO2 + 6C + 2N2 ® Si3N4 + 6CO carbon qualities; (puri®cation), variable 203, 215]
rice husks a/b content; raw material
SiO2: a content:
<97; raw material kaolin:
high b content
75
4.2
Powder Characterisation
Si3N4 powders have pronounced differences with respect to crystallinity, a/b

ratio, surface area, surface charge, particle shape, particle size, particle size
distribution, agglomeration and impurities. This is obvious from Table 6
to some extent and even more from Table 7. As a consequence a precise
characterisation of the powders is imperative.
The chemical composition with respect to Si and metallic impurities
(mainly Fe, Ca, Al) is generally determined by wet chemical methods in
combination with standard spectroscopic techniques (AAS, AES, XRF)
(Table 8) [224±226]. A precondition is the dissolution of the powder. Typical
dissolving processes are fusion with sodium carbonate or mixtures of sodium
carbonate and boric acid, with alkaline hydroxides [225, 226] and special acid
treatments [225]. A more effective analysis based on optical emission
spectroscopy allows the direct analysis of impurities in the solid state and
requires no dissolution step [227].
Silicon is a constituent of the main phase as well as of the most common
impurities SiC, free Si, Si2N2O, SiO2, FeSix. Therefore, besides the determi-
nation of the total amount of Si, additional methods are needed to determine
the amounts of free Si and the other compounds. Quantitative X-ray
diffraction (XRD) is often used; with modern XRD methods less than
0.1 wt.% free Si [232] and 0.2 wt.% Si2N2O [228] are detectable. The accuracy
for SiC is somewhat lower due to the overlapping of the main peak of SiC with
b-Si3N4. Free Si can be determined also with volumetric methods [225].
Nitrogen can be determined by wet chemical analysis [225, 226] or by inert
gas fusion technique at 2700 °C [233, 234].
The total amount of oxygen can be analysed by neutron activation [224] or
inert gas fusion technique. The inert gas fusion technique with controlled
heating rate allows a distinction between surface and bulk oxygen [234±236].
Surface oxygen can be analysed by XPS methods [233, 237]. The oxygen is
mainly in a thin SiO2 surface layer on the Si3N4 particles. The total bulk oxygen
content consists of different amounts of Si3N4 solid solution and Si2N2O
[228, 234].
Table 7. Typical Si3N4 powder characteristics (From different commercial data sheets)
Direct nitridation Diimide Plasma

synthesis chemical
synthesis
Refractory Low cost Higher cost
grade grade grade
b content, wt% >25 5±20 5±10 <5 30

Mean particle size, lm >10 2±5 0.3±0.7 0.3 0.1±0.7
Crystallite size, nm >200 70±150 70±150 30±70 15±70
Surface area (BET), m2/g 1 4±6 10±15 2±13 20±70
Fe impurity, wt% £1 <0.1 0.03 <0.01 £0.01
Oxygen content, wt% <0.6±2 <1.5 <2 <2 <4
Table 8. Standard methods for the analysis of Si3N4 powders
Elements/phases Method Literature
a/b ratio XRD [224, 228±231]

Si Wet chemical analysis [225, 226]
Free Si XRD, [228, 232]
Wet chemical analysis [225, 226]
O, N, C Inert gas fusion technique, [233±235]
Neutron activation, [224]
Wet chemical analysis [225, 226]
Metallic impurities Wet chemical analysis, [224, 225, 227]
AAS, AES, MS, OES [224, 227]
The carbon content, mostly SiC or WC inclusions, is usually determined by

the inert gas fusion technique [204].
Anionic impurities as F) and Cl) result mainly from speci®c powder
preparation and puri®cation processes and can be determined by spectro-
scopic methods [224].
It has to be mentioned that the chemical analysis of the impurities often is
insuf®cient for an adequate powder characterisation because of their very
speci®c in¯uence on the properties of the powder compacts (sintered bodies).
On the one hand, particle size and particle distribution of the impurity
inclusions are important; for instance it has been shown that the same
impurity contents of Fe cause a drastic decrease in strength of the compacts if
concentrated in few inclusions with diameters >50 lm, but have nearly no
in¯uence as inclusions <5±10 lm [238]. On the other hand, most metallic
impurities form silicides which cause a remarkable volume increase [238].
The a-/b-Si3N4 ratio is measured by X-ray diffraction [228±231]. In earlier
investigations only some peaks where chosen [229±231]. New techniques
based on the Rietveld method use the whole angle range of the XRD diffraction
pattern [219, 228]. This results in more accurate data even in textured samples
[228]. The amount of amorphous Si3N4 also can be analysed by this method
using an internal standard [224].
The commercial Si3N4 powders mostly consist of 85±98% a; a high amount
of a is desired, since it results in materials with high mechanical properties
(Sects. 6 and 7).
The crystallite size can be determined by XRD [219, 224], which is sensitive
for crystallites up to 0.1±0.2 lm; bigger crystallites cannot be determined. This
is disadvantageous because the large crystallites have a signi®cant in¯uence on
the formation of the microstructure (Sect. 6). The TEM/SEM investigation is
very time-consuming due to the large amount of particles necessary to analyse.
Additionally, the distinction between a and b is dif®cult because of the similar
diffraction patterns.
The particle size distribution affects the shaping processes, green density,
the sinterability and the ®nal density. The particles mostly are not single
crystallites but contain several crystallites. Therefore, the size measured by
XRD is different from that measured by common particle size measurements
such as sedimentation or light scattering techniques. As an integral value of

the particle size often the speci®c surface area is used, measured by the BET
method [239].
For the particle size and particle shape analysis several methods are
employed [224, 240]. The quantitative particle shape analysis is time-
consuming and expensive and rarely used in Si3N4 powder characterisation.
Since in regular powders the particles are agglomerated, a main problem of the
measurement is the reproducible deagglomeration, which needs ultrasonic
treatment and a control of the interaction between particles [241]. The particle
interaction can be controlled by adjusting of the particle surface charge with
the help of organic surfactants [241, 242].
For the water-based processing of the powders with submicrometer grain
size, the surface charge characterised by the f potential is essential for a
homogeneous distribution of the sintering additives (Sect. 5). Therefore the
knowledge of the surface properties is very important [241, 242]. The surface
charge depends signi®cantly on the pH (Fig. 13), on surfactants and on the
surface oxygen, which forms an oxide layer. With increasing surface oxygen
the isoelectric point shifts to low pH, i.e., to the isoelectric point of SiO2
(Fig. 14). Therefore Si3N4 powders mostly will be dispersed or mixed with
sintering additives at high pH values of the dispersion.
5
Consolidation of Si3N4 Powders
The compaction of Si3N4 powders to parts or components is multi-staged and
in many respects similar to that used in powder metallurgy:
± Processing of powders by cleaning, milling, screening, deagglomeration,
mixing with processing and sintering additives, avoiding environmental
contaminations, drying.
Fig. 13. Dependence of surface charge (f-potential) of Si3N4, Y2O3, Al2O3 powders on pH
[241]
Fig. 14. Dependence of the isoelectric point [pH(iep)] of Si3N4 powders as function of
surface oxygen [241]
± Shaping by axial or isostatic pressing, extrusion, injection moulding, slip

casting, tape casting, colloidal methods.
± Densi®cation by different sintering methods.
± Finishing by cutting, grinding, lapping, polishing, drilling for precision
parts with optimised surfaces.
± Testing and applications of components (Sect. 10).
Within each stage fracture-causing defects may be introduced which cannot
be corrected in the following stages (Table 9) [21, 243]. In all stages there are
close correlations with chemistry, chemical engineering as well as solid state
and surface chemistry [244]. Probably most efforts in the development of Si3N4
ceramics are directed to powder processing and the other consolidation steps,
because during these procedures microstructures may be tailored to control
properties.
5.1
Processing
In general during powder processing two types of substances are added to the
Si3N4 powders:
Processing and sintering additives.
5.1.1
Processing Additives
These are dispersants (e.g., amines, fatty acids, polycarboxylic acids), binders
(e.g., wax polyacrylate, polyvinyl alcohol), plasticisers (e.g., stearic acids,
dibutyl phthalate, polyethylene glycol) and solvents (e.g., water, organics)
[244±250]. They are needed to prepare a suitable suspension or slurry for the
subsequent shaping step. They must evaporate during a presintering treatment
without causing failures (Table 9). Suspensions as well as slurries should have
an uniform distribution of the Si3N4 particles, adequate viscosity, and no
Table 9. Possible defects occurring in the stages of fabrication of Si3N4 ceramics [243]
Powder production Powder preparation Powder shaping Powder densi®cation

(e.g., reaction process) (e.g., Conditioning) (e.g., Pressing) (e.g., Sintering)
Unsuitable Unsuitable Bubbles, voids, Large and small pores

± Particle size agglomerate size cracks Micropore clusters
distribution distribution Micropore clusters Cracks
± Mean particle size Hard agglomerates Density Voids
± Particle shape Hollow agglomerates Inhomogeneities Non-uniform grain
Deviation from Pores in the Non-uniform growth
composition agglomerates distribution of Exaggerated grains
Foreign particles Varying density binder and Zones of varying grain
Impurities distribution in the sintering additions size distribution
Oversize particles agglomerates Segregation of Undesirable grain
Hard agglomerates Unsuitable viscosity binder and sintering boundary phases
Inhomogeneous additions and of Inclusions
distribution of small particles Surface roughness
the additives Incomplete removal
Unstable suspensions of the binder
Insuf®cient binder Organic inclusions
suspension Dust
Low content of Texturing resulting
binder phase from particle shape
Organic inclusions
bubbles. To avoid environmental impurities (dust, organic particles, etc.),

often clean room or closed loop conditions are required.
The dispersion stability, rheology, and consolidation of numerous aqueous
and non-aqueous Si3N4 suspensions have been studied extensively [251±257].
Recently a novel class of dispersants for Si3N4 powders in non-aqueous media
has been designed and its interactions with the powder surface have been
characterised systematically on the basis of surface chemistry and fundamen-
tals of colloidal stabilisation [255, 258].
5.1.2
Sintering Additives
They are a precondition for dense Si3N4 ceramics. Because of its covalent
bonding and low diffusivity Si3N4 cannot be densi®ed by common dry
sintering. To create a liquid-phase sintering process yielding ®nally to full
densi®cation, additions of sintering aids are necessary. They react with the
adhered SiO2 on the powder surface to a silicate phase, which is molten at the
sintering temperature.
Common sintering additives are mixtures of metal oxides (Li2O, CaO,
MgO, SrO, Al2O3, RE2O3, ZrO2) or mixtures of oxides with non-oxides like
AlN, ZrN, and Mg3N2. Often small deviations in composition cause pro-
nounced variations in effectiveness. For instance, it has been shown that the
compound YAlO3 has more bene®cial effects on sintering than Y2O3 + Al2O3
mixtures [256]. Mostly, amount and composition of the mixture of additives

have been developed empirically. As a ®rst approximation for a good additive
the following can be stated: Firstly, under sintering conditions Si3N4 should
not react with the additives to form a nitride and SiO2
Si3 N4 3 Mex O2 , 3 SiO2 3 xMeNy 2 1:5yxN2 7
The free energies of the exchange reactions Eq. (7) DG7° must be positive,
otherwise Si3N4 decomposes. Secondly, Si3N4 and SiO2 should not react with
the oxides or nitrides to form silicides or metals i.e., the reactions
Si3 N4 3 Mex O2 , 3 SiO2 3 xMe 2 N2 8
must have large positive free energies (DG8 ). These conditions are expressed
for several oxides in Fig. 15. Oxides which satisfy both conditions and
therefore are suitable as sintering additives lie in the shaded area of Fig. 15
[259].
For instance, TiO2 reduces to TiN (DG7 is negative and DG8 positive) and
the d-elements in the 5th to 8th group of the periodic table of elements react to
silicides, forming inclusions.
In addition, additives must form eutectics with Si3N4 below the sintering
temperature. Densi®cation improves with lower eutectic temperature and
viscosity of the eutectic melt. Therefore the densi®cation decreases in the
direction MgO/Al2O3 > MgO/RE2O3; MgO; RE2O3/Al2O3 RE2O3 (RE = Y,
Sc, La and lanthanoides).
Amount and composition of the additives are not only of decisive in¯uence
on the sintering parameters (temperature, pressure, time, atmosphere), but
Fig. 15. Plot of DG08 [Eq. (8)] against DG07 [Eq. (7)] for some metal oxides at 2000 K. Shaded
area indicates stable oxides under sintering conditions, i.e., effective additives (after [259])
also on the resulting phase relations (Sect. 3) and microstructures (Sect. 6),
which emphatically determine many properties of Si3N4 ceramics.
The additives are conventionally mixed with the Si3N4 powders either by
ball milling, planetary milling, or attritor milling. It is problematic to achieve
full homogeneity in the powder mixtures by these mechanical methods
[260, 261]. A clearly improved homogeneity can be reached by incorporating
additives into Si3N4 powder by using sol-gel techniques [260, 262, 263] or by in
situ incorporation by a combustion process [261]. Experiments with combu-
sted Si3N4 powders and nanosized sintering aids result in a reduction of
sintering temperature compared with conventionally processed Si3N4 ceramics
[264]. Nevertheless, it should be mentioned that the advantage of perfect
homogeneity cannot be fully exploited in the following production steps
because of diminishing in¯uences. So far the highest strength values of Si3N4
ceramics (up to 1600 MPa) are produced by conventional mixing of additives
and Si3N4 [245, 246].
5.2
Shaping
All shaping methods (also called ``molding'') mentioned are well-known in the
ceramic industry [244, 247]. Selection criteria depend on economical and
technical requirements. Simple shaped parts mainly produced by uniaxial
pressing, cold isostatic pressing, casting and especially injection moulding are
favoured for components with complex geometry [263].
All shaping methods need processing additives (Sect. 5.1.1): dispersants and
plasticisers for suited rheological conditions for slip casting and injection
moulding as well as binders for ensuring stability of the powder compacts.
Prior to sintering all processing additives have to be burned out quantitatively
without inducing defects (Table 9); this procedure is called ``dewaxing''.
Burnout or dewaxing temperatures are comparatively low compared to
sintering temperatures to maintain an open porosity; therefore the required
times reach from several hours to several days, depending on component
thickness and kind of plasticiser. Air is the preferred atmosphere for burning
out the binder. To prevent oxidation of the powder the temperature must
remain below 750 °C. Following the burnout treatment the green body may be
shaped by green or white machining if a precise shape is required before the
®nal densi®cation process.
5.3
Densification
Table 10 summarises all methods for the densi®cation of Si3N4 used at present.
The resulting Si3N4 ceramics classi®ed according to the densi®cation routes
are also listed together with several remarks on manufacturing characteristics,
properties and applications. For comparison with the sintered qualities,
information on reaction bonded silicon nitride ceramics are also included; but
will be treated in more detail in Sect. 8.
5.3.1
Densification Methods
The most common densi®cation method is the gas-pressure sintering which

guarantees better reproducibility and improved properties by only moderate
increase in production costs compared to pressureless sintering. The ®rst
dense Si3N4 ceramics were produced by hot pressing [13]. Nowadays it is used
for the preparation of specimens in materials development or for some
applications where components of simple geometries and low quantities are
required.
The hot isostatic pressure (HIP) methods are high-cost technologies and
therefore used only for special applications. A full densi®cation of Si3N4 in
absence of sintering additives succeeds only with capsule HIP; the adherent
silica of about 3 vol% on the Si3N4 particles acts as sintering aid since it melts
during HIPing. The HIP procedure requires encapsulating the Si3N4 powder
compacts in gas-tight glasses, which soften at the sintering temperature and
therefore uniformly transmit the external gas pressure to the powder compact.
High densi®cation with very precise retention of shape is a great advantage of
capsule HIP, but removing the encapsulation material by mechanical methods
or chemical etching is time consuming and costly.
5.3.2
Densification Mechanism
Apart from HIP, all attempts to densify Si3N4 powder without additives have
not been successful. Densi®cation requires sintering additives (Sect. 5.1.2).
The role of the additives is to react with Si3N4 and its adhered silica to produce
a liquid at high temperatures which allows mass transport through solution-
reprecipitation to consolidate the solid Si3N4 by rearrangement and coales-
cence in equilibrium with the liquid. Thus the general reaction may be
expressed as:
a a=b-Si3 N4 SiO2 additives starting powder mixture

# sintering temperature
b a=b-Si3 N4 liquid of SiO2 ; additives and dissolved Si3 N4
# cooling 9
c b-Si3 N4 ss amorphous phase SiO2 ; additives
# devitrification temperature
d b-Si3 N4 ss secondary phases amorphous phase
Both modi®cations a and b are present in different amounts in the starting

powder (Table 10); mostly a-rich powders are used. At sintering temperature
the metastable a-phase and the subcritical b-particles (dcrit.<0.5 lm) dissolve
in the liquid and reprecipitate during cooling as b solid solutions on the initial
stable b-particles which act as nuclei, while the homogeneous and heteroge-
neous nucleation can be neglected in most cases (Sect. 6.1).
84
Table 10. Production technologies and resulting Si3N4 ceramics
Sintered materials Reaction bonded materials
Densi®cation Low gas pressure Gas pressure Hot pressing Hot isostatic pressing (HIP) Reaction Reaction bonding
method sintering sintering bonding and postsintering
Presintering Capsule
Material SSN GPSN HPSN HIP-SSN, HIP-SN RBSN SRBSN,

Sinter-HIP-SN HIP-SRBSN
Starting powders Si3N4 Si3N4 Si3N4 Si3N4 Si3N4 Si Si + additives
Additive content, 7±20 3±15 2±15 2±15 0±8 0 3±15
vol%
Heat treatment 1600±1800 °C 1750±2000 °C 1500±1800 °C Presintering to 1750±2000 °C 1200±1450 °C Nitridation as for
nitrogen nitrogen uniaxial closed porosity gas pressure up to 100 h RBSN, Sintering
pressure pressure pressure in a (similar to SSN); up to 200 MPa as for SSN,
up to 0.3 MPa up to 10 MPa graphite die HIP-process: GPSN
1750±2000 °C; or HIP-SSN
gas pressure up
to 200 MPa
Linear
shrinkage % 15±22 15±22 up to 50 15±22 15±22 0 Up to 8±14
Main advantages Low sintering Complex Good Complex shapes Complex shapes No shrinkage, low shrinkage,
cost, complex shapes, better densi®cation, possible, possible, low cost raw low cost raw
shapes possible, reproducibility, high reliability sintering of materials material, low material
continuous better reliability materials with without production
sintering than SSN, lower low sinterability additives, high costs
process possible additive content (low additive reliability
content), high
reliability
Main Lower strength Somewhat Only simple High sintering High sintering Low strength, Time consuming

disadvantages than GPSN, higher shapes, low costs costs, high porosity, low nitridation
good sintering cost productivity costs for hardness
sinterability than SSN encapsulation
necessary
Relative 95±99 98±100 100 100 100 70±88 95±100
density %
Strength, MPa 500±900 800±1500 800±1500 800±1500 500 no 150±350 500±1000
(20 °C) additives,
800±1500 with
additives
Toughness, 5±8 5±11 5±8 5±11 3 no additives, 2±4 5±11
MPa m1/2 5±8 with
additives
Applications Wear parts, Cutting tools, Mainly used Ball bearings, Materials with Refractories Same as
precision wear parts, for evaluation wear parts, no or very SSN/GPSN
parts ball bearings, of materials, precision parts low additive
seals, engine prototypes content
parts (valves, with simple
turbo charger geometries,
rotors) cutting tools
85
During densi®cation the green compacts shrink up to 25% depending on green

density, starting powder characteristics and amount of additives. Fig. 16 shows
typical densi®cation curves of Si3N4 ceramics with different amounts of the
same additive mixture. Densi®cation starts at 1200±1300 °C when the eutectic
melt begins to form, as can be seen in the shrinkage curve dL/L0 (Fig. 16a) as
well as in the densi®cation rate curve dL/dt (Fig. 16b). The ®rst maximum in
Fig 16b is caused by the rearrangement of the Si3N4 particles and the second
by the solution-reprecipitation mechanism, the main densi®cation process.
The sintering behaviour improves with increasing additive amount (Fig. 16).
However for amounts above 15±20 vol%, gas bubbling occurs caused by an
increased formation of gaseous SiO [265, 266].
After cooling, the stable b-phase is embedded in an amorphous phase, from
which an oxide nitride can be precipitated during a devitri®cation heat
treatment, which is done just below the eutectic temperature. But even after
a)
b)
Fig. 16. Typical shrinkage dL/dLo (a) and densi®cation rate dL/dt (b) curves of Si3N4
ceramics. Additives: Y2O3/Al2O3 mixture of weight ratio 2:1 and contents of 3.5, 6.0 and
8.5 vol%
very long annealing times a certain amount of amorphous phase remains

(Sect. 6). Depending on the composition of the starting powder mixtures the
steps c) and d) in Eq. (9) yield different phase relations according to the
relevant phase diagrams (Sect. 3).
Theoretically, single-phase bss ceramics can only be achieved with
additives in the systems Si-N-O-Al and Si-N-O-Be in which extended b solid
solutions exist (Sect. 3.2). With respect to Eq. (9) in step c) or at the latest in
d) homogeneous bss material without other phases emerges. Nevertheless,
small deviations of the compositions from the theoretical one, which is
normally the case, result in the formation of small amounts of residual grain
boundaries. Also, in the theoretically monophase materials thin amorphous
grain boundary ®lms are stable, and indeed, in HIPed (2273 K, Ar pressure
180 MPa) high purity Si-N-O-Al powders glass ®lms at the grain boundaries
are obtained (Sect. 6.1.4).
In compositions of the starting powder mixture relevant to Si3N4 ceramics
of low b solid solution (Sect. 3.2.2), a polyphase material according to Eq. (9d)
is originated. For compositions resulting in stabilised ass ceramics (Sect. 3.3),
theoretically a single-phase material can also be produced by tailoring the
composition. But usually in these materials thin grain boundary ®lms also
remain. The reaction steps b)±d) in Eq. (9) can be applied in the analogous
way for these materials. An a solid solution is existent because the stabilising
elements are dissolved in a mostly during heating between step a) and b) in
Eq. (9).
Finally, one should take note of the fact that in all Si3N4 ceramics a
residual, continuous glassy phase remains between the crystalline phases,
even after prolonged annealing time at temperatures above 1300 K. This
amorphous phase is neglected in all phase diagrams (Sect. 3), yet strongly
affects the properties. Therefore much effort has been made to characte-
rise the amorphous phase which will be treated in more detail in
Sect. 6.1.4.3.
In conclusion, it can be stated that Si3N4 ceramics are polyphased materials
including mainly bss, ass, secondary phases (mainly oxide nitrides, in ceramic
literature generally called oxynitrides) and an amorphous phase, all having
characteristic morphologies and can be arranged in a manifold of micro-
structures (Sect. 6).
5.3.3
Influence of Densification Parameters
Besides the in¯uence of the starting powders and the sintering additives,
temperature, time, pressure, and atmosphere are important parameters which
must be taken into account.
Of course, increasing sintering temperatures and time enhance densi®ca-
tion, but above optimum values they cause unfavourable changes of the
microstructure (Sect. 6). In general, increasing pressure during hot pressing
and HIPing causes increased densi®cation rate, re®nement and for hot
pressing, an anisotropic orientation of the microstructure as well.
5.3.3.1
Decomposition
Because of the decomposition of Si3N4, the sintering temperature is limited,

and the vaporisation of the melt formed by reaction of the additives with the
adherent silica begins. The decomposition can cause a shift of composition in
the near surface area of the sample during sintering [232, 267±269, 272±274].
To suppress decomposition, high nitrogen pressure in the sintering atmo-
sphere is required. This is achieved in the powder bed technique and
encapsulation technique. The formation of SiO according to
3 SiO2 l Si3 N4 , 6 SiO 2 N2 10

is the main reason of decomposition during sintering [232, 267, 270±272].
Overheating or a ¯owing nitrogen atmosphere may cause additional
decomposition by evaporation [274]:
Si3 N4 , 3 Si 2 N2 11
In a non-¯owing atmosphere the evaporation becomes more pronounced

when the Si3N4 nitrogen partial pressure reaches the equilibrium partial
pressure of Eq. (11). The reactions (10) and (11) explain the dependence of the
weight losses on the ratio of the mass or volume of the Si3N4 samples to the gas
volume in the furnace. The colour changes of the samples during sintering are
also caused by this ratio [232] (Sect. 7.4).
Amount and composition of the sintering additives may also be changed by
interaction with the atmosphere. For instance, Al2O3 can be partially reduced
to AlN which will be dissolved in the b-Si3N4 lattice (Sect. 3) according
Eq. (12):
0:5 z Al2 O3 2 0:25 zSi3 N4 , Si6 z Alz N8 z Oz 0:25 z SiO2 12
In this case Al2O3 acts like a buffer, the more SiO2 evaporates, the more Al
dissolves in the silicon nitride grains. The amount of Al in the Si3N4 solid
solutions depends, besides the sintering parameters, on the Al2O3 activity in
the oxide nitride liquid, i.e., on the ratio of Al2O3 to the other additives [269].
Strong reducing atmospheres and high weight losses cause an intensive
reduction of Al2O3 and SiO2 which may even result in a-SiAlON formation on
the surface [232, 274].
Rare earth oxides and Y2O3 are more stable than the other additives [267];
no change in concentration of yttria or lanthanoide oxides has been found
experimentally, even during long sintering times [232, 268, 272, 274].
MgO, commonly used as sintering additive, is unstable and evaporates
according to:
6 MgO Si3 N4 , 6 Mgg 3 SiO2 2 N2 13
The alkaline and other alkaline earth oxides behave similarly.
5.3.3.2
Influence of Carbon
Carbon or graphite heating elements and insulating materials, as commonly

used in sintering devices, may have a deleterious in¯uence on the densi®cation
process. Carbon or SiC can cause a mass loss in the sintered compact during
sintering because of the reduction of SiO2 according to:
3 SiO2 l 6 C 2 N2 , Si3 N4 6 CO high N2 pressure 14
SiO2 l 2 SiC 2 N2 , Si3 N4 2 CO low N2 pressure 15

At higher nitrogen pressure, C is in equilibrium with Si3N4, at lower
nitrogen pressures, SiC. Thus at 1800 °C the conversion of SiC to C and Si3N4
takes place at a nitrogen pressure of 6 MPa. There is a pronounced
dependence of the stability ranges of the species on nitrogen pressures and
temperature [275]. The formation of CO is the reason for the mass loss, which
should increase with increasing N2 pressure, assuming equilibrium. But under
usual sintering conditions a reduced weight loss with increasing nitrogen
pressure is observed [232, 272, 274]. This indicates a non-equilibrium of the
CO partial pressure. The formation of CO seems to be kinetically determined
by transport processes involving SiO, which is reduced by increasing nitrogen
pressure [232, 272]. The weight loss can be minimised by reduction of the H2O
and O2 content in the atmosphere because both react with graphite heating
elements and form CO. An SiC surface layer on the carbon parts in the
sintering devices reduces CO formation for kinetic reasons, and that is why
aged carbon crucibles are advantageous as experience shows.
Another way to reduce the decomposition due to the interaction with the
atmosphere in a carbon containing furnace is sintering under CO partial
pressure [268]. If the CO partial pressure in the furnace reaches the
equilibrium pressure (Eqs. 14, 15), CO is no longer formed and consequently,
the weight loss is reduced. But if the pressure is higher than the CO
equilibrium pressure, additional SiO2 and C are formed (up to 0.6% C) [268].
The presence of carbon causes structural defects (pores, concentration
gradients, dark coloured spots), and therefore has to be suppressed or
minimised for a reproducible sintering process. Additionally, high CO partial
pressure results in dif®culties in temperature control, and the life of the
heating elements is reduced.
5.3.3.3
Influence of Powder Bed
Si3N4 ceramics have been, and sometimes still are, sintered in a powder bed to
prevent the decomposition and interaction with carbon and/or graphite
elements of the sintering devices [270]. The powder beds usually consist of
Si3N4/BN mixtures and occasionally sintering additives are added [276, 277].
The powder bed may react with carbon parts in the furnace [274]. A critical
evaluation of the relevant results must take into account the dependence of SiC
and C stability on the nitrogen pressure and compositional changes of the gas
phase.
Sintering in a powder bed has several disadvantages compared with normal
sintering (low thermal conductivity, temperature gradients, increased scatter
of sample shape and properties in large furnaces, and more complicated
handling during packing and removal of the powder). Therefore sintering
without a powder bed is increasingly favoured and has gained acceptance
because of improved sinterability of Si3N4 powders, better sintering furnaces
and by better understanding of the sintering processes. Control of interactions
between atmosphere and material is essential for reproducible sintering.
Nearly complete densi®cation and weight losses below 1% can be achieved
without powder bed.
5.3.4
Densified Materials
A classi®cation of dense Si3N4 ceramics follows from the densi®cation

techniques (Table 10):
Sintered Si3N4 SSN
Gas-pressure sintered Si3N4 GPSN
Hot pressed Si3N4 HPSN
Presintered hot isostatically pressed Si3N4 HIP-SSN (Sinter-HIP SN)
Encapsulated hot isostatically pressed Si3N4 HIP-SN
Silicon nitride ceramics are not merely only one material but several
classes of materials. All of them are multiphased, i.e., they exhibit a
heterogeneous microstructure which has formed during sintering (Sect. 6).
Therefore in all classes a large variety of properties is predominant and as a
consequence also a large variety of potential applications (Sect. 10). Often
little variations in the powders and the processing parameters cause
remarkable changes in the microstructure which have a pronounced effect
on properties (Sects. 6 and 7).
5.4
Finishing
In general, components of Si3N4 ceramics have to ful®l high performances.

Therefore, precise and careful ®nishing steps are indispensable. Usually
cutting, grinding, drilling, and polishing are used but they are time consuming
and expensive. Because of the great hardness of Si3N4 (Table 1) mainly
diamond tools are necessary. Ultrasonic erosion, laser cutting and spark
erosion are very promising and increasingly used tools for ®nishing
components of Si3N4 ceramics [278±280].
5.5
Precursor-Derived Si3N4 Ceramics
During the past years the pyrolysis of precursors, mainly Si- and N-containing
polymers, has gained interest as a production method and to some degree as
an alternative to the described powder route [213]. Both monophase Si3N4 and
Si3N4 composites with compositions in the systems Si-C-N and Si-B-C-N have
been prepared free of additives. After pyrolysis precursor-derived ceramics are
amorphous and require an additional heat treatment >1300 °C for crystalli-
sation [Structure and Bonding, Vol. 101 (2002) p 137]. So far, the precursor
line is restricted to small and thin components because a short diffusion path
is required for the release of the gaseous species developed during pyrolysis.
The gas causes uncontrolled cracks, pores and shrinkage with the consequence
of pronounced deterioration of the mechanical properties. To overcome these
disadvantages the `àctive ®ller concept'' has been developed [281]. It is based
on the idea of using additional ®llers (for instance, Ti, Cr, Mo) which react
with the gaseous products and reduce them. In addition, the volume change
caused by the solid reaction products (mainly nitrides, carbides and
carbonitrides) allows a better control of shrinkage.
6
Microstructures
In addition to the crystal structure, the microstructure has a signi®cant
in¯uence on the characteristic properties of a material. Type, amount,
arrangement, size, shape and orientation of the various phases all contribute to
the actual microstructure of a material which thus results from the
combination of all phases and the defects they contain. Such defects are
vacancies, dislocations, grain boundaries, pores and cracks. The microstruc-
ture is an important domain within the science of materials. The higher the
requirements of a material, the more stringent are the requirements on its
microstructure, i.e., the more accurately must its microstructure be estab-
lished. The aim is to create a microstructure speci®cally designed to produce a
given property pro®le. The terms ``microstructural engineering'' and ``micro-
structural design'' are the key words used to describe this problem.
The amount, distribution, size, morphology of the a- and b-particles,
secondary and amorphous phases of Si3N4 ceramics are decisive factors for
their quality and reliability. In the following, mainly sintered and gas pressure
sintered Si3N4 ceramics will be discussed because of their extraordinary
economical interest and because they show all the microstructural features
also present in the other Si3N4 ceramics.
6.1
Development of Microstructures in bss Ceramics
Silicon nitride ceramics are mostly produced from a-rich Si3N4 powders which
transform during sintering by a solution reprecipitation process to b solid
solutions (bss) according to the reaction sequence in Eq. (9). Under

conventional sintering conditions the low bss and, but to a lesser amount,
the extended bss, form elongated needle-like grains with ratios of grain length
to thickness (aspect ratios) comparable to whiskers. This is the basis for the
concept of an in-situ whisker reinforcement of Si3N4 ceramics [282].
The microstructural development is controlled mainly by the Si3N4 starting
powders, the additives and the sintering parameters.
6.1.1
Influence of Starting Powder
Model experiments with oxide nitride glasses have revealed that the growth of
b-grains starts from initial b-particles present during liquid phase sintering
[Eq. (9), step b] [283, 284]. Until now no homogeneous nucleation was
observed in the oxide nitride liquid [283, 285]. Only sometimes heterogeneous
nucleation of bss on a-particles was detected during sintering of a-rich starting
powders, obviously due to the similar crystal structures [286, 287]. Epitaxial
growth of each modi®cation on the other was found in bss and ass as well,
because of the lower differences of the lattice parameters and the reduced
nucleation energy compared to the pure a- and b-modi®cations [288]. In
general the starting powders contain b-particles (³2 vol%). During sintering,
grain size and morphology of the growing b-particles are strongly in¯uenced
by the number and size of the initial b-Si3N4 particles. A small number results
in a large interparticle distance between b-grains, and therefore the grains are
able to grow during liquid phase sintering [Eq. (9), step b] without signi®cant
impingement by other grains. In case of b-rich powders (>30%) with particle
sizes of about 1 lm, the driving force for the solution reprecipitation process
is much smaller. The resulting microstructure is more or less ®ne grained and
characterised by equiaxed grains [282, 289]. Fig. 17a and b show two
microstructures of different starting powders densi®ed under the same
conditions. Similar or even ®ner microstructures can be obtained from very
®ne b-powders (grain size £50±100 nm measured by XRD) by hot pressing
and further heat treatment (Fig. 17c and d) or gas pressure sintering [219,
290]. Obviously the growth rate of bss in pure b-powders does not differ so
much from that in a-rich powders in case of ®ne powders, but decreases
noticeably in coarser b-rich powders [291].
Nevertheless, the microstructure cannot be directly correlated to the initial
b-content in the starting powder. The experimentally determined particle
density in sintered samples indicates that only a part of the initial b particles
are able to grow [283]. The number of growing b-particles depends on a
critical particle diameter dcrit (Fig. 18). Particles below dcrit will dissolve in the
oxide nitride liquid during phase transformation and reprecipitate on the
overcritical b-particles according to an anisotropic Ostwald ripening process
[284, 292, 293].
Besides the sintering parameters (Sect. 6.1.3), dcrit is determined mainly
by three parameters [290, 294, 295]. First, the size distribution of a- and
b-particles due to a higher stability of large a-particles compared to small
Fig. 17a±d. Microstructures of bss ceramics (plasma etched). a from a-rich starting powder.
b, c, d from b-rich starting powders; a and b pressureless sintered; c hot pressed, and d
sample c additionally heat treated
Fig. 18. Schematic of b particle size distribution of a Si3N4 starting powder
b-particles in the early stage of sintering. A high volume fraction of large

a-particles cause an increase in dcrit. Second, the nitride solubility in the
siliceous melt which is determined by the additive composition. A higher
nitride solubility shifts dcrit to higher values. Third, the volume fraction of b. If
the powder does not contain enough a, a higher amount of smaller b-particles
will dissolve to saturate the siliceous melt, and dcrit increases (e.g. in cases of
b-rich starting powders).
Furthermore, detailed experiments showed that the microstructural evolu-

tion is controlled by grain impingement. Theoretical consideration of the Si3N4
crystal structure (Sect. 2) revealed a higher stability of the prism planes of the
hexagonal shaped grains compared to basal planes. However, the growth rate
of basal planes is much higher than that of prism planes, resulting in needle
like grain morphology. The difference in growth rates can be related to the
Si3N4 crystal structure and is attributed to an energetically more favourable
attachment of a surface nucleus on a basal plane. The (001) planes are
`àtomically rough'', and the growth mechanism is diffusion controlled and
much faster [283]. The surface energy of the basal plane, i.e., the stability in the
melt depends on the thickness of the grains [15, 285, 296]. The (100) plane is
`àtomically ¯at'' and the growth velocity is determined by the formation of
surface nuclei [297, 298]. The surface energy of the prism plane increases with
decreasing aspect ratio and thickness. Consequently thick elongated grains
grow anisotropically during the densi®cation process and may result in
exaggerated grain growth at temperatures above 1800±1850 °C [296, 299]. This
fact is used in the production of seeded materials (Sect. 9.3).
Based on the assumption of a high stability of prism planes and a higher
growth rate of the basal planes, three microstructure controlling cases of grain
impingement can be derived (Fig. 19). The growth of a basal plane will be
stopped, if it hits a more stable prism plane (A). The prism plan can cause a
diameter reduction of a growing basal plane (B). If the diameter of a basal
plane is large enough it will grow around smaller grains and dissolve them (C).
It is important to point out that cases (A) and (B) will stabilise the
microstructure against grain growth. However, at higher temperatures and
long sintering times exaggerated grain growth of particles with a large initial
diameter occur by the mechanism schematically shown in Fig. 19C.
Fig. 19. Schematic of the steric hindrance of grain growth

6.1.2
Influence of Sintering Additives
The in¯uence of the densi®cation aids is pronounced, as can be seen for

example from Fig. 20 [300]. The results presented are based on studies by
using an additive combination of Al2O3 and RE2O3 (RE = La, Nd, Gd, and Yb).
The additive compositions were calculated in consideration of the different
nitride solubility to keep the volume fraction (Vliq = 0.15) of liquid-forming
additives constant. The plots in Fig. 20 indicate a ®ner-grained microstructure
with high aspect ratios for the La-containing sample compared to the material
densi®ed with Yb2O3.
The ratio of additive combinations also in¯uences the microstructural
evolution. For instance, with decreasing ratio of Y2O3/Al2O3 the microstruc-
ture becomes ®ner and the aspect ratio lower [301, 302]. MgO as well as CaO
additives accelerate the grain growth and increase the aspect ratio [303, 304].
6.1.3
Influence of Sintering Conditions
The heating rate can in¯uence the microstructural development. Fast heating
leads to a coarser microstructure due to the dissolution of ®ne b-Si3N4
particles during the heating period in the oxide nitride melt [290, 305]. On the
other hand, low heating rates in the interval 1500±1700 °C intensify an
exaggerated grain growth [306]. This seems to be caused by the reduced steric
hindrance of growing particles and growth processes involving gas phase
Fig. 20. Grain diameter and aspect ratio distribution for Si3N4 ceramics densi®ed with
various sintering additives
transport processes. Similar exaggerated grain growth was observed in a two-

step sintering procedure, lowering the temperature of the ®rst step below
1750 °C [307].
Sintering temperature and time are other important parameters, not only
for densi®cation (Sect. 5.3.4) but also for the microstructural development
[308]. After complete phase transformation and partial devitri®cation accord-
ing to Eq. (9) step c, grain growth starts by dissolution of smaller b-grains, as
concluded from Fig. 21. An increasing aspect ratio with increasing sintering
temperatures up to 1950 °C and a decrease with further temperature increase
is observed [309]. The observed development of a needle- or rod-like
microstructure is caused by the particle size distribution at the beginning of
the grain growth stage. Broad particle size distributions lead to a faster grain
growth than very narrow ones. Some grains with a large initial diameter can
grow in the length direction with a minor steric hindrance up to a length of
100±200 lm [310, 311]. Fig. 21 shows clearly the different types of steric
hindrance explained schematically in Fig. 19. Big elongated grains can also
form by coalescence during the ®nal sintering [286].
6.1.4
Microstructural Features
From the micrographs shown in Figs. 17 and 21 it is evident the bss grains
are embedded in an amorphous or partially crystallised matrix; their amount
Fig. 21. Microstructure of gas pressure sintered Si3N4 after 35 min at 1835 °C (a) and
360 min at 1900 °C (b), respectively
and chemistry are determined by the sintering aids. According to the

corresponding phase diagrams the amorphous phase is not in an equilibrium
condition. In order to permit a controlled crystallisation, postsintering heat
treatments are necessary for devitri®cation. Nevertheless, the devitri®cation
of the glassy phase is not completed by this treatment. Detailed high-
resolution transmission electron microscopy (HRTM) studies reveal that heat
treating only partially crystallises the residual amorphous pockets, as shown
in Fig. 22a. Complete crystallisation of these pockets cannot be attained
because residual glass remains at the tip of each triple junction. Furthermore,
homophase (between two Si3N4 grains) as well as heterophase (between Si3N4
and a crystalline secondary phase) two-grain junctions always remain
amorphous. Generally it can be stated that the microstructure of Si3N4
ceramics is characterised by three features: the Si3N4ss grains, the secondary
crystalline phases and the amorphous grain boundary ®lms, as shown
schematically in Fig. 22b.
Fig. 22a, b. Microstructural features in Si3N4 ceramics. a HRTM of a pressureless sintered

Si3N4 ceramic with 10 vol% Yb2O3. Grain boundary ®lm between the Si3N4 grains (SN) as
well as between the secondary phase and SN grains. b schematic of grain boundary region
after post-sintering/devitri®cation treatment. Complete devitri®cation cannot be achieved
6.1.4.1
bss Grains
As mentioned before, the morphology is given by the number and size of the
initial b particles, sintering temperatures and time. An important goal is to
have a uniform or bimodal grain size distribution of grains with high aspect
ratio in a ®ne-grained matrix of elongated grains [295, 299, 312±315].
Due to the low diffusion coef®cient in the bss grains a concentration
gradient of Al arises, which clearly outline the nuclei of the large bss grains
(Fig. 23) [283]. In large grains different subgrain boundaries and inclusions of
the grain boundary phase have been observed by TEM [286].
6.1.4.2
Secondary Phase
The composition of the crystalline secondary phases can be predicted by the

phase diagram (Sect. 3). According to the phase diagrams the resulting Si3N4
ceramics would either contain two phases, i.e., compositions reaching
equilibrium on a tie line (Si3N4 + sec. phase), or three phases, i.e., compo-
sition reaching equilibrium in one of the compatibility triangles containing
Si3N4 as an end-member (Si3N4 + sec. phase I + sec. phase II). When further
components are included, the phase equilibrium must be viewed in compo-
sitional space, and therefore, dense Si3N4 ceramics may contain more than
three equilibrium phases. Thus number and type of equilibrium secondary
phases depend on the amount of each of the starting constituents (Si3N4, SiO2,
additives, impurities) and their phase relations. Generally the secondary
phases are oxide nitrides; a few examples are given in Table 11. From
analytical electron microscopy investigations it has been shown that the
secondary phases are highly enriched with rare earth and Al2O3 additives at
the triple grain junctions while they are less concentrated at the two grain
junctions (Fig. 22b). It is also found that the secondary phases are not
uniformly distributed within the triple junctions where they are more
Fig. 23. bss needle grown on b particle as nucleus

Table 11. Grain boundary phases in Si3N4 ceramics [101]

Compound Structure type La Ce Pr Nd Pm Sm Eu Gd Tb Y Dy Ho Er Tm Yb
RE4Al2O9 RE4Al2O9 X X X X X X X X X X
REAlO3 Perovskite X X X X X X X
RE3Al5O12 Garnet X X X X X X X X
REAl11O18 PbFe12O19 X X X X X X X
REAl12O18N PbFe12O19 X X X X X X X
REAl11+xO18Nx (0 £ x £ 1) PbFe12O19 X X X X X X X
RE2AlO3N K2NiF4 X X X X X X X
RE2SiO5 Orthosilicate X X X X X X X X X X X X X X X
RE2Si2O7 Pyrosilicate X X X X X X X X X X X X X X X
RE9,33(SiO4)6O2 Apatite X X X X X X X X
RE10(SiO4)6N2 Apatite X X X X X X X ? ? X
RE2Si3O3N4 Melilite X X X X X X X X X X X X X X X
RE2Si3)xAlxO3+xN4)x Melilite X X X X X X X X X X
RESiO2N Wollastonite X X X X X X ? ? X
RE4Si2O7N2 WoÈhlerite X X X X X X X X X X X X X X X
RE4Si2)xAlxO7+xN2)x (0 £ x £ 2) WoÈhlerite ? ? X X X X X X
RE3AL3+xSi3)xO12+xN2)x U-Phase (RE3Ga5GeO14) X X X X X X X X X X X X
RE4Si9Al5O30 N W-Phase (Latiumite) X X X X
X: stable phase.
Empty cells: phase not stable.
?: no data.
99
concentrated at the central regions, as compared to the surrounding regions at

the triple pockets [316].
6.1.4.3
Amorphous Phase
The amorphous phase remains as a stable ®lm at the grain boundaries

[317±321]. Its thickness depends on the temperature [312, 313], the additive
composition [314, 322, 323] and on the impurities of the starting powder [324].
It varies between 0.5 and 1.5 nm [325] and is constant for a certain additive
composition and independent of the amount of additives [317, 318]. An
increasing additive content causes only an enlargement of the three- and four-
grain junctions, but no change in ®lm thickness of the two grain junctions.
Small variations in the chemistry of the amorphous ®lm may result from the
bss grain growth [316, 326].
The experimental observations con®rm theoretical considerations which
predict a stable grain boundary ®lm in Si3N4 ceramics through a balance of the
attractive van der Waals forces and the various repulsive interactions across
the grain boundary [318]. In a more sophisticated treatment of this problem
the diffuse interface approach has been used to describe a ¯at interface
between the two coexisting phases [327]. The main assumption of this
approach is that the free energy vs. concentration between the two phases can
be represented by a continuous function of the Gibb's energy, as if a miscibility
gap existed in the system considered. In other words, an interaction energy
term exists, describing equilibrium conditions for the amorphous ®lm and its
®nite thickness. Recently, molecular dynamic calculations have proved that
atomic Si-N and Si-O bonds contribute to this energy [320]. However, the
apparent stability of the glassy phase that would be expected to crystallise
during subsequent heat treatment (devitri®ation) demands an additional
explanation. It has been suggested that the crystallisation of the intergranular
®lms can be inhibited due to stresses caused by volume changes during
crystallisation [319]. On the other hand, a model experiment has proved the
stability of the amorphous intergranular ®lms, even in a geometry free of
residual stresses and capillary effects [328]. Thus the question arises whether
the existence of amorphous phases in the grain boundary can be explained by
the thermodynamic equilibrium in the ceramic system considered; in other
words, whether or not the amorphous phase must be taken into account in the
phase relations and therefore must be included in phase diagrams. Further
arguments for this possibility are given from a molecular dynamic simulation
in silicon as a model for covalent materials, identifying a thermodynamic
criterion for the existence of thermochemical stable disordered intergranular
®lms based on the relative energies of the atoms in the grain boundaries and in
the bulk amorphous phase [329]. A subsequent step is a thermodynamic
treatment of boundary phase admitting computations of realistic phase
structure in liquid phase sintered ceramics [330]. Thus the additional
interaction responsible for getting the thin intergranular ®lm have been
considered as a constraint for the thermodynamic equilibrium of the system.
The method of parallel tangent construction developed for representation of

equilibrium between bulk and liquid phase at sintering temperature can be
used for the description of the stability of the grain boundary ®lms. To
examine the true equilibrium of the system the Gibb's energy G* (Dl)
computed assuming the amorphous phase as the boundary interface, has to be
compared with the Gibb's energy G0 (Dl) calculated in the same way, but
taking for the grain boundary interface the crystalline equilibrium phase
conforming to the conventional phase diagram. A necessary condition for the
amorphous grain boundary phase in equilibrium with the matrix is a positive
value of the thermodynamic driving force DG(Dl):
DGDl G0 Dl G? Dl > 0 16
This approach has been applied for Si-Al-O-N ceramics. The results can
explain the existence of the amorphous phase in the grain boundaries
(Fig. 24), depending on the overall composition of the system ([equ% O] = 16,
12, 8 and [equ% Al] = 10.6) and on the strength of interaction (Dl). In a strict
sense it must be concluded that the conventional phase diagrams of Si3N4
ceramics are incomplete representations of materials containing an amor-
phous intergranular phase.
6.2
Development of Microstructures in ass Ceramics
Compared to bss, the investigations on ass microstructures are less numerous

and intensive, but many analogies exist. Similar to bss ceramics the ass
ceramics (Sect. 3.3) are produced by liquid-phase sintering involving the steps
explained in Eq. (9). The starting powder consists of a-rich Si3N4 (a > 80%),
the additives (mainly AlN, Al2O3) and a compound with an appropriate cation
which is solvable in the a structure (e.g. Li+, Mg2+, Ca2+, Y3+, Nd3+-Lu3+). This
reaction starts at temperatures above 1450 °C [331]. The amount of liquid
Fig. 24. Driving force for an amorphous grain boundary phase in equilibrium with bss as a
function of composition in the system Si-N-O-Al and strength of interaction [330]
available for densi®cation is quickly reduced due to the formation of ass solid
solutions. Pure ass ceramics with low amounts of stabilising cations are
therefore dif®cult to densify completely; improved densi®cation can be
realised at higher amounts [Eq. (6); high n and m values] [332, 333] or in two-
phase ceramics ass/bss [334]. Whilst compositions with low amounts of
stabilising cations form microstructures with ®ne equiaxed grains in the
densi®ed ceramics, the compositions with relatively high n and m values [127]
or mixed cation compounds (rare earths with Sr or Ca) lead to microstructures
with elongated ass grains. Elongated grain growth can be obtained by using b-
Si3N4 as starting powder [333] as well as a-Si3N4 [121, 332, 335, 336]. The size
of the elongated grains is larger than those in bss ceramics, indicating the
importance of grain growth. However, the reason for the anisotropic grain
growth of ass has not been as intensively studied as in bss but the mechanism
seems to be quite similar [337]. Starting with a-Si3N4 powders the ass grains
form epitaxially on the existing a particles. Also, epitaxial growth of ass on bss
has been observed [288]; nucleation in the oxide nitride liquid does not seem
necessary [340].
A precondition for growing of elongated ass grains is enough liquid at
sintering temperatures. Besides the mentioned larger amounts of the cation
stabiliser this can be achieved by sintering at higher temperatures (³1900 °C),
reducing the formation rate of the ass by using b powder [337, 338], special
heating rates [337], seeds [338, 339] or by an additional stable liquid [340,
341]. Recently it has been shown that elongated ass ceramics with low n and m
values can be produced by gas pressure sintering, using an appropriate liquid
phase [340]. The growth is also in¯uenced by the nature of the stabilising
cations [337].
b-Si3N4 as starting powder retards the reaction rate, and therefore the liquid
phase exists up to higher temperatures, allowing a more pronounced elongated
grain growth [342]. Coarse b starting powder (3 lm) can lead to a reduced
formation of the ass for kinetic reasons and increase the grain size of the
resulting ass [339].
6.3
Development of Microstructures in ass/bss Ceramics
Since the relevant systems include two phase regions between a and b solid
solutions a production of ceramics with microstructures of both phases is
possible [343]. Depending on the composition of the a-rich starting powder
mixture, the formation of the phase assemblage of the ass/bss ceramic can be
realised by all the densi®cation methods described in Sect. 5.3.1. In general,
the equilibrium in the two phase region ass + bss is approached in about 1 to
2 h at temperatures between 1700 °C and 1800 °C; times increase with
decreasing temperatures. Amount and ratio of the two phases, as well as the
lattice parameters, vary with composition, in agreement with the equilibrium
conditions in the relevant phase diagrams (Sect. 3). The aim of those
compositions is to combine the advantages of ass, namely high hardness with
the toughness of bss [334, 336]. Typical microstructures of mixed ass + bss
ceramics consist of needle-like bss grains in a matrix of equiaxed ass grains

with some amount of secondary phase at the grain boundaries [344].
6.4
Characterisation of Microstructure
For reliable determination of the microstructure, a perfect preparation of the

specimens is an absolute necessity [345].
Materials are three-dimensionally arranged. Thus, the two-dimensional
information which is gained from the examination of microsections or thin
foils, the types of specimen usually used for materialographic investigations,
does not truly re¯ect the whole structure of a material. This becomes clear
from the photomontage (Fig. 25a). It shows a cube of silicon nitride ceramic in
Fig. 25a, b. Stereological analysis of microstructures of Si3N4 ceramics. a photo montage of

an Si3N4 ceramic superimposed upon a pile of Si3N4 crystallites. b the statistic extrapolation
of a microstructure with a high proportion of elongated grains
which the matrix phase has been dissolved away leaving behind a pile of Si3N4
grains. The upper surface of the cube represents a normal two-dimensional
section. It is obvious that the microstructural morphology shown in the
section, and in particular the grain size (D) and degree of elongation (aspect
ratio) of the grains which signi®cantly affect the fracture behaviour of the
material, are not correctly quanti®ed. The quantitative characterisation of the
two microstructural features, grain size and aspect ratio, becomes possible
only if the visible surface in the two-dimensional section can be extrapolated
to reveal the spatial shape of the grains. The transformation shown in Fig. 25b
is achieved using a stereographic-based variation of the principles used in
microstructural modelling [346, 347]. The histogram shown in the left part of
the diagram has been constructed solely from two-dimensional measurements
on the specimen. Accordingly more than 80% of the grains would have an
almost equiaxial shape. A statistical extrapolation of the grain population,
however, reveals a very different size-aspect ratio distribution, as can be seen
in the right-hand part of the diagram. This is because in this type of
microstructural modelling, the many different ways in which the individual
grains can be intersected (Fig. 26) are taken into account, according to their
size, aspect ratio and orientation. Often two-dimensional pictures show only
minor deviations between the microstructures of materials with pronounced
differences in properties, if one ignores the single elongated grain which has
quite by chance been sectioned completely along its longitudinal axis. In
comparison, the three-dimensional evaluation reveals quite signi®cant differ-
ences between the materials.
This basic approach to the special interpretation shows that there are
correlations which cannot be studied satisfactorily using two-dimensional
sections. However, they also show that there are very ef®cient methods to
extrapolate three-dimensional realities from two-dimensional measurements.
6.5
Microstructure/Property Relations
As a consequence of the large diversity of microstructures which can adjusted

by microstructural design, Si3N4 ceramics are a whole class of materials with
an inherent large variety of properties and therefore a large variety of potential
applications (Sect. 10). Different qualities depend on amount and distribution
of the microstructural features (Sect. 6.1.4). Often small variations have severe
Fig. 26. Some selected cases for planar sections through the hexagonal prism. Generally,
triangles up to octagons have to be considered
consequences with respect to the property pro®le. A more general overview on

the microstructure/property relation is in Table 12.
Materials with a high room temperature strength exhibit a ®ne-grained,
elongated microstructure, while materials with a high fracture toughness are
more coarse-grained [348]. In both cases, a weak grain boundary phase is
required to introduce transgranular fracture. Since all bss grains are
completely wetted by the grain boundary phase, the interface strength is
determined by the additive composition [349]. Nevertheless, a contradiction
arises between the development of high strength, high toughness Si3N4
ceramics and high temperature resistant materials because the grain boundary
phase is responsible for the excellent properties at low temperatures, but limits
the properties at temperatures above its softening point.
There is a signi®cant reduction in strength when microstructures consisting
of a broad grain diameter distribution are generated. When a ®ne equiaxed
microstructure is generated, both the fracture strength and the fracture
resistance are reduced [33, 350].
In Young's modulus and hardness of extended bss decreases linearly with
increasing number of replacing ions z [Eq. (5)]. The decrease was ascribed to
the lattice softening which was con®rmed by Raman spectroscopy; at high
Table 12. Overview on microstructure/property relations of Si3N4 ceramics
Property Microstructural features
Grain size/shape Grain boundary phase
High strength up to Fine grained/needle-like grains Median additive content

1000 °C
High strength at Fine grained/needle-like grains Al2O3 free, special compositions
T > 1200 °C
High fracture toughness Large, needle-like grains, Low Al2O3 and SiO2 content
or large, needle-like grains
in a ®ne matrix
High hardness Fine grained or ass/bss Low additive content
or ass materials
High fatigue strength Fine grained/needle-like grains Low additive content
(cyclic mechanical
load)
High heat conductivity Large grains, low amounts of No components solvable
impurities and defects and in the Si3N4 crystal lattice,
solid solutions (e.g., Al, Be), no impurities
High creep resistance Large, needle like grains, Al2O3 free with special composi-
composites with SiC, or tions or no sintering additives
refractory silicides
High oxidation Large, needle-like grains, Al2O3 free with special
resistance at composites with SiC, or compositions or no
T > 1200 °C refractory silicides sintering additives
High corrosion Special compositions depending
resistance on the corrosive media
Good wear behaviour Fine-grained microstructure Homogeneous distribution
temperatures, creep rates were signi®cantly enhanced with increasing z values

[351]. In general, the creep resistance decreases with increasing amounts and
decreasing viscosity of the intergranular amorphous phase, but depends also
on size and distribution of Si3N4 grains and the oxidation behaviour
[352±355]. The oxidation behaviour strongly depends on the additives [352,
356, 357].
The hardness of ass is signi®cantly higher than that of bss (Table 1)
[332±337, 340] and, in case of elongated grain morphology, the fracture
toughness can be signi®cantly improved by crack de¯ection, crack bridging
and grain pull-out mechanism up to KIC values of 6.3 MPa1/2 [358].
The rising fracture resistance with the extension of the crack (R-curve
behaviour [349, 359]) is more expressed in coarse-grained materials [360]. The
R-curve behaviour also has a signi®cant contribution with respect to thermal
shock behaviour; during thermal shock a continuous strength degradation is
observed in materials with a pronounced R-curve behaviour, in contrast to a
catastrophic failure of ®ne-grained materials with a minor R-curve behaviour
[360, 361]. In Si3N4 ceramics with gradient microstructures compositions,
microstructures and properties change gradually from the hard ass with
spherical grains on the surface to the tough and strong bss with elongated
grains in the core, a promising combination of microstructures for wear
applications [362].
The microstructure/property relations are treated in more detail in Sect. 7.
7
Properties
Si3N4 ceramics have a bunch variety of interesting properties. They are light,
have good mechanical and thermomechanical behaviour, they are compatible
with metals, and they are wear and corrosion resistant. The unique inherent
properties of the Si3N4 modi®cations (Sect. 2), the reasonable alloying
behaviour (Sect. 3) and the variability in the production processes (Sects. 4
and 5) open a wide range for tailoring microstructures (Sect. 6) and
remarkable property combinations.
7.1
Physical Properties
The electrical resistivity of Si3N4 is >1014 Wcm at room temperature and

>106 Wcm at 1200 °C [363]. The electrical conductivity of bss only slightly
increases with increasing degree of alloying substitution z according to Eq. (5)
at 700 °C (2 ´ 10)9 (Wcm))1 for z = 1.5; 2 ´ 10)7 (Wcm))1 for z = 3.2)
[364, 365]. The conductivity is ionic at temperatures above 900±1000 °C and
electronic by impurities below [364, 365]. At temperatures below 800 °C the
activation energy is around 1 eV and 1.8±2 eV greater around 1000 °C.
Therefore Si3N4 or bss ceramics are used as electronic insulators. Li-
containing ass has an electrical conductivity at high temperatures which is
similar to that of solid electrolytes, e.g., b-alumina; the conductivity of the ass
composition Li2Si9Al3ON15 is 9´10)6 (Wcm))1 and the activation energy of

0.93 eV is low. An electroconductive bss material (ionic conduction) was
developed by designing the grain boundary phase [366]. Dense sintered and
hot pressed Si3N4 ceramic exhibit an electric break down strength
>200 kV cm)1 [367]. The low dielectric losses tan d and dielectric strength e
(6±8.0 at 1 MHz) in a broad frequency range is advantageous for radar
windows [368, 369], but disadvantageous for microwave sintering [370].
The speci®c heat of Si3N4 ceramics is in the temperature range 293 up to
1200 K [Cp (293 K) = 0.67 KJ (K kg))1] nearly independent of the composi-
tion of the additives. The isobaric speci®c heat values agree well with the
isochoric speci®c heat calculated by Debye's theory. Also the Dulong Petit's
rule can applied as an approximation of the Cv values [25 J(K mol))1] at
temperatures >1100 K [371]. From the Cp values at around 100 K the amount
of the amorphous grain boundary phase can be calculated [371].
Thermal diffusivity a and thermal conductivity k strongly depend on the
composition (Fig. 27). Materials containing a high residual a content show a
very low thermal diffusivity (88 vol% a; a = 5 mm2 s)1) whereas the same
composition after complete a/b transformation has 13 mm2 s)1 at RT. ass
ceramics also show low thermal diffusivity (Fig. 27). The intrinsic anisotropic
thermal diffusivity inside the individual grains is without connection with the
macroscopic diffusivity, as a consequence of the dispersion of grain
orientations in the ceramics [375].
The thermal conductivity of bss depends strongly on the amount of Al
impurities and/or sintering additives because the incorporation of Al and O in
the b structure reduces the thermal conductivity of the grains due to the
reduced free path of phonons. At RT the thermal conductivity initially
increases with increasing grain size before reaching constant values [372]; it
depends on internal defects in the grains (dislocations, Al and O impurities,
point defects) [373]. Al-free sintering additives are a precondition for the
production of Si3N4 ceramics with high thermal conductivity. With increasing
sintering time the thermal conductivity increases, because of grain growth and
Fig. 27. Thermal diffusivity versus temperature of bss with different Y2O3/Al2O3 ratio from
6/0 to 0/6 (mol/mol) [371] and ass (n = 1.4, m = 1) [379]
defect healing [374]. The thermal conductivity of b-Si3N4 ceramics strongly

decreases with increasing grain boundary ®lm thickness [372]. In hot pressed
ceramics with different rare earth oxide additives the thermal conductivity
increases in the order La [31.6 W(mK))1], Nd [81.6 W(mK))1], Gd
[100.7 W(mK))1], Y, Yb [115 W(mK))1], because of the decreasing amount
of oxygen dissolved in the b grains [374, 376]. The highest values for the
thermal conductivity of isotropic Si3N4 ceramics are around 100 W(mK))1, for
ceramics with anisotropic grain orientations values of 160 W(mK))1 were
measured [377] (Sects. 2 and 9.3).
The linear thermal expansion coef®cient a of the bss materials depends
only slightly on the additive composition [301]. Between RT and 1000 °C
values between 3.1 up to 3.6 ´ 10)6 K)1 are common; they increase slightly
with increasing temperatures. For b-Si3N4 ceramics containing 6 wt% Y2O3
and 4 wt% Al2O3 a = 2.8 ´ 10)6 K)1 between RT and 200 °C, and between RT
and 500 °C a = 3.2 ´ 10)6 K)1.
Compared to bss, the linear thermal expansion coef®cient a(RT±1000) for
ass is higher; depending on the cation RE in RE0.6Si9.3Al2.7O0.9N15.1 values
between 3.8 ´ 10)6 K)1 (RE = Er; Nd) and 3.7 ´ 10)6 K)1 (RE = Y) and for
the Ca containing ass 3.9 ´ 10)6 K)1 were measured [378].
7.2
Mechanical Properties
Especially strength and fracture toughness are essential for structural

applications. Therefore, great efforts are made to optimise all preconditions
for a reproducible production of reliable Si3N4 ceramics with high strength
and toughness [19].
The strength r of a brittle material is proportional to the fracture toughness
KIC and indirectly proportional to the square
p root of the highest defect size
Öa in the loaded volume: r Y KIC = a (Y is a geometry factor). For
optimisation a must be reduced and KIC increased. Whilst the reduction of a is
possible by optimising all the processing steps (Sect. 5), an improvement of
KIC is mainly possible by microstructural engineering (Sect. 6).
Stresses arising during application may cause defect growth (subcritical
crack growth) which at room temperature only is activated for stresses close to
the fracture stresses, but at elevated temperatures above the softening point of
the glassy grain boundary phase, they are more pronounced. In addition,
softening of the grain boundary phase (Tg) is the reason for deformation
processes at high temperatures (creep). The typical behaviour of Si3N4 ceramics
under static load is shown schematically in a fracture map (Fig. 28). At
temperatures up to the softening point of the glassy phase (for oxide nitride
glasses formed by common sintering additives between 800±1000 °C) the
materials behave in a brittle fashion with low subcritical crack growth and high
oxidation resistance (Sect. 7.2.1). At temperatures higher than the transforma-
tion temperature, creep, creep fracture, subcritical crack growth and damage
due to oxidation become the dominating processes (Sect. 7.2.2). For the
measurement of mechanical properties several methods are available [380±383].
Fig. 28. Fracture map of Si3N4 ceramics schematic [382]
7.2.1
Room Temperature Properties
The strength of commercial Si3N4 ceramics are in the range of 800 to

1400 MPa, depending on defects like pores, cracks, inclusions such as iron
silicides or agglomerates of sintering additives. In materials with higher
strength or high fracture toughness, unusual elongated grains can be strength-
determining defects. So far, the highest strength for isotropic Si3N4 ceramics is
2000 MPa measured by three point bending tests [246]. Recently a strength of
2100 MPa was measured for materials with anisotropic grain orientation
produced by super plastic forging [384]. Materials with strengths between 1400
to 1500 MPa usually have defect sizes of about 10 lm [245, 310]. Therefore
their grain size should be below 10 lm, because larger grains can act as
strength reducing defects.
For strength levels above 1000 MPa a special surface ®nishing is necessary
because surface defects are strength limiting. Compression stresses in the
surface reduce the effective tensile stresses acting on surface defects and thus
increase the strength. In contrast tensile surface stresses reduce the strength.
Internal stresses, different sizes and orientations of surface defects may reduce
the strength by more than some hundred MPa [385]. A careful reproducible
®nishing of the ceramics for strength testing and application at high loads is
absolutely necessary [385]. In the as-sintered state the strengths of the Si3N4
ceramics are usually lower than after removing of the surface layer which is
decomposed during sintering. 750 MPa is measured for materials in the as-
sintered state with minimised decomposition reactions during sintering [386].
Porous Si3N4 ceramics with oriented b whiskers and high strength have a high
strain to failure relation at RT due to reduced elastic constants [387].
The fracture toughness varies in a wide range from 3 to 12 MPa m1/2, this is
on one hand connected with variations in the microstructure and on the other,
by different methods of determination giving slightly different values [380,
381, 383]. The fracture toughness depends strongly on the microstructure. Two
main factors in¯uence the fracture toughness: grain shape and size, and the
composition of the grain boundary phase.
The comparatively high fracture toughness of Si3N4 ceramics in comparison
to other ceramic materials is related to the toughening mechanisms which are
similar to those in whisker reinforced composite materials: grain bridging,
pull-out, crack de¯ection and grain branching around large, elongated grains
[359, 388±390]. Due to these mechanisms, the fracture toughness increases
with increasing volume and the square root of the mean grain thickness of the
elongated grains (grains with aspect ratio > 4) [304, 349, 359, 388, 390±393].
The dominant toughening mechanism depends on the grain thickness of the
elongated grains. Elastic bridging and pull-out were observed for thin, needle-
like grains (thickness < 1 lm). Crack de¯ection was mainly observed for
thick, elongated grains (thickness > 1 lm), whereas grain bridging was
detected independent of the grain size [388]. These mechanisms are also
responsible for a pronounced rising fracture resistance with crack extension
(R-curve behaviour) [303, 304, 390, 391].
The toughening mechanisms can only operate when the dominant fracture
mode is intergranular. The ratio of transgranular to intergranular fracture
depends on the relative strengths of the grain boundaries and the grains. For a
material with a high toughness, the grain boundary must be weak in
comparison to the grains (Fig. 29) [33, 389]. Materials with large, needle-like
grains in the matrix (in-situ-reinforced) can be produced by prolonged
sintering times, special additive combinations or by seeding the materials with
b-Si3N4 whiskers. Seeding leads to a more narrow size distribution of the large
grains and allows one to achieve textured materials with large, oriented
b-Si3N4 needles [377, 394] (Sect. 9.3). A similar anisotropy was observed in
hot pressed materials with low amounts of sintering additives [395] or in
superplastic forged materials [384]. Perpendicular to the hot press direction a
higher strength and fracture toughness is observed than in the parallel
direction.
Recent investigations show that ®ne-grained materials have a higher
toughness than coarse-grained materials in the small crack region (crack
length < 30 lm) but lower toughness in the large crack region (crack
length > 50 lm) [396±398]. This can be important for the applications in
which high local stresses exist (e.g., for ball bearings).
The strength of the grain boundary is connected with two different
mechanisms: local residual stresses [399] and special chemical interactions
between the grain boundary phase and the Si3N4 grains [33, 389]. Generally,
the amorphous or partially crystallised grain boundary phases have thermal
expansion coef®cients different from Si3N4. When the thermal expansion
coef®cient of the grain boundary phase is higher than that of the Si3N4 grains,
the grain boundary phase is under tensile stresses and the fraction of
intergranular fracture is high. This is the case for nearly all Si3N4 ceramics. In
Fig. 29. Crack path in a Si3N4 ceramic a transgranular with low and b intergranular with high
fracture toughness
consequence the fracture toughness is high. In contrast, materials with grain

boundary phase under compression (e.g., HIPSN without sintering additives)
have low fracture toughness due to a high percentage of transgranular fracture
(Fig. 29a) [400].
The residual stresses can be in¯uenced during the sintering cycle, either by
changes of composition of the grain boundary phase (e.g., by evaporation of
SiO2), or by crystallisation of the glassy phase or by partial relaxation of
stresses (e.g., by slow cooling) [357]. These changes are of minor in¯uence and
are usually outweighed by grain size, shape and composition, which have a
more pronounced in¯uence on stress state and fracture toughness.
The composition of the grains also in¯uences the fracture toughness
because of the special chemical interactions that occur between the grains and
the grain boundary phase. The formation of Al and O rich bss layers on Si3N4
grains, which is especially pronounced in bss materials, results in an increase
of transgranular fracture and a decrease in the fracture toughness [33, 401].
ass materials with relatively coarse, elongated grains show a high percentage
of intergranular fracture and exhibit a high fracture toughness [332, 333, 337,
342].
The subcritical crack growth in Si3N4 ceramics takes place only at stresses
very close to the fracture stresses, it is low in comparison to other ceramic
materials. The growth rate exponent is in the range of n = 30±300 for static
loads and n = 20±120 for cyclic loading [293, 396, 402±404]. The growth rates
under static and cyclic loading accelerates in water and independent of pH
between 2 and 14, similar to that of glasses [404]. Oxide nitride glasses are
more resistant to subcritical crack growth than oxide glasses [405]. This,
together with toughening mechanisms, may be responsible for the lower
subcritical crack growth rate of Si3N4 ceramics under static load in comparison
to oxide ceramics. Until now, no clear correlation could be made between the
microstructure and composition of the material, on the one hand, and the
subcritical crack growth on the other hand.
Cyclic fatigue of Si3N4 ceramics is more pronounced than static fatigue, i.e.,
in cases where under static loading no crack growth occurs, pronounced crack
growth can take place under cyclic loading. This is connected with the
degradation of bridging grains, i.e., the toughening mechanisms can only work
partially [403, 404]. The most intensive damage takes place under cyclic
loading conditions with alternating compressive and tensile stresses, as a
result of the destruction of the bridging large grains. Under cyclic loading,
ceramics with smaller grain size have the higher crack growth exponent, i.e.,
have a lower degradation of strength during loading and a higher life time at a
given strength level [402].
The hardness of Si3N4 ceramics depends on the phase composition because
of the different values of the different phases (Sect. 2). Compositions with a
high amount of a, which is not transformed during densi®cation, have
hardnesses as high as ass (up to 20 GPa) [17, 406]. The macrohardness of
different b-Si3N4 ceramics increases with decreasing grain boundary phase and
grain size [392]; the dependence of hardness on the microstructure is opposite
to that of the fracture toughness. Usually the values for b-Si3N4 ceramics
(HV10) are in the range of 12 GPa (coarse grains, high additive content) to
16 GPa (®ne grains, low additive content). The hardness of two phase ass/bss
materials changes linearly with the phase ratio [407]. The dependence of the
microhardness on the grain size is more complex and depends on the ratio of
indentation to grain size [408]. The hardness at elevated temperatures depends
additionally on the softening of the glassy grain boundary phase. Materials
with MgO/Al2O3 as sintering additives show a faster degradation of the
hardness than materials with more refractory grain boundary phases.
Si3N4 exhibits a high thermal shock resistance due to the combination of
low thermal expansion exponents, high strength, medium elastic constants,
and reasonable thermal conductivity [360, 409 ], e.g., Si3N4 ceramics can
withstand quenching in cold water from up to 800±1000 °C whereas Al2O3 or
ZrO2 ceramics withstand such a procedure only up to 200±400 °C [360, 409].
Especially in-situ reinforced materials have high thermal shock resistance
[383].
7.2.2
High Temperature Properties
As a consequence of the strong covalent bonding the properties of the Si3N4

grains do not change up to temperatures of 1600 °C but the grain boundary
phase already begins to soften at lower temperatures. Depending on amount
and composition of the grain boundary phase various processes (diffusion,
creep, slow crack growth, oxidation, corrosion) may occur at elevated
temperatures (Fig. 28) with the consequence that a new defect population is
generated which determines the failure behaviour and limits the lifetime. The
extent to which these processes occur is mainly in¯uenced by the softening
point and viscosity of the amorphous grain boundary phase [382, 410].
Improvements in all processing steps, and quality of powders allow the
increased use of refractive additives, e.g., rare earths which result in
remarkable increases of creep resistance, strength and life time of Si3N4
ceramics (Figs. 30 and 31) [411±415, 420].
Creep curves of Si3N4 at high temperatures generally consist of three
regimes: transient, steady-state, and accelerated creep, similar to metals. The
creep rate under tensile stresses is some orders of magnitude higher than
under compression [412, 416].
Different creep mechanisms are discussed in references [383, 412, 413,
416±421]. Diffusional creep is unlikely to be the rate controlling process in
Si3N4 ceramics with considerable amounts of glassy grain boundary phase.
Also dislocation motion contributes only little to creep below 1700 °C. The
creep mechanism in Si3N4 ceramics is strongly correlated with the grain
boundary phase. The dominating processes are material transfer by solution-
reprecipitation through the viscous phase and rearrangement by viscous ¯ow
[417, 421, 422], formation of cavities [418, 421], or cracks accompanied by
Fig. 30. Creep behaviour of Si3N4 ceramics with different sintering additives under 150 MPa
static load [413]
Fig. 31. Time-to-failure behaviour of Si3N4 ceramics under bending load at 1400 °C.
Si3N4/MoSi2; Si3N4/SiC composites with 8 wt% Y2O3, HPSN Si3N4 ceramic with 8 wt% Y2O3,
SSN6Y4Al shows the typical behaviour of Si3N4 ceramic with Al2O3-containing sintering
additives
grain boundary sliding [417, 420], and redistribution of the secondary phase
resulting in cavitated multigrain junctions [412, 416]. The rearrangement by
viscous ¯ow leads to a logarithmic dependence of the creep rate on stress and
explains the high stress exponents for Si3N4 ceramics (e_ rn ; e_ creep rate; r
applied load, n stress exponent) [412, 413, 416]. Thus, creep behaviour of Si3N4
ceramics is largely dependent on characteristics and behaviour of the glassy
phase. Some disadvantages can be avoided by partial crystallisation of the
amorphous phase; this is especially bene®cial to creep and slow crack growth
behaviour [413].
In HIPSN with no sintering aids the only liquid phase during sintering is the
silica adhered even to high-purity Si3N4 powders. As the consequence of the
small amount and the high softening point of the grain boundary phase,
the materials exhibit excellent creep behaviour at temperatures up to 1500 °C
[400, 423]. A more refractory grain boundary could only be obtained by
removing the silica from grain boundaries and triple junctions. But this is not
possible and each addition of only small amounts of impurities or sintering
additives will change the chemistry of the grain boundary phase and weaken
the material [411, 424]. Besides the costly fabrication, the main disadvantage of
these materials is their relatively low strength (500 MPa) and fracture
toughness (3±4 MPa m1/2) at RT.
Glass-forming and stabilising sintering additives which result in a silicate
phase with a low softening point and low viscosity such as MgO, Al2O3 or AlN
are categorically unsuited for Si3N4 ceramics for applications above 1200 °C.
Refractory silicates forming intergranular phases with a high crystalline
content were obtained by yttria or other rare earth oxides as sintering
additives [413, 414, 423, 425]. Superior creep behaviour is found with Lu2O3 as
sintering additive exhibiting Lu2Si2O7 and Lu4Si2O7N2 grain boundary phases

[413, 414]. It is supposed that the main creep mechanisms are solution-
reprecipitation steps accompanied by signi®cant suppression of cavitation
[413]. It is known, that the bond strength of the rare earths increases with
decreasing ionic radii, i.e., in the order La¼Y, Yb, Lu. However, the change of
the glass properties of oxide nitride glasses with changing rare earth additive
can not alone explain the drop in creep rate going from Yb2O3 to Lu2O3 [413,
414]. There are some indications, that the thickness of the grain boundary
phase is reduced and that the superior properties are connected with a more
complete crystallisation of the grain boundary phases [413, 414]. Fine grained
ceramics with grain boundary phases of low viscosity have high creep rates
(e.g., 10)4 s)1 at 60 MPa and 1550 °C) and show superplasticity (deformation
without remarkable mictrostructural deterioration) which can be used in
shaping [426].
At high temperatures exist two regions in the fracture mechanism map of
Fig. 28: slow crack growth failure and creep fracture [413, 414]. The former
occurs when a crack grows subcritically from a pre-existing ¯aw and
reaches the critical size. This is predominant in the high-stress, short-life
time region and varies with the kind of additives [427, 428]. The creep
fracture is due to the formation of a macrocrack of the critical size by cavity
nucleation and coalescence, which prevails in the low-stress region and long
times. For materials with rare earth additives the change between crack
growth failure and creep fracture occurs at tensile stresses of
>200 300 MPa at 1400 °C.
For long-term high temperature application, oxidation damage has a
signi®cant in¯uence on the lifetime. The migration of sintering additives
toward the outer surface, the pore formation in the bulk and the pit formation
due to local enhanced oxidation leads to an environmentally caused
degradation of the properties. High long term stability in oxidising atmo-
spheres was found for HIPed materials without sintering additives and
materials with Lu2O3 as sintering additive. A high long-term stability was
found also for Si3N4/SiC and Si3N4/MoSi2 composites having a different
oxidation mechanism compared to monolithic Si3N4 ceramics (Fig. 31). This is
caused by a reduced redistribution of the liquid phase and pore formation in
the bulk [410, 423, 429] (Sect. 9.2).
Two phase ass/bss materials are of interest with respect to their HT
behaviour [334, 430]. Due to the ability of the ass grains to incorporate
cations from the sintering aids, it is possible to modify the grain boundary.
With a large amount of ass, a skeleton of strong grain boundaries between the
ass grains is formed due to incorporation of cations in the ass structure, with
the consequence of an improved creep resistance. The oxidation resistance
cannot be improved by the increase in the ass content; therefore these
materials can only be used up to 1300±1350 °C for long-term applications
[334].
In principle, the amount, composition and degree of crystallisation of the
grain boundary phase are key factors which must be considered for successful
development of Si3N4 ceramics for applications at elevated temperatures. It is
possible to provide Si3N4 ceramics for long-term applications up to temper-

atures of 1500 °C.
7.3
Chemical Properties
Like all other non-oxide ceramics Si3N4 is metastable in air or combustion

gases, both at room and at elevated temperatures. Detailed understanding of
oxidation and corrosion mechanisms and the in¯uence of the surrounding
atmosphere on the lifetime are necessary before Si3N4 ceramics can be applied
under oxidising or corrosive conditions [431±437].
7.3.1
Oxidation
One can distinguish between active and passive oxidation. During active
oxidation the oxidation products are immediately removed from the surface,
causing weight loss, whereas during passive oxidation a weight gain takes
place, because oxide layers formed.
Normally active oxidation occurs at low oxygen pressure and high
temperatures. The transition from active to passive oxidation depends on
temperatures and pressures, which are predictable by coupling local equilibria
with transport processes (Fig. 32) [431, 436, 438, 439]. With increasing
nitrogen pressure the temperature increases and the oxygen partial pressure
decreases [438]. In addition, the transition is in¯uenced by the activity of SiO2,
i.e., for glassy surface layers with higher amounts of additional compounds
Fig. 32. Dependence of the active to passive transition of Si3N4 ceramics on the temperature
and oxygen partial pressure [438]
such as Y2O3 or Al2O3 it takes place at higher temperatures and/or lower

pressures [438].
For perfect surface layers the passive to active transition occurs at much
lower oxygen pressure than the active to passive transition [431, 439]. In
real systems, however, the transition temperatures are identical because of
imperfections (cracks, bubbles, etc.) in the layer [438, 440].
In air the active/passive transition takes place at about 1800 °C (Fig. 32). In
high temperature oxidation experiments the begin of active oxidation in air
was observed between 1600 and 1650 °C, caused by destruction of the layer by
bubble formation and evaporation of gaseous reaction products such as
Si(OH)4 or SiO(OH)2. The active to passive transition and the accompanying
weight loss is altered by gas stream velocity, gas viscosity, laminar or turbulent
¯ow regime of the gas. Beside the oxygen pressure, H2O, CO or H2/H2S in the
gas can initiate active oxidation [431, 434, 436]. Active oxidation takes also
place in high velocities of water vapour and gas ¯ow below the values
predicted in classical theory (Fig. 32) [431, 436, 441±443]. In this case the SiO2
surface layer reacts with water to gaseous Si(OH)4 or SiO(OH)2 [441]. This is a
most serious problem for application of Si3N4 ceramics in high temperatures
gas turbines. Intensive research on special coatings or surface layers for
environmental barrier coatings has not been very successful yet [442, 443].
The different reactivity of phases in multiphase systems results in selective
attack and pit formation, reducing strength or causing an enrichment of
sintering additives such as Y2O3 or Al2O3 at the surface [444].
The passive oxidation of Si3N4 ceramics is mainly in¯uenced by the
protective layer formed at the surface by the oxidation process and the ability
of this layer to prevent oxygen from diffusing into the material. A surface layer
of pure SiO2 causes a very low rate of oxygen diffusion into the material [445,
446]. However, diffusion of sintering additives or impurities into the oxide
layer lowers its viscosity and increases the oxygen diffusion, i.e., it decreases
the protection and enhances oxidation [432, 436, 445, 447±449]. Therefore
CVD-Si3N4 and HIP SN (no additives) have the highest oxidation resistance of
all Si3N4 ceramics because of their perfect protective layer (Fig. 33). For these
materials a diffusion/interface controlled reaction was found [450, 451]. At low
oxidation temperature (<1400 °C) in CVD-Si3N4 a thin additional Si2N2O sub-
layer with lower diffusion rate has been observed. Presence or absence of an
additional Si2N2O sublayer is still under discussion and depends on the
condition and purity of the material [436, 450, 452, 453]. An amorphous oxide
nitride layer with changing oxygen/nitrogen ratio may also occur [453, 454].
No Si2N2O sublayers have been observed in additive containing materials, but
Si2N2O forms in the bulk [445, 449].
In Si3N4 ceramics containing additives, the protective layers are complex
and composed mostly of the amorphous glassy phase containing some of the
sintering additives and cristobalite and silicate phases (Fig. 34) [432, 435, 445,
447, 449, 455]. The temperature of the SiO2 rich eutectic of the oxide nitride
systems including the additives is the limit for long-term applications. Above
the eutectic temperature higher amounts of liquid are formed, causing
accelerated degradation. For example, the temperature of dramatic oxidation
Fig. 33. Parabolic rate constant Kp of different Si3N4 ceramics [455], (SSN Mg/Al [465];
SSN-Y/Al [466]; HPSN Y/0.2 Al [466]; HPSN Mg/Y [467]; HPSN Y and HIPSN (no additives)
[445]
increase is 1350 °C for materials with Y2O3/Al2O3 additives while the materials
only with Y2O3 are stable up to 1500 to 1550 °C [15, 445].
Whereas CVD Si3N4 and additive-free HIPSN oxidise at the outer surface,
only ceramics containing sinter additives are susceptible to internal oxidation,
because of their lowered protection by the surface layer and higher grain
boundary diffusion of the oxygen (Figs. 34, 35).
The formation of SiO2 by the oxidation causes ¯uxes of cations from the
bulk to the surface. In the near surface area an enrichment of additives is
found, leading to a damage also of the bulk. Grain boundary diffusion of
oxygen into the bulk and the oxidation of the Si3N4 grains below the outer
surface results in an SiO2 enrichment, causing ¯uxes of additives from bulk to
surface and an enrichment of additives in the near surface region thus leading
to bulk damage (pore formation) and a strong degradation of strength
(Figs. 34, 35) [445, 448, 455±458]. This process is especially pronounced at
high temperatures, when the protection by the surface layer is reduced and the
grain boundary diffusion is faster than the interfacial oxidation [448].
For MgO-containing materials the Mg diffusion is the rate controlling step
[459]. The in¯uence of MgO and Y2O3 on the oxidation rate depends on the
changing amount of cristobalite formed on the interface between the oxide
skin and the bulk. The in¯uence of rare earth elements on the oxidation
kinetics is much smaller than that of MgO [447].
In composites of Si3N4 with MoSi2 and SiC, a moderate damage of the bulk
due to oxidation was observed, leading to an increase in life time at high
temperature [429, 445, 460]. The reason is the rapid formation of Si2N2O near
Fig. 34a±d. Oxide layers on Si3N4 ceramics oxidised at 1500 °C. a HIP-SN (no additives;
2500 h), b SSN (Y2O3/Al2O3 additives; 1000 h), c SSN (Y2O3 additive; 5000 h), d Si3N4/MoSi2
composite (Y2O3 additive; 5000 h)
the surface, reducing the enrichment of SiO2, which is the driving force for the
diffusion of additives from the bulk toward the surface. Additionally, the
diffusion of oxygen into the material is reduced.
The oxidation stability of Si3N4 ceramics with different sintering additives at
temperatures above 1200 °C increases in the order: MgO; MgO/Al2O3;
MgO/R2O3 < R2O3/Al2O3 R2O3 (Fig. 33). The oxidation rates of ceramics
with different additives of rare earth oxides are reduced with decreasing ionic
radii, i.e., the oxidation rate decreases in the order La, Ce > Y > Yb > Lu > Sc
[413, 461]. Additionally, the oxidation rate depends on the amount of additives
and on the additives to SiO2 ratio. So the weight gain during oxidation of
Y2O3 containing materials can change by a factor of two, depending on the
SiO2/Y2O3 ratio (Fig. 36). Materials with Lu2Si2O7 and Lu4Si2O2N7 as grain
boundary phase show a very high oxidation and creep resistance, which
cannot be explained by the change of the radius of the rare earth ions in
comparison to Y2O3, Yb2O3-containing ceramics. The reason for this high
oxidation resistance appears to be a nearly complete crystallisation of the grain
boundary phase [413]. The oxidation (parabolic oxidation constants) of
ceramics with MgO increases non-linearly with rising MgO/SiO2 ratio at
constant amount of additives.
The strong dependence of the oxidation on composition of the ceramic as
well as the amount and state of the grain boundary phase (composition and
crystallisation ability) is the reason for the scatter of the oxidation constants
shown in Fig. 33, i.e., small amounts of impurities may change the oxidation
rate by several orders of magnitude [455, 459]. Especially cation impurities
Fig. 35a±c. Schematic representation of the processes during oxidation. a HIP-SN no additives,
b SSN containing Y2O3/Al2O3 additives, c Si3N4/MoSi2 composite with Y2O3 additive
Fig. 36. Dependence of weight gain during oxidation (1500 °C 1000 h) and residual strength
after oxidation of HPSN with different SiO2/Y2O3 ratios [446]
such as Na+, K+ inhibit the crystallisation and lower the viscosity of the oxide
®lm and reduce the oxidation resistance dramatically [459].
The oxidation rates do not correlate with the strength degradation in
all cases. For materials with different SiO2/Y2O3 ratios, the oxidation rate
increases with increasing Y2O3 content and the residual strength also increases
after oxidation (Fig. 36). This tendency is more pronounced for composite
materials with MoSi2 or SiC, exhibiting nearly the same oxidation rate as the
monolithic materials but a twice the strength after oxidation at 1500 °C. The
reason is the different segregation of the grain boundary phase and a reduced
pitting tendency (Sect. 9.2) [429, 445, 460].
Different rate laws were proposed for the oxidation mechanisms. The most
common is the parabolic law (diffusion controlled process) [436]. For the
starting period of the oxidation more complex laws are proposed, including
linear, logarithmic or arctan functions of time [436, 462]. The simultaneous
crystallisation of the oxide layer, causing a reduction of the diffusion
coef®cients in the oxide scale, leads to a logarithmic law [436, 455].
Additionally cracks, bubbles and other defects in the surface layer will
in¯uence the kinetics and cause deviations from the parabolic law [436, 455,
462]. Ceramics with nitrogen-rich grain boundary phases exhibit accelerated
oxidation in the range of 900±1100 °C ± the so-called catastrophic oxidation.
This is caused by the absence of a dense oxide layer [463]. The stresses caused
by the volume change during the oxidation of the grain boundary phase lead to
cracks and new surfaces undergoing oxidation and ®nally to a fast destruction.
This process increases with increasing additive content, but can be prevented
by a short heat treatment between 1200 and 1400 °C to develop a protective
SiO2 surface layer.
The oxidation behaviour of Si3N4 ceramics strongly depends on impurities
in the gas atmosphere. Impurities like alkaline or alkaline earth metals, SO2,
and vanadium drastically decrease oxidation [431, 433, 434]. The main
in¯uence of the different impurities is caused by a change of the viscosity or
the destruction of the oxide scale, accelerating the diffusion of oxygen or water
vapour into the ceramic and increasing the corrosion. Of coarse, the effect
strongly depends on temperature and gas composition.
The corrosion by molten salts was intensively investigated in connection
with impurities of combustion gases [431, 433, 434, 436, 464]. The corrosion
effect of NaCl in combustion environments is less pronounced, when the
sulphur concentration in the fuel is higher [431]. The reason for this behaviour
is that at high sulphur concentration Na2SO4 is stable and thus the Na2O
activity and the formation of sodium silicates is reduced [431]. These
processes are analysed in [431, 433, 436].
7.3.2
Interaction with Metals
Information on the interaction of Si3N4 ceramics with metals is important for

understanding of the behaviour of metallic impurities during sintering, for the
joining to metals [468, 469], for application of Si3N4 materials in metallurgy,
and as cutting tools [470, 471]. The interactions strongly depend on the
nitrogen/oxygen pressure in the atmosphere and on temperature.
In Table 13 the interactions of Si3N4 ceramics with common metals are
summarised (see also [18, 472, 473]). For application in metallurgy not only
the interaction with the metal but also the interaction of oxide slags on the
surface of the metals has to be taken into account. At higher temperatures
most metal oxides react with the grain boundary phase. For example, V2O5,
122
Table 13. Interaction of Si3N4 ceramics with metals
Metal Chemical interaction Wetting angle Remarks
Si Some solubility of Si3N4 in Depend on the oxide scale on the

Si: 2±4 á 10)6 at % N in Si(l) surface; at low oxygen pressure
[474, 475] wetting angle 45±50 ° [476, 477];
at decomposition conditions of
Si3N4 wetting was observed [478]
Alkali Li Li react at >500 °C forming Li2SiN2, Oxides react very strong with the
metals Si solves in Li [479]; can cause grain boundary phase; Si3N4 ®lms
cracking of the ceramic; can act as diffusion barriers for
Na Pure Na does not react up to 1700 K; Na(l) 20 ppm O2 at 400 °C±500 °C alkaline metals in semiconductor
impurities have an aggressive 85 °; at lower temperature it is devices [481]
in¯uence [480]; higher
K Pure K does not react [480]
Earth alkaline Be No reliable data exist Reaction of the oxides with the
metals Mg Reacts only slightly at ³750 °C; No wetting of pure Mg at 850±950 °C glassy grain boundary phase
at ³1000 °C under 100 torr Ar in vacuum or air 90±150° [482, 484]
MgSiN2 formation is observed [482]
Ca Ca SiN2 can be formed at 1300 °C Wetting, 10° at melting point [484]
[483]
3a-group Al Stable with molten Al [474, 485, 486]; No wetting near the melting point Extensively used in aluminium
metals different protective layers at the due to the existence of an oxide industry
interface can exist of 15-R polytypes layer on the surface; at 900 °C
[487, 488]; amorphous Al2O3; at 65° [468] at 1000 °C 126°[484]
high temperature AlN formation
[489]; Al 7 wt% Mg in®ltrate and
decompose RBSN [490];dense
materials are stable [490];Al/Ca-
alloys in®ltrate and react with RBSN
[474]; no reaction [474, 491]
Ga, No reaction [474, 491] No wetting [468, 492] Used in the GaAs crystal growth

In No reaction [474, 491] No wetting [468, 492] technique
4a-group Ge, No reaction No wetting [473, 496]
metals Sn Reaction [491]
Pb No reaction in vacuum or protective No wetting in vacuum or protective Strong reaction of Si3N4 with PbO,
atmosphere up to 800 °C [493] atmosphere [491] Pb: 100±150° PbO2 [493]
at 450 °C
3b-group Sc Depending on nitrogen pressure at
metals 1000 °C ScN, ScSi2)x or Sc5Si3 are
stable [494, 495]
La Si3N4 is stable with molten La; at
nitrogen pressure lower than
10±4 atm LaSi2 is in equilibrium with
Si3N4; at about 1600 °C and 1 atm
nitrogen LaSi2 and Si3N4 are
compatible [472]
3f-group Lan- Ce, Ho react with Si3N4 under Ar at Gd wets Si3N4 [497]
metals thanoides 1000 °C to form RESi2 ; at higher
nitrogen pressure the nitrides are
stable [495]
4b-group Ti, Zr, Silicide or nitride formation Due to reaction good wetting Ti is used as active element in
metals Hf depending on temperature and [468, 469, 499] many brazes [468, 469, 499, 500]
nitrogen pressure [468, 469; 498]
6b-group Cr CrN or silicides are formed [501]
metals
Mo In equilibrium with MoSi2 or Mo5Si3
at 1000 to 2000 °C depending
on nitrogen pressure [502]
123
124
Table 13. (Contd)
Metal Chemical interaction Wetting angle Remarks
W W is in equilibrium with Si3N4

at 1800 °C and nitrogen pressure
>3 MPa , at lower N2 pressure WSi2
or W5Si3 is formed
8b-group Fe Fe reacts at temperatures >700 °C due
metals to formation of solid solution of Si, N
in Fe, at higher temperatures silicide
formation [503, 504]
Ni, Co Ni and Co similar behavior as
Fe [504, 505]
Pt, Pd PtSix formation is found >1100 °C Pd is used in high temperature
in inert atmospheres brazes
9b-group Cu, Stable in absence of oxygen [474] Poor wetting [474] Cu: 150° at Are used in brazes as non-reactive
metals Ag, 1100 °C; Ag : 155° at 985 °C [484] components
Au
CuO, and PbO react at temperatures ³600±700 °C quite strongly with the grain
boundary phase and accelerate the oxidation and degradation. At tempera-
tures below the transition temperature Tg of the glassy phase this interaction
can be neglected because of the low ion diffusion into the grain boundary.
The interaction with Fe is the main reason for the unfavourable wear
behaviour of Si3N4 cutting tools for machining steel. The temperature at the
cutting edge during the turning of cast iron is about 800 °C causing only
moderate wear. But Si3N4 cutting tools are unsuited for steel cutting due to the
higher temperature at the cutting edge (>1000 °C) and the resulting fast wear
[470, 471].
7.3.3
Corrosion in Liquids
Si3N4 ceramics are promising engineering materials for application under

corrosive and wear conditions [18e, 506±511]. Si3N4 ceramics are heavily
attacked by hot acids, hydrothermal conditions, and bases, as can be seen in
Fig. 37. The weight loss of pure CVD Si3N4 is much lower compared to Si3N4
ceramics. The corrosion behaviour of Si3N4 ceramics in liquids is determined
mainly by the stability of the grain boundary phase. Therefore the corrosion
resistance can be altered with the composition by orders of magnitude
(Fig. 37) [506, 510, 511].
The corrosion behaviour can divided into a few main classes (Table 14). In
most organic liquids (excluding organic acids) no corrosion was observed.
Si3N4 ceramics with Y2O3/Al2O3 additives degrade strongly above RT in
medium concentrated HCl, H2SO4 and HNO3 solutions (Figs. 38 and 39). With
increasing temperature the corrosion resistance decreases, whereas with
decreasing additive content the corrosion resistance increases.
MgO-containing materials have better corrosion resistance in HCl, H2SO4,
and HNO3 solutions than Y2O3/Al2O3 containing materials [506, 510]. The best
corrosion resistance in acids was obtained for HIPSN with no additives. Si3N4
ceramics are more stable in concentrated (>5 N) than in diluted acids [18e,
506, 507, 516]. H2SO4 and HCl do not attack intergranular ®lms as strongly as
Fig. 37. Weight loss of pure CVD Si3N4 [18e] and Si3N4 ceramics with different grain
boundary phases (measuring points j, r at 90 °C [525] ; +, d at 80 °C [506])
Table 14. Classi®cation of corrosion conditions
Conditions Corrosion Literature
Organic components (oil, Wear reducing lubricants

hydrocarbons)
Acids (HCl, H2SO4, HNO3 . . .) Main attack at grain boundaries [18e, 506±518]
corrosion resistance can be
improved signi®cantly by
tailoring the composition
Media that solves the SiO2 Dissolution of grain boundaries [18e, 513, 515, 518]
protective layers intensively and Si3N4 grains
(HF, alkaline melts, concentrated Intensive corrosion
alkaline solutions at temperatures
>100 to 150 °C); Hydrothermal
conditions at ³250 °C
Bases at medium temperatures Main attack of grain boundaries, [18e, 507, 510, 513,
<100 to 150 °C corrosion resistance can be 519]
Hydrothermal conditions at Main attack at grain boundaries, [506, 510, 520±522]
temperatures £200 °C corrosion resistance can be
the triple points (Fig. 38). Even after corrosion for 200 h there is no sign, that
these boundaries are attacked, whereas the triple points are strongly corroded.
Si3N4 ceramics with a leached-out grain boundary phase have a low strength of
about 400 MPa, similar to RBSN (Sect. 8). The different corrosion behaviour
between triple junctions and thin grain boundary ®lms may be caused by
deviations in composition; the grain boundary ®lms are richer in SiO2 [316].
The SiO2 content of the grain boundary phase is the main parameter
governing the corrosion resistance of the Si3N4 ceramics. Materials with high
SiO2 concentration in the grain boundaries are more resistant to attack by
acids; changes of more than one order of magnitude were measured [511, 517].
Additionally the corrosion is in¯uenced by the fact that corrosion takes place
only in small channels between the Si3N4 grains, which can cause a change of
the rate controlling step (diffusion or reaction controlled, formation of
protective layers) [510].
The reduction of strength after short time corrosion in acids (pit formation)
cannot be correlated with the weight loss [511], whereas after intensive
corrosion a correlation exists between thickness of the corroded layer and
strength [507]. Acid corrosion appears to in¯uence subcritical crack growth in
Si3N4 ceramics [507, 523].
The corrosion resistance of the Si3N4 ceramics in H3PO4 solutions differs
from that in H2SO4 and HNO3 because a protective phosphate layer is formed
[507, 513].
Corrosion in HF-containing solutions is much more intensive than in other
acids, because of the ability of HF to dissolve the SiO2 protective layers. HF
solutions also dissolve the Si3N4 grains.
a)
2 µm
b)
2 µm
c)
Fig. 38a±c. SEM micrographs Si3N4 ceramic with Y2O3/Al2O3 additives, a before, b and c after
corrosion in H2SO4 at 90 °C
The corrosion resistance of selected Si3N4 ceramics in NaOH is given in

Fig. 40. The attack by bases is less pronounced than by acids. The extent of
corrosion increases with increasing temperature and concentration of the
bases [524]. Also, in bases the intergranular ®lms are not attacked as strongly
as the triple points. In NaOH solutions often a linear dependence of the weight
loss on time is reported. Materials which are less stable in acids are more
stable in bases. This can be explained by the stability of the grain boundary
phase [510].
Under hydrothermal conditions ceramics with Y2O3/Al2O3 additives are
stable (Figs. 41, 42). The HIPed ceramics without additives and the
MgO-containing materials are less stable. Under hydrothermal conditions
the grain boundary phase of these ceramics is dissolved completely, leading to
a removal of Si3N4 grains from the surface. This behaviour is different from
corrosion in acids, where the grain boundary is not completely dissolved and
the corroded layer is quite strong and stable. Under corrosive conditions in
Fig. 39. Weight loss and residual strength of b Si3N4 ceramics with 3.2 (1), 5.0 (2) and 7.3
(3) vol% Y2O3/Al2O3 additives in 1 N HCl at 60 °C as function of time
acids, mainly the network modi®ers (e.g., Mg, RE, Ca, alkalines) are dissolved,
whereas under hydrothermal corrosive conditions and in strong basic
solutions the silica-containing boundary network dissolves. Therefore ceram-
ics with a high SiO2 content in the grain boundary are less stable under
hydrothermal conditions than ceramics with a low SiO2 content. Under
hydrothermal corrosion at 270 °C a signi®cant dissolution of the Si3N4 grains
takes place. In materials with Y2O3/Al2O3 additives, the dissolution rate of the
grains is higher than that of the grain boundary [510].
Investigations of the corrosion behaviour of Si3N4 ceramics in acids, bases,
and under hydrothermal conditions show that composition and amount of the
grain boundary phase govern the corrosion resistance. Materials with high
SiO2 content in the grain boundary phase are stable in acids and less stable in
bases and under hydrothermal conditions. Materials with a low SiO2 content in
Fig. 40. Weight loss and residual strength of b Si3N4 ceramics with 3.2 (1), 5.0 (2) and 7.3
(3) vol% Y2O3/Al2O3 additives in 1 N NaOH at 60 °C as a function of corrosion time
the grain boundary phase are stable in bases and under hydrothermal
conditions but less stable in acids.
7.4
Colours
Si3N4 by itself is colourless due to the big gap between the valence and the
conduction band (3.5±5.5 eV) [526]. But normally Si3N4 ceramics are more or
less grey, because of the in¯uence of additions, impurities and sintering
conditions.
With increasing silicide-forming impurities (d-elements, excluding ele-
ments of the third and fourth group of the periodic table of elements) the
colour becomes darker. Addition of Ti-containing compounds leads to the
formation of TiN. Low concentrations of TiN produce a dark colour, higher
concentrations a typical brown to golden colour by the red-brown TiN(1±x)Cx.
Fig. 41. Weight loss of Si3N4 ceramics with different additives under hydrothermal
conditions at 210 °C as function of corrosion time
Under normal sintering conditions Hf or Zr additions form HfO2 or ZrO2

(stabilised by REs or MgO) which are bright. Strong reducing conditions
generate yellow ZrN(1±x)Cx or read brown HfN(1±x)Cx. There are some
indications that Cl) impurities intensify the dark coloration. Impurities of
free C lead to a black [232, 272] and SiC additions to a green colour.
Additionally, RE ions have colour effects. For example, Yb may colour the
material brown or green depending on the oxidation state. The coloration of
ass by different REs varies, e.g., violet for Nd [527].
Transparent Si3N4 ceramics were produced by compaction of amorphous
Si3N4 under 5 GPa pressure [528]. Translucent ass can be produced by
minimisation of the grain boundary phase [335].
The possibilities of colouring described above can be altered by the
sintering conditions. Without additional colour centres, Si3N4 ceramics appear
between almost white to black.
The intensity of grey colouration corresponds to the free Si precipitates of
several nm to several lm which act as colour centres [232, 272, 529].
Impurities of transition metals can be nuclei for precipitation of elemental Si
[232, 268, 272], which is most likely formed by the following reaction sequence
[232, 271, 272]:
3 SiO2 Si3 N4 , 6 SiO 2 N2 17
6 SiO , 3 Si 3 SiO2 18
SiO exists not only in the gaseous phase but also dissolved in the oxide
nitride liquid. The solubility is quite high, because its structure is very similar
Fig. 42. Weight loss and residual strength of Si3N4 ceramics under hydrothermal conditions
at different temperatures (corrosion time 200 h)
to that of the oxide nitride melt. Alternatively, this process can be described as
a simple formation of oxygen vacancies in the oxide nitride melt or as a simple
reduction of silicon from a valence of four to two. Due to the decreasing
stability of the divalent Si during cooling from sintering temperatures, the SiO
will decompose during cooling according to Eq. (18).
Free silicon is kinetically stable due to the low diffusion rate of nitrogen in
the dense ceramic, whereas in ceramics with open porosity, the diffusion path
is short enough for the reaction of nitrogen with Si or SiO to Si3N4. This agrees
with experiments showing that sintered Si3N4 ceramics with an open porosity
are lighter in colour than those with closed porosity [232, 272].
The grey colour often is not homogeneous and features a bright surface
layer followed by a dark layer and a grey coloured bulk material (Fig. 43).
On the basis of experimental data and thermodynamic considerations it can
be concluded that the colour of the bulk is less dependent on the weight loss,
but strongly depends on the temperature and the nitrogen pressure. With
increasing nitrogen pressure during sintering at constant temperature or with
decreasing temperature at constant nitrogen pressure the samples become
Fig. 43. Optical micrograph of a cross-section through a GSSN rod (Additives : 6 wt% Y2O3,
4 wt% Al2O3; sintered at 1800 °C; 6 bar N2; 4 h)
brighter [232, 272]. The dark coloration of the near-surface area depends on
the weight loss and the oxygen partial pressure in the sintering atmosphere
and is caused by an enhanced reduction of Si4+ to Si2+ during sintering. The
outer bright near surface area results from a decolouration processes during
the high pressure step (isothermal sintering), or during cooling from the
sintering temperature. The cooling regime will in¯uence signi®cantly the
thickness and the colour of the outer layer [232, 272].
Sometimes Si3N4 ceramics exhibit the so called ``snow ¯akes'' with
dimensions up to several millimetres (Fig. 44). They appear bright in optical
dark ®eld images. Such features are observed in dense HPSN, SSN, GPSN, and
SRBSN, independent of the production method. This phenomenon can be
explained by local changes in the refractive constants of the grain boundary
phase causing a local enhanced light scattering [530]. However, the reason for
the ``snow ¯ake'' formation remains unclear. Several possible mechanisms are
discussed: formation of microvoids [531], locally different crystallisation [530,
532] and microcracks caused by volume change during crystallisation of the
Fig. 44. Dark ®eld image of an silicon nitride material with ``snow ¯akes''
grain boundary phase [533]. For some reason this ``snow ¯ake'' formation is
connected with the crystallisation of the grain boundary phase; all steps
leading to a stabilisation of the glass phase, e.g., higher SiO2, Al2O3, MgO
contents, faster cooling, reduced interaction with the atmosphere cause a
suppression of crystallisation and ``snow ¯akes''.
Fe impurities may cause ``snow ¯akes'', because of the nucleation and
crystallisation of YAG (Y3Al5O12) [532]. The volume change during crystal-
lisation of the grain boundary phase leads to internal stresses which can be
cause micropores, or microcracks, or relax by other mechanisms [534, 535].
Such microcracks have only been detected in ceramics with crystallised
b-Y2Si2O7 as grain boundary phase [533].
The ``snow ¯akes'' can be reduced by heat treatment slightly above the
eutectic temperature, even under slightly oxidising conditions [530]. A
controlled crystallisation of the grain boundary leads to a more homogenous
microstructure. ``Snow ¯akes'' are inhomogeneous and cause locally changing
microhardness and polishing behaviour [530]. This is especially disadvanta-
geous when a high local mechanical load is applied, as it is the case for ball
bearings because these inhomogeneities are potential starting points for
pitting damage.
Colouring and ``snow ¯akes'' can be controlled by reducing the interaction
of Si3N4 during sintering, and optimising the temperature, time and pressure
regime (Sect. 5.3.3).
8
Reaction Bonded Silicon Nitride (RBSN)
The reaction bonded silicon nitride (RBSN) is the ®rst silicon nitride based
material, already produced in the early 50s of the last century [536]. The
reaction bonding of Si powder with N2 results in a material with 12±30%
porosity. During nitridation no shrinkage occurs, permitting the production of
complex shaped components without expensive ®nishing. These possibilities
and the resulting low cost were the main reasons for an increasing attention
and its application as refractory material, especially for thermocouple sheaths
and in the metallurgy of aluminium alloys [537]. With reduction in the cost of
Si3N4-powders and improvements in their densi®cation technology, RBSN
ceramics are increasingly being replaced by sintered Si3N4 ceramics, which
have higher strength, hardness, and long-term stability (Sects. 7 and 10).
By a post sintering process including additional sintering aids RBSN can
densi®ed to porosities of about 0±5% [538±540] This material, SRBSN, is
applied for different wear parts [540]. A comparison of the production and
material properties is given in Table 10.
8.1
Production and Nitridation Process
The production of RBSN starts from Si powder (mean grain size 3±40 lm),
which is shaped to parts and then nitrided [541]. The nitridation of the shaped
body is the same process as for the synthesis of Si3N4 powders by direct
nitridation (Sect. 4.1). During nitridation different reactions take place
simultaneously (nitridation, sintering of Si, oxidation by oxide impurities of
the atmosphere, and evaporation of the adhered SiO2 layer) in¯uencing the
rate of conversion, microstructure and properties [440, 541±546].
Since the reaction is strongly exothermic, local melting of Si might be
caused, leading to microstructural defects and reduced strength. Therefore low
reaction rates and nitridation times of 24 to 100 h are recommended.
Especially in the initial stages a precise control by low heating rates or nitrogen
pressures is required.
In the beginning of nitridation a bridging network of Si is built, ®xing the
overall dimensions of the nitrided part, i.e., no classical sintering process with
dimensional changes takes place. Nevertheless the pore volume decreases
during nitridation, because of an increase of the solid phase volume of 21.2%
by forming of Si3N4. The reduction of the pore volume and pore radius causes
a reduced transport of nitrogen through pores resulting in a decreased
reaction rate with increasing conversion of Si.
Depending on the time/temperature regime, the a/b ratio in the Si3N4
compacts varies as a result of varying mechanisms of formation of a- and b-
Si3N4. a-Si3N4 is as ®ne-grained needles or whiskers in the pores formed by a
CVD process of gaseous Si with molecular N2 [544±546]. Additionally a
vapour-liquid-solid mechanism was found, i.e., impurities (e.g., Fe) form
a liquid from which the a-Si3N4 grows. The whiskers are formed during
nitridation of SiO2 surface layer [440]. All these mechanisms are connected
with the evaporation of Si, and all factors that reduce the evaporation also
reduce the reaction rate [544]. Therefore the removal of the SiO2 prior
nitridation accelerates the reaction, leading to a coarser microstructure [440,
544]. SiO2 signi®cantly in¯uences growth rate and density of Si3N4 nucleation
[542, 544]. Intensive nitridation takes place only when the SiO2 surface layer is
locally removed [440, 541, 542]. Si3N4 formed on the surface also reduces the
evaporation rate of Si. As a consequence, the nitridation cannot be completed
by isothermal nitridation at low temperatures when intensive nucleation of
Si3N4 occurs.
The formation of b-Si3N4 takes place on the surface of solid or liquid Si. The
growth of b-Si3N4 is epitaxial on the surface of Si [546]. The relatively high
diffusion coef®cient in the direction of the c-axes of the b crystals promotes
growth in this direction. The growth rate is determined by kinetic processes on
the interface between Si and Si3N4. In liquid Si a fast growth of large b-Si3N4
needles is observed. Liquid Si results in an increased b content due to a high
nitridation temperature or impurities forming low melting silicides (e.g., Fe
forms an eutectic at 1212 °C) [543, 546]. Iron impurities can also accelerate the
formation of a-Si3N4. Therefore the resulting a/b phase ratio depends not only
on the time-temperature regime but also on the amount and distribution of Fe
impurities.
Hydrogen or ammonia in the reaction atmosphere increase the a content,
the ®neness of microstructure, and the strength, they reduce the dependence
of the nitridation process on variations in green density and powder properties
[440, 544]. Both gases accelerate the mass and heat transport through the pores
and the local removal of the SiO2 surface layer (it follows from thermodynamic
calculations that only in H2 or NH3 containing atmospheres can SiO2 be
nitrided [440]). Factors in¯uencing reaction and microstructure are summar-
ised in Table 15. It is obvious that the nitridation is very complex and can
modelled only partially at present [547, 548].
8.2
Microstructure and Properties
The development of the microstructure is strongly affected by the perfection of

the Si green body. Inhomogeneities in density, pore size and distribution are
Table 15. In¯uence of processing variables during nitridation on the conversion rate,
microstructure and properties of RBSN [541±546, 549, 550]
Processing variable Product characterisation
Low temperature <1300 °C High a/b ratio, ®ne microstructure, normally low
conversion rates
Medium temperature 1300±1412 °C Coarser microstructure with increasing
temperature, pose size increases
High temperature >1412 °C b formation pronounced, coarse structure,
low strength and fracture toughness (KIC)
Low grain size of Si Faster nitridation, higher a content,
higher strength and KIC
Big grains of Si (>40 lm) No complete nitridation, reduced strength
Low green density <50% th. density Faster nitridation (especially of bigger parts),
lower strength and KIC
High green density >65±70% Dif®cult to achieve full nitridation, strength
reduces if inhomogeneous nitridation
takes place
No SiO2 surface layer Fast nitridation, coarse microstructure,
high a content
Thin SiO2 surface layer Finer microstructure, high a content, high strength
(<2 á 10)3 g/m2) and KIC
Thick SiO2 surface layer Dif®cult to control the nitridation, lower strength
and KIC
Fluor addition Accelerated nitridation, high a content, high
strength
Fe impurities (®ne dispersed, low Accelerated nitridation, high a content,
content) low temperature high strength
Fe impurities (inhomogeneous) Defect formation due to silicide formation,
reduced strength
Alkaline, alkaline earth oxides, Accelerated nitridation, high a content, high
Y2O3, Sc2O3, RE2O3 strength, reduced high temperature strength,
subsequent sintering at high temperature
is possible
Al2O3 Accelerated nitridation, increased b-content
Hydrogen or ammonia in the More homogenous reaction. High a-content,
nitridation atmosphere ®ne microstructure, high strength and KIC
Flowing N2 atmosphere Coarse microstructure, low strength and KIC
unfavourable. The microstructural features normally are open pores (in the
range of 12±30%), a/b-Si3N4 grains and residual Si; their size, amount and
distribution signi®cantly in¯uence the properties, especially strength. Addi-
tionally, the strength may be in¯uenced by defects caused by local melting of
the Si or silicides and by large inclusions of non-reacted Si. Whereas the defect
size is mainly determined by the green body structure, the fracture toughness
is strongly in¯uenced by the nitridation process. A perfect nitridation results
in a very ®ne overall porosity (mean pore channel radius of several 10 nm).
This is caused by the growth of ®ne grained mainly a-Si3N4 into the pores. The
strong interconnection between these ®ne a-Si3N4 grains is responsible for
fracture toughness values of about 3 MPa m1/2 and strength values of up to
350 MPa. A coarser microstructure leads to a lower fracture toughness and
reduced strength.
In inert atmospheres the mechanical properties of RBSN are constant up to
1200±1400 °C because of the absence of a glassy grain boundary phase, which is
also the reason for the excellent thermal shock and creep behaviour. The thermal
shock resistance, hardness and elastic constants depend on the microstructural
parameters but are much lower than for dense Si3N4 ceramics [539].
In air, the mechanical properties are in¯uenced by the oxidation processes
[543]. In materials with a ®ne overall porosity the oxidation at > 1100 °C
closes the pores with the help of an SiO2 surface layer. This layer protects the
material from further oxidation and heals surface defects. This and the
formation of compressive stresses due to the different thermal expansion
coef®cients between SiO2 and RBSN are the reasons for strength increase after
oxidation. Materials with a high amount of macropores (>1 lm) oxidise not
only at the surface but also inside the volume due to longer closing times of the
surface pores. In consequence these oxidation mechanisms result in more
intensive oxidation at low temperatures £ 1100 °C, due to the slow rate of pore
closure and higher internal oxidation.
8.3
Sintered Reaction Bonded Si3N4 (SRBSN)
An remarkable improvement of RBSN is made by including additives in the

starting silicon powder and post sintering between 1700±2000 °C after
nitridation. During post sintering porosity is reduced and strength increased.
The properties of the SRBSN are very similar to SSN or GPSN materials
(Table 10) [540, 541, 551]. The additives are the same as for SSN (Sect. 5.1.2).
The densi®cation can be attained by pressureless sintering, gas pressure
sintering, or sinter-HIPing (Table 10) [538, 552]. The microstructure of the
starting RBSN causes some differences compared to sintering of Si3N4
powders. RBSN has a higher green density, but the very strong skeleton of
interconnected a needles, as well as the formation of oxide nitride grain
boundary phases, noticeably delay the densi®cation due to slow rearrangement
mechanisms [552]. SRBSN has the advantage of lower shrinkage and better
green machinability. The microstructure formation is very similar to that
during sintering of Si3N4 powders.
The strength of the SRBSN is usually between 600±900 MPa and comparable
to that of common GPSN. The highest observed values (980 MPa HIP-SRBSN
[538]) are much lower than the highest observed values for GPSN materials
(1400±1600 MPa). It is an open question whether or not this difference is
caused by the lower level of optimisation or by the differences in the sintering
behaviour. In general, the properties of SRBSN are very similar to sintered
qualities, which can achieved with low cost Si3N4 powders. Therefore the
former cost advantage of the SRBSN is being lost with fewer applications as a
consequence.
9
Composites
To improve, mechanical or electrical properties of Si3N4 ceramics different
types of composite materials have been developed (Table 16). Among them
Si3N4/TiN and Si3N4/SiC are the most extensively investigated.
Special composites are Si3N4 ceramics with additions of b-Si3N4 whiskers,
or large b-Si3N4 seeds are developed to generate b-Si3N4 ceramics with
bimodal microstructures and improved toughness and strength. They contain
two different fractions of Si3N4 grains and may considered to be Si3N4/Si3N4
composites.
9.1
Si3N4/TiCN Composites
The interest in composites containing TiN or TiCxN1±x (x < 1; TiN and TiC
form a continuous solid solution with a NaCl structure type) is based on
advantageous properties of TiCN such as high hardness, electrical conductivity
and chemical compatibility with Si3N4. The Si3N4/TiN mixtures are stable in a
wide range of temperatures and nitrogen pressures [565]; they have good
sinterability without decomposition. Details about the solid solubility of Si in
TiN of the Si3N4/TiN composites without additives differ considerably: up to
2.3 wt% [566] and 10.7 wt% Si [567]. No solubility of Si in TiN was observed
in Si3N4/TiN composites produced by hotpressing of Ti- and Si-containing
organometallic precursors. Although a clear decision between this contradic-
tory data is not possible, a low solubility of Si in TiN is more likely. With
respect to the solubility of TiN in b-Si3N4 an analogous situation exists. From
EDX investigations, and con®rmed by a slight change of the lattice parameter,
a solubility of 0.6 wt% of TiN in b-Si3N4 has been determined [567]. But this is
not in agreement with the ®ndings of submicrometer-sized TiN precipitates in
composites with only 0.2 wt% TiN prepared by liquid phase densi®cation,
indicating a solubility of TiN in b-Si3N4 below 0.2 wt% [525].
TiC is thermodynamically not compatible with Si3N4 under sintering
conditions. TiC reacts with Si3N4 forming TiCxN1±x where the x values are in
the range of 0.3 [568, 569] to 0.24 [567]. The x value in TiCxN1±x depends on
nitrogen pressure and temperature. As an additional reaction product, SiC
may be formed or the oxide nitride liquid phase can be reduced by carbon
138
Table 16. Overview of Si3N4 ceramic composites
Composite Improved properties Remarks Literature
Si3N4/BN Wear behaviour and machine stability, BN retard the sintering, materials with higher [553±556]
resistance against metals, high thermal shock BN content are dif®cult to sinter
resistance. Hot pressed nanocomposite with [366]
high strength and improved machinability
Si3N4/TiN/TiC High electrical conductivity, wear behaviour Sect. 9.1
Si3N4/SiC High mechanical properties at RT and HT Sect. 9.2
Si3N4/Si3N4 High fracture toughness and strength, thermal Sect. 9.3
conductivity
Si3N4/TaN Electrical conductivity TaN react under sintering conditions forming [557, 558]
silicides and mixed nitrides
Si3N4/ZrO2/HfO2 Only moderate increase in toughness due to Under strong reducing conditions ZrN1)xCx [559, 560]
transformation toughening; reduced or HfN1)xCx can be formed leading to
hardness and oxidation stability with instability
higher ZrO2 content.
Si3N4/TiB2 Only moderate improvement of KIC or Thermodynamically not stable under sintering [555, 559]
cutting behaviour conditions; decomposition of TiB2 in TiN and
BN or TiSix
Si3N4/MoSi2 Signi®cant improved oxidation resistance Depending on temperature and nitrogen [410, 561, 562]
and life time at high temperatures, pressure Mo3Si5 can be formed during
electrical conductive sintering
Si3N4/W W wire in Si3N4 for heating elements Depending on nitrogen pressure and [563]
temperature WSi2 can be formed leading
to a destruction of the wire
Si3N4/other silicides Less important due to less stability against Under sintering conditions liquid silicides can [564]
oxidation or low melting point concentrate in big defects, Fe silicide can
improve the wear behaviour in engine
applications
bss/Al2O3/carbon Increased stability against metals used for Al2O3 particles partially dissolve during [555]
refractories sintering, materials usually not dense
[568, 569]. Therefore the TiC-containing materials are more dif®cult to sinter
and are used to a lesser extent [568, 570].
Most of the composite materials were produced by mixing TiN or TiCN
powders with Si3N4 [559, 569, 570]. Also additions of nanosized TiN particles
to Si3N4 was tested successfully [571]. The nitridation of mixtures of Si and Ti
or TiN results in an Si3N4/TiN particulate composite [572].
To describe the formation of composites with low TiN content a simple
exchange reaction can be used:
3 TiO2 Si3 N4 , 3 TiN 3 SiO2 1=2 N2 19
The result is a composite with a very ®ne homogeneous distribution of TiN

of nanometer size in the Si3N4 matrix. Higher starting amounts of TiO2 result
in Si2N2O formation [525, 573]. The composites can also be prepared from
metalorganic precursors [574].
The densi®cation of TiN containing composites begins at lower tempera-
tures than the densi®cation of the pure Si3N4 matrix. The reason is that the
solution of TiO2, which adheres to the surface of TiN powder particles, reduces
the viscosity of the liquid phase. At temperatures >1350 °C TiN precipitates
from the liquid according to Eq. (19) [568, 570].
The thermal expansion coef®cient of TiN and TiCxN1±x is higher than that
of Si3N4. Therefore around the TiCN grains, local tangential compressive and
radial tensile stresses are formed. These stresses change the crack path and
increase the fracture toughness. Systematic investigations of the dependence of
the fracture toughness on TiN grain size and content has shown that with
increasing particle size and volume contents up to 30 vol%, the fracture
toughness increases slightly without strength increase [568, 570, 575]. This is
connected with the intrinsic high fracture toughness level of the Si3N4 matrix
[559, 570, 575]. Nanosized TiN particles in the composites give no improve-
ment of the properties compared to composites with TiN particles in the lm
range [571].
Composites with more than 30 vol% of TiN have a high electrical
conductivity, which is very useful for the ®nal shaping by electrical discharge
methods or other applications where the electrical conductivity is involved
[559, 576, 577]. The speci®c resistivity depends on the grain size of the TiN
particles and changes drastically between 20±30 vol%, expressed by a drop
from high values (1013 Wcm)1) to low ones (10)1±10)3 Wcm)1). This drastic
drop shifts with decreasing particle size to lower volume contents of TiN [577].
The formation of duplex microstructures, consisting of TiN free areas of up to
200 lm surrounded by TiN, results in materials with high electrical conduc-
tivity down to low TiN contents (9.8 vol%) [578].
The oxidation resistance of TiN and TiCN is lower and therefore the
composites are less stable at high temperatures than the monolithic materials.
When the TiN particles do not form a continuous skeleton, then only surface
TiN particles oxidise rapidly. The further oxidation is determined by the
stability of the matrix. Composites with a TiN skeleton oxidise more readily
than the Si3N4 matrix [559, 570, 579].
The composites have an improved wear behaviour under load [580±582]. In

comparison to monolithic Si3N4 ceramics, the differences at 800 °C were more
pronounced than at room temperature. This was attributed to the formation of
TiOx layers, which are known as solid lubricants [580, 581].
Also, materials with small amounts of TiN additions show lower friction in
ball bearing applications leading to higher loading capacities of the bearings
[525, 583].
Si3N4/TiCN composites were tested for cutting tool applications under
different conditions with little success [559, 570].
Layered ceramic Si3N4 composites on the basis of TiN or TiCN have been
prepared by tape casting and hot pressing and showed highly anisotropic
electrical and mechanical characteristics. One idea behind this development is
the detection of crack formation under loading conditions by electrical
conductivity measures [584, 585].
9.2
Si3N4/SiC Composites
Already in the early 70s of the last century the ®rst investigations of Si3N4/
SiC composites were carried out [586]. Unfortunately the strong interaction
of SiC with the liquid phase during sintering causes deterioration of the
densi®cation behaviour [275, 586±588] and the interest in this material
declined. In the late 1980s new investigations in connection with SiC
whisker reinforcement began. But the increase in fracture toughness was
not substantially greater than that of the in-situ reinforced monolithic
Si3N4 ceramics [589±593]. Additionally, the carcinogenity of the SiC-
whiskers was a strong argument for discontinuing the research activities
[594]. The incorporation of SiC platelets instead of whiskers yields only a
moderate improvement in the properties [595, 596]. With the beginning of
the 1990s the interest in Si3N4/SiC particulate composites was rekindled in
connection with the nanocomposite concept [223, 587, 597, 598]. The
improved quality of the powders and advanced technology enabled
development of materials with high strength at room and high tempera-
tures [223, 597]. These promising results have intensi®ed the investigation
of powders, their preparation and the properties of the resulting
composites [213, 587, 598].
The recently developed Si3N4/SiC composites are based on
± amorphous or crystalline composite powders [599±603],
± mixtures of Si3N4 and SiC powders [587, 604, 605],
± in-situ synthesis of SiC during sintering [606, 607],
± polymer derived routes [213, 608±611].
Most of the nanocomposite materials were produced by hot pressing [223,
275, 587, 597±599] to overcome the dif®culties during densi®cation. For gas
pressure sintering and HIPing, the decomposition reaction of SiC under high
nitrogen pressure and the interaction of SiC or residual free carbon with the
oxide nitride liquid during sintering must be taken into account [275].
However, gas pressure sintered high strength composites were also developed
[275, 604, 612, 613].
The SiC particle distribution is intergranular and intragranular. The ratio of
inter- and intragranular particles depends on the SiC grain size and the growth
conditions of the Si3N4 grains [223, 275, 587, 597, 612, 614, 615]. There are SiC
particles with no glassy phase between Si3N4 and SiC grains and others which
are surrounded by a glassy phase [615, 616]. In materials with an SiC content
up to 30 vol%, size and shape of the Si3N4 grains are similar to those in
monolithic material, and the SiC particle size distribution is broad between
some nm up to 1 lm. The SiC particles reduce the exaggerated grain growth by
pinning (Zener mechanism), but the overall grain size changes only slightly.
The pinning effectivity increases with decreasing grain size and increasing
volume content of SiC. In composites with high SiC content (>50 vol%)
produced from amorphous Si3N4/SiC powders the SiC and Si3N4 particles are
nanosized [223, 597, 599]; these materials are superplastic [617].
Ceramics with ®ne SiC and Si3N4 grains can also directly produced from
amorphous precursors with no sintering additives, however they are not dense
after crystallisation and therefore have relatively low strength and fracture
toughness [213, 609±611].
The strength of Si3N4/SiC ceramics increases compared to the monolithic
state especially with Y2O3/Al2O3 additives [223, 597, 600, 606, 611, 612]. The
positive effect on the strength is connected with the pinning mechanism,
reducing the accelerated growth of big grains which are strength limiting
defects in high strength materials [587, 605, 612]
The wide scatter of the maximum values of fracture toughness [587, 598]
or sometimes even a reduced fracture toughness by adding SiC [587, 598, 599,
612] indicate that changes of powder characteristics, processing variables, or
of the composition of the grain boundary phase are of more in¯uence than
special mechanisms connected with the nanoparticles [587, 604]. The
increased residual stresses [604], caused by the different thermal expansion
coef®cients of the different phases, might not be the reason for the changes in
fracture toughness and strength [587, 604] as assumed earlier [606].
The improvements of high temperature strength and creep resistance are
mostly connected with shift of composition of the grain boundary phase
caused by the interaction of SiC or residual free carbon with SiO2 during
sintering [587, 604, 612, 618]. For instance the high temperature strength of
Si3N4 ceramics is substantially affected by the crystallisation of the grain
boundary phases (Sect. 7.2.2), which is in¯uenced by the SiC addition (shift of
the composition of the grain boundary phase) [587, 619]. Materials with
constant compositions of the grain boundary phase have very similar high
temperature strength [605, 618, 620]. In addition to the change of composition
of the grain boundary phase, the SiC particles in the grain boundary have a
positive in¯uence on creep. SiC particles without a glassy phase between SiC
and Si3N4 particles form rigid skeletons, reducing the grain boundary sliding
[587, 612, 616]. In HIPed materials with SiC inclusions but without other
additives the creep rate decreases only by a factor of 2 [621]. This change
seems to be connected with a true strengthening effect by the SiC particles.
Higher differences between the creep rate of the composites and the
monolithic materials might be caused by changes in the grain boundary
phase, impurities or grain size [429, 460, 622, 623]. Only slight altering of
existing creep mechanisms known from pure materials were found [604, 605,
621].
In Si3N4 ceramics with SiC additions the residual strength after oxidation
increases because of a change in the oxidation mechanism [429, 460, 605, 624].
Just below the outer oxide layer a protective layer of Si2N2O is formed which
reduces the damage in the bulk (reduced migration of additives toward
surface, and pore formation) and the size of pits created by oxidation [429,
460, 605]. Similar behaviour was found in Si3N4/MoSi2 composites [429, 620].
This changed oxidation mechanism is also the reason for higher life times,
under load, of such materials at high temperatures [460, 605]. No differences
in the wear resistance between Si3N4 and Si3N4/SiC composites was found
[625]. Si3N4/SiC nanocomposites with SiC contents>25 vol% have reduced
electrical resistivity (<107 Wcm) [610, 611, 626].
Summarising the data about the Si3N4/SiC composites, it can be stated that
no signi®cant increase in the mechanical properties at room temperature has
been achieved which can not be realised in monolithic Si3N4 ceramics.
However, at high temperatures substantially improved long term behaviour
connected with a change in the oxidation mechanism can be realised.
9.3
Si3N4/Si3N4 Composites (Seeded Materials)
High fracture toughness of Si3N4 ceramics is obtained by activation of

toughening mechanisms such as crack-de¯ection or bridging via interfacial
debonding at large elongated grains [33, 310, 377]; but large elongated grains
can also act as strength limiting defects (Sect. 7.2). To overcome this
discrepancy, ceramics with large elongated grains of narrow size distribution
embedded in a ®ne grained matrix with oriented rod-like grains are desirable.
Such ceramics can be realised by large elongated b-Si3N4 crystallites (usual
length 2±10 lm; thickness 0.5±1.3 lm; aspect ratio 4±5) as seeds [299, 377,
627±635]. Such seeds usually are prepared by sintering of Si3N4 powders with
Y2O3/SiO2 additives at 1850 °C, subsequent acid treatment to remove the
additives and classi®cation by size [629, 633, 635]. The seeds (whisker-like
large b nuclei) grow anisotropically consuming the ®ne grains of the matrix as
described in Sect. 6.1. The amount of the seeds should be moderate to avoid
impingement of grains. The best results were observed with amounts between
2 and 5 vol% [299, 377]; larger amounts (>10 vol%) cause a fast consumption
of the matrix grains by the growing seeds. Exaggerated grain growth results in
a coarse grained microstructure with a less pronounced anisotropic shape
[299]. With tape casting [377, 629, 632, ] or extrusion [631] the seeds can be
strongly oriented in the starting materials as a precondition for an optimised
anisotropic order. The degree of order of the seeds is the higher, the smaller
the amount of seeds [629]. For instance, tape casting of seeded Si3N4 powders
leads to a high degree of orientation of the seeds (>60%) and hence to
anisotropic fracture toughness and strength (5 MPa m1/2 parallel to the grain
alignment and 12 MPa m1/2 normal to it). In tape cast materials a simulta-
neous increase of strength (from 1100 to 1400 MPa) and fracture toughness
from 7.1 MPa m1/2 to 12.5 MPa m1/2 by addition of 5 vol% seeds was observed
[377]. Such ceramics with oriented grains also have high thermal conductivity,
up to 160 W(mK))1.
Seeding can be used also in reaction bonded materials (Sect. 8). Increased
fracture toughness and strength were observed in reaction bonded and post
sintered materials [633].
An increase of the fracture toughness from 2 to 6 MPa m1/2 was observed in
hot pressed Si3N4 ceramics with 50 vol% residual a-Si3N4 by adding 5 vol%
seeds [635].
10
Applications
During the last 40 years Si3N4 ceramics have developed from exotic to
commercial materials with increasing application. At present, cutting tools are
the most important market; followed by engines components, ball bearings,
metal forming and processing devices, and gas turbines. An overview on the
applications of Si3N4 ceramics is given in Table 17. A rough idea of the market
and the share of the different applications can be derived from the powder
consumption. One third of the total production of Si3N4 powders (300±350 t in
1998) was used for cutting tools, 25% for engine components, 25% for metal
processing and wear parts and 2% for ball bearings. The remaining 10% are
used in research [636].
Si3N4 ceramics cover only 1% of the total market of advanced ceramic
materials, i.e., electronic and structural applications, but about 5% of the
structural ceramics. They have the highest growth rates among structural
ceramics [636±638].
The use of Si3N4 ceramic components in engines is now seen more
realistically than in the past. The adiabatic full ceramic engine, as it was
proposed, has not been realised, but there are ongoing efforts to improve the
design of this engine [648, 654]. Also, the production of more than 300,000
turbochargers per year in the beginning 1990s in Japan did not result to series
introduction in passenger cars [430, 654]. But the development of components
for different engines are still under way. Injector links, check balls, brake pads
and fuel pump rollers are used in diesel engines for trucks and have been
produced in several thousand pieces per month for years [646, 647]. These
applications are not really at high temperatures, but very bene®cial due to
reduced wear and weight. Also in a newly developed high pressure common-
rail injection pump system, valve plates made of Si3N4 ceramics are used to
reduce wear [402].
The production costs are a main problem in the application of Si3N4
ceramics. The production of small series is more expensive compared to
existing materials. This is connected on the one hand with the high costs of
raw material and on the other hand with the production technologies. In
144
Table 17. Applications of Si3N4 ceramics
Area of Application State of the market Literature

Application
Cutting Cutting tools for turning and milling of cast Stable market; approx. 50 million $/year; [471], [637], [639]
iron; ass/bss for turning and milling of moderate growth
Ni- based superalloys; Granulation of
polymers, cutting of textiles and ®bres
Bearings and sealing Ball bearings especially for high speed Fast growing market for hybrid and ceramic [583], [637], [640], [641]
high stiffness applications; No or low bearings 10 times increase in hybrid
lubrication (vacuum pumps, dental drill bearings between 1995±1999. Increasing
handpieces); Under corrosive conditions numbers of balls reduces the price.
(food ndustry; chemistry; metallurgy); Development of other rolling bodies is
high and low temperatures in the beginning
Metallurgy Different parts in Al metallurgy (components Growing market, RBSN is more and more [636, 637], [640]
of pumps working in different metals Al, replaced by low cost SSN
Pb); locating pins and other parts for
welding operations
RBSN refractory materials 20 t/year [14]
Springs, jigs for brazing of electronic Small market [637]
components, jigs for glass sealing,
Metal forming bss Rollers for cold rolling operations Small market [430]
Punches, dies, plugs, mandrels (Fe; Al; Cu) Small, but fast growing market [642], [643]
Automotive Valve plate in common rail systems Several million pieces/year [402], [637]
Glow plugs for diesel engines and Increasing market
additional car heating
Fuel injector parts, valves, Used in trucks. Components in [645±649]
valve train components, rocker arm pads, test cars and limited series
tappet disks, cam follower roller,
turbocharger rotors; tappet shim
precombustion chambers
Aircraft engines Ceramic turbine nozzles in the APU More than 65,000 h and 5500 starts in ®eld [650, 651]

(Auxiliary Power Unit Engine). tests.
Seals runners installed in different business Growing market: better speci®cation
jets such as Falcon, Citation and Learjet. and standardisation needed.
Ceramic oil pump spacer enhance engine
could start capability on APUs on Boeing
777 and 737 and all Airbuses.
Cutter pins, which are used as a safety More than 30,000 cutter pins are installed
feature for air turbine starters. Ceramic
wear indicator on Airbus APU starter
brushes since 1997
Space technology Turbo pump of the space shuttle; radar Small market [644]
windows for rockets
Electronic industry Thin ®lms for insulating barriers or masks Increasing market [14]
for etching processes; XY machine tables
for VLSI semiconductor manufactures;
vacuum jigs
Chemistry/oil industry Different chemical machinery parts as valves Growing market, [642]
sealing,
Micro®lter; high strength, chemical resistant No commercial application at present [652, 653]
porous materials
Wear parts (mill liners, milling balls) Small market [642]
Gas turbine 300 kW output power turbines Test runs [654, 655]
Household Cooking plates with integrated heater for Introduction in the market at present [656]
electrical cooking for kitchens and camping
(low voltage systems)
145
recent years, substantial improvements in the production technology have

been made and it was shown that mass production can lower the cost
substantially [644]. The improvement of properties and production technology
result in an increasing application of Si3N4 ceramic components in engines.
They can be produced with high reliability, as demonstrated in extensive ®eld
tests of valves carried out in several thousand 1600 C200-ML cars by Daimler-
Chrysler in more than several million miles [644, 649]. The higher production
costs of the ceramic valves compared to steel valves and also changes in engine
design are reasons that mass production has not been achieved to date. There
still exist test cars of low weight and high ef®ciency loaded with Si3N4 ceramic
valves [644, 649].
The high reliability of Si3N4 ceramics is demonstrated by hybrid bearings in
the main engine pump of the space shuttle [644] and different components in
the aircraft's auxiliary power units [650, 651].
The application of Si3N4 ceramics as components for gas turbines is under
development. First successful tests of those components have been carried out
with 1350 °C turbine inlet temperatures. The further improvement of the
turbine ef®ciency requires higher turbine inlet temperatures; for this,
additional coatings are necessary (Sect. 7.3).
Si3N4 thin ®lms and coatings are ®nding increasing application in
electronics as electrical diffusional barrier or as masks for etching processes.
Si3N4 coatings produced by plasma-enhanced chemical vapour deposition are
increasingly used in silicon photovoltaics as passivating antire¯ection coating
[657].
Low cost Si3N4 powders and advanced processing technologies facilitates
the production of materials with 800 MPa bending strength for cooking plates
with integrated heater. These plates are the central component of an effective
cooking system allowing automatic control of the cooking process [656].
An interesting application of Si3N4 powders is as a foaming agent in the
production of glass foams for thermal insulating in uses ranging from
microelectronic devices to ®re resistant non ®bre containing insulation
materials for buildings [658]. Low-grade Si3N4 powder is used in the steel
industry to increase the nitrogen content of the metal; this could also be a
possible application for recycled Si3N4 ceramics.
The largest quantity of Si3N4 by weight is applied in nitride bonded silicon
carbide refractories which are produced in amounts of 20,000 t/year [14].
11
Conclusions
Si3N4 ceramics represent a whole class of different compositions with many
facets and a wide range of properties, and as a consequence they have a high
potential for speci®c applications. They are light, have good mechanical and
thermomechanical properties and they are wear and corrosion resistant.
Differences between the individual types are founded in grain morphology,
and amount and chemistry of the grain boundary phase. Not accidentally the
comparison is made to the broad variety of steels which satisfy a multitude of
technical requirements. After more than three decades of intensive research

good progress has been made in the scienti®c and practical exploration of
these sophisticated materials which characteristically combine all advantages
and all problems of advanced ceramics. This is particularly true of all the
sintered Si3N4 ceramics (SSN, GPSN, HPSN, HIP-SSN, HIP-SN) which are
suitable for many applications not attained by the reaction bonded grades
(RBSN, SRBSN, HIP-RBSN). Nowadays, Si3N4 ceramics can be processed with
high reliability and can be adjusted over a broad range of properties for
desired applications. However, this state of the art is based mainly on
empirical experience. A quantitative understanding of the whole chain of
events from the production of powders, their densi®cation up to the
microstructure and its connection with properties of the components has
yet to be achieved.
Si3N4 ceramics are multicomponent systems of higher order and it is
obvious that the corresponding phase diagrams are very powerful tools in
understanding the relations between processing, microstructural development
and ®nal properties. In establishing these interrelations, all chemical constit-
uents and phases must be considered; even small amounts of impurities can
produce large effects. So far, all available systems have been investigated only
partially, and more than a few experimental results have been misinterpreted.
Therefore, a systematic study of constitutional phase diagrams of Si3N4
ceramics, including their promising composites, is necessary and this should
be combined with thermodynamic calculations of the equilibria and
metastable states.
The advances that have achieved so far are mainly based on increasing
understanding of processing and microstructure/property relationship. New
analytical methods and high resolution transmission electron microscopy have
provided new insight into the grain boundary region and offer the possibility
to tailor the microstructure for speci®c applications.
But in the area of speci®c microstructural engineering more knowledge is
desirable concerning the interplay between the microstructural features ±
Si3N4 (ss) grains, secondary phases (in general oxide nitrides) and amorphous
phase ± during sintering heat treatments and long-term high temperature
exposure to avoid improper deviations and defects.
Especially the phenomenon of the amorphous grain boundary phase has
not been clari®ed completely and is a challenge for basic research. Similarly
open questions exist concerning the optimisation of ®ne-grained and
nanoscaled materials for applications under extreme conditions. Here, the
recently developed ®ne b-powders may have greater promise than the
commonly used a-rich powders. The correlations between microstructure
wear, creep, corrosion and long-term behaviour are by far less understood
than the correlations between microstructure, strength and toughness. More
intensi®ed research on this topic should include the Si3N4 composites which
may have some advantages. Concerning improvement of fracture toughness
ass and ass/bss compositions are the prime candidates for study, their
mechanical properties can be substantially improved by tailoring their ®brous
microstructure.
The possibility to produce Si3N4 ceramics of precise composition and high

purity from low molecular weight, inorganic or organoelement precursors is a
topic of increasing relevance for small and thin components.
Si3N4 ceramics have not only the potential to compete successfully with
other engineered materials, but also to initiate new technical concepts which
have not caught on for lack of suitable materials. On the other hand, the
introduction to the market is slow and Si3N4 ceramics still are a niche market.
Cost still remains a major barrier. To overcome this handicap a cost reduction
is urgent. In this connection, a less-expensive raw material for powder
production and lower grade qualities of the starting powders are decisive
factors. Furthermore, improving the steps in the production and densi®cation
processes may cut the costs. These include intimate mixing of Si3N4 powders
with additives, avoidance of agglomerates, and homogeneous particle distri-
bution in the green state, pore minimisation during densi®cation and
elimination of all defects occurring in the various stages. Lower sintering
temperatures, cheaper surface ®nishing methods (diamond free) and higher
reliability of the components cut costs and therefore are crucial subjects of
study. In addition to cost problems holding back progress are gaps in
knowledge and experience and even unfounded caution on the part of
application engineers. Sound internationally accepted standards would be
helpful in overcoming these obstacles.
Si3N4 ceramics have developed into a family of very well established
materials with many useful properties, of which resistance to thermal shock,
high temperature stability, hardness and wear resistance are the most
important. The high potential of this versatile class of ceramics is evident,
but is far from being exhausted.
12
References
1. Nittler LR, Hoppe P, Alexander CMOD, Amari S, Eberhardt P, Gao X, Lewis RS, Strebel
R, Walker RM, Zinner E (1995) The Astrophysical J 453: L25
2. Lee MR, Russel SS, Arden JW, Pillinger CT (1995) Meteoritics 30: 387
3. Sainte-Claire Deville H, WoÈhler F (1859) Ann Chem Pharm 34: 248
4. Melner H (1896) German Patent No 88999
5. Weiss L, Engelhardt T (1910) Z Anorg Chem 65: 38
6. WoÈhler L (1926) Z Electrochem 32: 420
7. SchroÈder F (ed) (1989) Gmelin Handbook of Inorganic and Organometallic Chemistry,
8th edn., Silicon, Suppl Vol B4, Springer, Berlin, Heidelberg, New York
8. DP 235 421 (1908), DP 238450 (1909)
9. Collins JF, Gerby RW (1955) J Met 7: 612
10. Reinhardt F (1958) Glas-Email-Keramo Technologie 9: 327
11. Parr NL, Martin GF, May ERW (1960) Preparation, Microstructure and Mechanical
Properties of Silicon Nitride. In: Popper P (ed) Special Ceramics. Heywood, London,
p 102
12. Popper P, Ruddlesden SN (1961) Trans Br Ceram Soc 60: 603
13. Deeley GG, Herbert JM, Moore NC (1961) Powder Metall 8: 145
14. Riley FL (2000) J Am Ceram Soc 83: 245
15. Herrmann M, Klemm H, Schubert C (2000) Silicon Nitride Based Hard Materials. In:
Riedel R (ed) Handbook of Ceramic Hard Materials. Wiley-VCH, Weinheim, p 747
16. Chen IW, Becher PF, Mitomo M, Petzow G, Yen TS (1993) Symp Proc, MRS, Pittsburgh
17. Hampshire S (1994) Nitride Ceramics. In: Swain MV (ed) Structure and properties of
ceramics. Cahn RW, Haasen P, Kramer EJ (eds) Materials Science and Technology,
Comprehensive Treatment. VCH, Weinheim, p 119
18. Gmelin Handbook of Inorganic and Organometallic Chemistry, 8th edn. Springer,
Berlin, Heidelberg, New York; (a) SchroÈder F (ed) (1996) Silicon Nitride: Mechanical
and Thermal Properties: Diffusion. Silicon Suppl Vol B5b1; (b) SchroÈder F (ed)
(1997) Silicon Nitride: Electronic Structure; Electrical, Magnetic and Optical
Properties; Spectra; Analysis. Silicon Suppl Vol B5b2; (c) SchroÈder F (ed) (1991)
Silicon Nitride in Microelectronics and Solar Cells. Silicon Suppl Vol B5c; (d)
SchroÈder F (ed) (1995) Silicon Nitride: Electrochemical Behavior; Colloidal Chemistry
and Chemical Reactions. Silicon Suppl Vol B5d1; (e) SchroÈder F (ed) (1995) Silicon
Nitride: Chemical Reactions (continued). Silicon Suppl Vol B5d2; (f) SchroÈder F (ed)
(1994) Non-Electronic Applications of Silicon Nitride. SiNx. SiNx:H. Silicon Suppl Vol
B5e
19. Hoffmann MJ, Petzow G (eds) (1994) Tailoring of Mechanical Properties of Si3N4
Ceramics. NATO ASI Ser E 276. Kluwer Acad Publ, Dordrecht
20. Hoffmann MJ, Becher PF, Petzow G (eds) (1994) Silicon Nitride 93 ± Proc Int Conf on
Silicon Nitride-Based Ceramics, Stuttgart Trans Tech Publications Ltd, Aedermanns-
dorf, Switzerland
21. Lange H, WoÈtting G, Winter G (1991) Angew Chem Int Ed Engl 30: 1579
22. Wang CM, Pan X, RuÈhle M, Riley FL, Mitomo M (1996) J Mat Sci 31: 5281
23. Zerr A, Miehe G, Serghiou G, Schwarz M, Kroke E, Riedel R, Fueû H, Kroll P, Boehler R
(1999) Nature 400: 340
24. Toraya H (2000) J Appl Cryst 33: 95
25. GruÈn R (1979) Acta Cryst B35: 800
26. Schwarz M, Miehe G, Zerr A, Kroke E, Poe BT, Fuess H, Riedel R (2000) Adv Mat 12: 883
27. Hiraga K, Tsuno K, Shindo D, Hirabayashi M, Hagashi S, Hirai T (1983) Phil Mag A 47:
483
28. Wendel JA, Goddard III WA (1992) J Chem Phys 97: 5048
29. He H, Sekine T, Kobayashi T, Hirosaki H, Suzuki J (2000) Phys Rev B62/17: 1
30. Kohatsu I, Mc Cauley JW (1974) Mat Res Bull 9: 917
31. Suematsu H, Petromic JJ, Mitchell TE (1996) Mat Sci Eng 209: 97
32. Chakraborty D, Mukerji J (1983) Mat Res Bull 17: 843
33. Sun EY, Becher PF, Plucknett KP, Hueh CH, Alexander KB, Waters SB (1998) J Am
Ceram Soc 81: 2821 and 2831
34. Dusza J, Eschner T, Rundgren K (1997) J Mat Sci Lett 16: 1664
35. Ching W, Xu Y, Gale J, RuÈhle M (1998) J Am Ceram Soc 81: 3189
36. Hay JC, Sun EY, Pharr GM, Becher PF, Alexander KB (1998) J Am Ceram Soc 81: 2661
37. O'Hare PAG, Tomaszkiewicz I, Beck II CM, Seifert HJ (1999) J Chem Thermodynamics
31: 303
38. Liang JL, Topor L, Navrotsky A (1999) J Mat Res 14: 1959
39. Andrievskii RA, Spivak II (1984) Nitrid Kremnija i Materialy na ego ocnove,
Metallurgija, Moskau, p 19
40. Kitayama M, Hirao K, Toriyama M, Kanzaki S (1999) J Am Ceram Soc 82: 3105
41. Watari K (2001) J Ceram Soc Jpn 109: S7
42. Kunz KP, Sarin VK, Davis RF, Bryan SR (1988) Mat Sci Eng A 105/106: 47
43. Kijima K, Shirasaki S (1976) J Chem Phys 65: 2668
44. Clancy WP (1974) Microscope 22: 279
45. Wild S, Grieveson P, Jack KH (1972) Special Ceramics 5: 385
46. Konstanovskii AV, Evseev AV (1994) High Temp (translated from Teplo®zika
Vysokikh Temperatur, Russia) 32: 25
47. Cerenius Y (1999) J Am Ceram Soc 82: 380
48. Schneider J, Frey F, Johnson N, Laschke K (1994) Z Kristallographie 209: 328
49. Kitayama M (1999) J Am Ceram Soc 82: 3263
50. Thompson DP (1993) New Grain Boundary Phases for Nitrogen Ceramics. In: Chen IW,
Becher PF, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics. Mat Res Soc
Symp Proc 287: 79
51. Milhet X, Demenet JL, Rabier (1999) Phil Magaz Lett 79: 19
52. Wang CM, Riley FL (1996) J Eur Ceram Soc 16: 679.
53. Bowen LJ, Carruthers TG, Brook RJ (1978) J Am Ceram Soc 61: 335
53 a. Saito T, Iwamoto Y, Ukyo Y, Ikuhara Y (2000) Interface Characterisation of a-b
Phase Transformation in Si3N4 by Transmission Electron Microscopy. In: Sakuma T,
Sheppart LM, Ikuhara Y (eds) Grain Boundary Engineering in Ceramics, Am Ceram
Soc, Westerville OH, p 173
54. Colquhoun J (1973) Proc Br Cer Soc 22: 207
55. Chase MW JR, Davies CA, Downey JR, Frurip DJ, McDonald RA, Syverud AN (1985)
J Phys Chem Ref Data 14 Suppl. 1
56. The American Ceramic Society (ACerS) (ed) (1996) Cumulative Indexes of ``Phase
Equilibria Diagrams ± Phase Diagrams for Ceramists''. Vol I-XII and Annuals 91±93,
Westerville, Ohio
57. Alper AM (ed) (1995) Phase Diagrams in Advanced Ceramics. Academic Press, London
58. Massalski TB (ed) (1990) Binary Alloy Phase Diagrams. American Society for Materials
(ASM), Materials Park, Ohio
59. Petzow G, Effenberg G, Aldinger F (eds) (1988±2000) Ternary Alloys ± A Comprehen-
sive Compendium of Evaluated Constitutional Data and Phase Diagrams Materials
Science International Series (MSI),Vol. 1±17, Stuttgart
60. Blegen K (1976) PhD Thesis, Techn University of Trondheim, Norway
61. Hillert M, Jonsson S, Sundman B (1992) Z Metallkde. 83: 648
62. Heuer AH, Lou VL (1990) J Am Ceram Soc 73: 2785
63. Hallstedt B (1992) CALPHAD 16: 53
64. Swamy V, Saxana SK, Sundman B (1994) J Geoph Res 99: 11 and 787
65. Wriedt HA (1990) Bull Alloy Phase Diagr 11: 43
66. Rocabois P, Chatillon C, Bernard C (1996) J Am Ceram Soc 79: 1351 and 1361
67. Weiss J, Lukas HL, Petzow G (1983) Calculation of Phase Equilibria in Systems based
on Si3N4. In: Riley FL (ed) Progress in Nitrogen Ceramics. NATO ASI Series E. 65: 77
68. Richter HJ, Herrmann M, Hermel W (1991) J Eur Ceram Soc 7: 3
69. Lukas HL, Weiss J, Krieg H, Henig ET, Petzow G (1982) High Temp High Press 14: 607
70. Gauckler LJ (1976) PhD Thesis, University of Stuttgart
71. Jack KH (1976) J Mat Sci 11: 1135
72. Gauckler LJ, Lukas HL, Petzow G (1975) J Am Ceram Soc 58: 346
73. Lumby RJ, Noith B, Taylor AJ (1975) Spec Ceram Br Ceram Res Association 6: 283
74. Naik JK, Gauckler LJ, Tien TY (1978) J Am Ceram Soc 61: 332
75. Bergmann B, EkstroÈm T, Micski A (1991) J Eur Ceram Soc 8: 141
76. Boskovic S, Gauckler LJ, Petzow G, Tien TY (1977) Powder Metal Intern 9: 185; (1978)
10: 184; (1979) 11: 169
77. Hillert M, Jonsson S (1992) Z Metallkde 83: 720
78. Dumitrescu LFS, Sundman B (1995) J Eur Ceram Soc 15: 89 and 239
79. Anya CC, Hendrix A (1992) J Eur Ceram Soc 10: 65
80. Sekercioglu J, Wills RR (1979) J Am Ceram Soc 62: 590
81. Hillert M, Jonsson S (1992) CALPHAD 16: 199
82. Tien TY, Petzow G, Gauckler LJ, Weiss J (1983) Phase Equilibrium Studies in Si3N4-
Metal Oxides Systems. In: Riley FL (ed) Progress in Nitrogen Ceramics. NATO ASI Ser
E 65. Kluwer Acad Publ, Dordrecht, p 89
83. Thompson DP, Korgul P, Hendry A (1983) The structural characterisation of SiAlON
Polytypoids. In: Riley FL (ed) Progress in Nitrogen Ceramics. NATO ASI Ser E 65.
Kluwer Acad Publ, Dordrecht, p 61
84. Gauckler LJ, Weiss J, Tien TY, Petzow G (1978) J Am Ceram Soc 61: 397
85. Zhou Y, Vleugels J, Laoui T, Ratchev P, Van Der Biest O (1995) J Mater Sci 30: 4584
86. Liang JJ, Navrotzky A, Leppert VJ, Paskowitz MJ, Risbud SH, Ludwig T, Seifert HJ,
Aldinger F (1999) J Mater Res 14: 4630
87. Mitomo M, Kuramato N, Tsutsumi M, Suzuki H (1987) Yogyo Kyokaishi 86: 880
88. EkstroÈm T, KaÈll PO, Nygren M, Olsson PO (1989) J Mater Sci 24: 1853
89. Huseby C, Lukas HL, Petzow G (1975) J Am Ceram Soc 58: 337
90. Thompson DP, Gauckler LJ (1977) J Am Ceram Soc 60: 470
91. Kim DJ, Greil P, Petzow G (1987) Adv Ceram Mat 2: 817; 2: 822
92. Ekelund M, Forslund M, Erikson G, Johansson T (1988) J Am Ceram Soc 71: 956
93. Neidhardt V, Schubert H, Bischoff E, Petzow G (1994) Key Eng Mat 89: 187
94. Wada H, Wang HJ, Tien TY (1988) J Am Ceram Soc 71: 837
95. Jha A (1993) J Mat Sci 28: 3069
96. Wang MJ, Wada H (1990) J Mat Sci 25: 1690
97. Lange FF (1980) J Am Ceram Soc Bull 59: 249
98. Hampshire S, Park HK, Thompson DP, Jack KH (1978) Nature 274: 880
99. Huang ZK, Chen IW (1996) J Am Ceram Soc 79: 2091
100. Huang ZK, Tien TY, Yen TS (1986) J Am Ceram Soc 69: C241
101. Kaiser A (1999) PhD Thesis, RWTH Aachen
102. Mitomo M, Izumi F, Horiuchi S, Matzui J (1982) J Mater Sci 17: 2359
103. Mueller R (1981) PhD Thesis, University of Stuttgart
104. Lange FF (1978) J Am Ceram Soc 61: 53
105. Clarke DR, Lange FF (1980) J Am Ceram Soc 63: 585
106. Slasor S (1986) PhD Thesis, University of Newcastle Upon Tyne
107. Slasor S, Liddell K, Thompson DP (1986) The Role of Nd2O3 as an Additive in the
Formation of a¢- and b¢-Sialons. In: Howlett SP, Taylor D (eds) Proc Special Ceramics,
p 35; (1986) Br Ceram Proc 37: 61
108. Hohnke H, Gauckler LJ, Schneider G, Tien TY (1979) Am Ceram Soc Bull 58: 885
109. Jack KH (1978) Mat Sci Res 11: 561
110. Gauckler LJ, Hohnke H, Tien TY (1980) J Am Ceram Soc 63: 35
111. Schubert H, Gehrke E (1994) Processing, Phase Formation and Creep Behavior of Si3N4
with Y2O3 and Al2O3. In: Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical
Properties of Si3N4 Ceramics. NATO ASI Ser E 276, Kluwer Acad, Dordrecht, p 245
112. Sun WY, Huang ZK, Tien TY, Yan DS (1991) Mat Lett 11: 67
113. Hoffmann MJ (1994) High Temperature Properties of Yb-Containing Si3N4. In:
Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical Properties of Si3N4 Ceramics,
Kluwer Acad Publ, Dordrecht, p 233
114. Weiss J, Gauckler LJ, Tien TY (1979) J Am Ceram Soc 62: 632
115. Hampp E (1993) PhD Thesis, University of Stuttgart
116. Nishimura T, Mitomo M (1995) J Mater Res 10: 240
117. Huang ZK, Sun WY, Yen TS (1985) J Mat Sci Lett. 4: 255
118. Hewett CL, Cheng YB, Muddle BC, Trigg MB (1998) J Eur Ceram Soc 18: 417
119. Hewett CL, Cheng YB, Muddle BC, Trigg MB (1998) J Am Ceram Soc 81: 1781
120. Wood CA, Cheng YB (2000) J Eur Ceram Soc 20: 357
121. Wang PL, Zhang C, Sun WY, Yan DS (1999) J Eur Ceram Soc 19: 553
122. Wood CA, Zhao H, Cheng YB (1999) J Am Ceram Soc 82: 421
123. AS3. EkstroÈm T (1997) a-SiAlON and a/b-SiAlON composites; Recent research. In:
Babini GN, Haviar M, Sajgalik P (eds) Engineering CeramicsÔ96; Higher Reliability
through Processing, Kluwer Academic Publications, The Netherlands, p 147
124. EkstroÈm T, Nygren M (1992) J Am Ceram Soc 75: 259
125. Nordberg LO, Shen Z, Nygren M, EkstroÈm T (1997) J Eur Ceram Soc 17: 575
126. Sun WY, Yan DS, Gao L, Mandal H, Liddel K, Thompson DP (1995) J Eur Ceram Soc
15: 349
127. Nordberg LO (1997) PhD Thesis, University of Stockholm
128. Mandal H, Thompson DP (1999) J Eur Ceram Soc 19: 543
129. Mukerji J, Das PK, Greil P, Petzow G (1987) Ceram Intern 13: 215
130. Mandal H, Thompson DP (1996) J Mater Sci Letters 15: 1435
131. Mandal H, Hoffmann MJ (2000) Novel Developments in a-SiAlON Ceramics.

In: Sajgalik P, Lences Z (eds) Engineering Ceramics: Multifunctional Properties±New
Perspectives. Trans Tech Publications Ltd, Switzerland, p 131
132. Izhevsky VA, Genova LA, Bressiani JC, Aldinger F (2000) J Eur Ceram Soc 20: 2275
133. Wang PL, Tu HY, Sun WY, Yan DS, Nygren M, EkstroÈm T (1995) J Eur Ceram Soc 15: 689
134. Zhang C, Sun WY, Yan DS (1998) J Mater Sci Letters 17: 583; (1999) J Eur Ceram Soc
19: 33
135. Mitomo M, Ishida A (1999) J Eur Ceram Soc 19: 7
136. Sun WY, Tu HY, Wang PL, Yan DS (1997) J Eur Ceram Soc 17: 789
137. Liddell K, Thompson DP, Wang PL, Sun WY, Gao L, Yan DS (1998) J Eur Ceram Soc 18:
1479
138. Mandal H, Thompson DP, EkstroÈm T (1993) J Eur Ceram Soc 12: 421
139. Lewis MH, Jumali MHH, Lumby RJ (1997) Nd- and Gd-doped a¢/b¢ SiAlON Ceramics.
In: Niihara K, Hirano S, Kanzaki S, Komeya K, Moriuaga K (eds) Ceramic Materials
and Components for Engines. Japan Fine Ceramics Association, Tokyo, p 643
140. Zhang E, Liddell K, Thompson DP (1996) Brit Ceram Trans 95: 169
141. Yu ZB, Thompson DP, Bhattit R (2000) J Eur Ceram Soc 20: 1815
142. Nunn SD, Hohnke H, Gauckler LJ, Tien TY (1978) J Am Ceram Soc Bull 57: 321
143. Weiss J (1977) MS thesis, University of Stuttgart
144. Chee KS, Cheng YB, Smith ME (1995) J Eur Ceram Soc 15: 1213
145. Sun WY, Yan DS, Gao L, Mandal H, Lidell K, Thompson DP (1996) J Eur Ceram Soc 16:
1277
146. Kaiser A, Telle R, Richter HJ, Herrmann M, Hermel W (1996) Konstitutionsunter-
suchungen im System Nd-Si-Al-ON. In: Petzow G, Tobolski J, Telle R (eds)
Hochleistungs-Keramiken. VCH, Weinheim, p 627; (2001) Z. Metallkd. 92: 1163
147. Rosen¯anz A, Chen IW (1999) J Am Ceram Soc 82: 1025
148. Shen Z, Nygren M (1997) J Eur Ceram Soc 17: 1639
149. Shen Z, Nygren M (2000) Nd-Doped a-Sialon and Related Phases: Stability and
Compatibility. In: Hampshire S, Pomeroy MJ (eds) Nitrides and Oxynitrides. Trans
Tech Publications Ltd, Switzerland, p 191
150. Nordberg LO, Nygren M, KaÈll PO, Shen Z (1998) J Am Ceram Soc 81: 1461
151. Shen Z, EkstroÈm T, Nygren M (1996) J Am Ceram Soc 79: 721
152. Herrmann M, Kurama S, Mandal H (2002) J Eur Ceram Soc 22: 109
153. Zhao R, Cheng YB (1995) J Eur Ceram Soc 15: 1221
154. Chen WW, Li YW, Sun WY, Yan DS (2000) J Eur Ceram Soc 20: 1327
155. Sun WY, Tien TY, Yen TS (1991) J Am Ceram Soc 74: 2547 and 2753
156. Wisnudel M (1991) MS Thesis, University of Michigan, Ann Arbor
157. Hwang CM (1988) PhD Thesis, University of Michigan, Ann Arbor
158. Naik JK, Tien TY (1979) J Am Ceram Soc 62: 642
159. Thompson DP (1986) Phase relationships in Y-Si-Al-O-N ceramics. In: Tressler RE,
Messing GL, Pantano CG, Newnham RE (eds) Mat Sci Res Proc 21st Univ Conf in
Ceramic Science; Tailoring Multiphase and Composite Ceramics 20. Plenum Press, New
York, p 79
160. Sun WY, Huang ZK, Chen JX (1983) Trans J Brit Ceram Soc 82: 173
161. Huang ZK, Greil P, Petzow G (1983) J Am Ceram Soc 66: C96
162. Huang ZK, Tien TY (1996) J Am Ceram Soc 79: 1717
163. Wisnudel M, Tien TY (1994) J Am Ceram Soc 77: 2653
164. Klemm H, Herrmann M, Schubert C, Hermel W (1995/1996) High Temp±High Press
27/28: 449
165. Redington Keely W, Redington M, Hampshire S (2000) Liquid Formation in the
Y-Si-Al-O-N System. In: Hampshire S, Pomeroy MJ (eds) Nitrides and Oxinitrides.
Mater Sci Forum 325±326. Trans Tech Publications Ltd, Switzerland, p 237
166. Bandyopadhyay S, Hoffmann MJ, Petzow G (1996) J Am Ceram Soc 79: 1537
167. EkstroÈm T, McKenzie KJD, Ryan MJ, Brown WM, White GV (1997) J Mater Chem 7:
505
168. Kolitsch V, Seifert HJ, Ludwig T, Aldinger F (1999) J Mater Res 14: 447
169. Stutz D, Greil P, Petzow G (1986) J Mat Sci Lett 5: 335
170. Izumi F, Mitomo M, Suzuki J (1982) J Mater Sci Lett 1: 533
171. Shen Z, EkstroÈm T, Nygren M (1996) J Phys D Appl Phys 29: 893
172. Weiss J (1980) PhD Thesis, University of Stuttgart
173. Seifert HJ (1999) Z Metallkd 90: 1016
175. Cao GZ, Huang ZK, Yan DS (1989) Sci China Ser A 32: 429
176. Mahoney FM (1992) PhD thesis, University of Stuttgart
177. DoÈrner P (1982) PhD Thesis, University of Stuttgart
178. Wang PL, Li YW, Yan DS (2000) J Eur Ceram Soc 20: 1333
179. Gauckler LJ, Hucke E, Lukas HL, Petzow G (1978) CALPHAD VII: 49; (1979) J Mat Sci
14: 1513
180. Ran Q (1987) PhD Thesis, University of Stuttgart
181. Gauckler LJ, Petzow G (1977) Representation of Multicomponent Silicon Nitride Based
Systems. In: Riley FL (ed) Nitrogen Ceramics, Noordhoff-Leyden, p 41
182. Metselaar R, Yan DS (1999) Pure Appl Chem 71: 1765
183. Oyama Y, Kamigaito O (1971) Jap J Appl Phys 10/11: 1637
184. Jack KH, Wilson WJ (1972) Nature (London), Phys Sci 283: 28
185. Willems HX, Hendrix MMRM, Metselaar R, de With G (1992) J Eur Ceram Soc 10: 327;
10: 339
186. Hallstedt B, Hillert M, Selby M, Sundman B (1994) CALPHAD 18: 31
186a. Sekine T, He H, Kobayashi T, Tansho M, Kimoto K (2001) Chem Phys Let 344: 395
187. Cao GZ, Metselaar R (1991) Chem Mater 3: 242
188. Mandal H, Thompson DP, Jack KH (1999) Key Eng Mat 154±160: 1
189. Mandal H, Hoffmann MJ (1999) J Am Ceram Soc 82: 229
190. Lathrop D (2000) Ceram Bull 79: 54
191. Schwier G, Nietfeld G, Franz G (1989) Mat Sci Forum 47: 1
192. Herrmann M, Schulz I, Hintermayer J (1995) Materials From Low Cost Silicon Nitride
Powders In: Galassi C (ed) Proc 4th EcerS Conf, Riccione, Gruppo Editoriale Faenza
Editrice, p 211
193. Chang FW, Liou TH, Tsai FM (2000) Thermochimica Acta 354: 71
194. Liu YD, Kimura S (1999) Powder technology 106: 160
195. Hirotsuru H, Isozaki K, Yoshida A (1993) Trans Mat Res Soc Jap 14A: 815
196. Nakamura M, Kuranari Y, Imamura Y (1987) Characterisation and synthetic process of
Si3N4 material powders. In: Somiya S, Mitomo M, Yoshimura M (eds) Silicon Nitride I.
Elsevier, London, p 40
197. Kohtoku Y (1987) Developments in Si3N4 powder prepared by the imide decompo-
sition method. In: Somiya S, Mitomo M, Yoshimura M (eds) Silicon Nitride I. Elsevier,
London, p 71
198. Cochran GA, Conner CL, Eismann GA, Weimer AW, Carroll DF, Dunmead SD, Hwang
CJ (1994) The Synthesis of a High Quality, Low Cost Silicon Nitride Powder by the
Carbothermal Reduction of Silica. In: Hoffmann MJ, Becher PF, Petzow G (eds) Silicon
Nitride 93. Key Eng Mater 89±91. Trans Tech Publications Ltd, Switzerland, p 3
199. Bandyopadhyay S, Mukerji J (1991) Ceram Inter 17: 171
200. Ishii T, Sano A, Imai I (1987) a-Si3N4 powder produced by nitriding silica using
carbothermal reduction. In: Somiya S, Mitomo M, Yoshimura M (eds) Silicon Nitride
I. Elsevier, London, p 59
201. Bandyopadhyay S, Mukerji J (1992) Ceram Inter 18: 308
202. Arik H, Saritas S, Gunduz M (1999) J Mat Sci 34: 835
203. Rahman A, Riley FL (1989) J Eur Ceram Soc 5: 11
204. Mukerji J, Bandyopadhyay S (1988) Adv Ceram Mat 3: 369
205. Brink R, Woditsch P (1998) c®/Ber DKG, 75/7: 22
206. Dorn FW, Krause W, SchroÈder F (1992) Adv Mat 4: 221
207. Jennett TA, Harmsworth PD, Jones AG (1994) Ultra ®ne crystalline Silicon Nitride from
a continuous Gas Phase Plasma Route. In: Hoffmann MJ, Becher PF, Petzow G (eds)
Silicon Nitride 93. Key Eng Mat 89±91. Trans Tech Publications Ltd, Switzerland, p 47
208. Grabis J, Zalite I, Reichel U (2000) c®/Ber DKG 77/7: 9
209. Kubo N, Futaki S, Shiraishi K (1987) Synthesis of ultra®ne Si3N4 powder using the
plasma process and powder characterisation. In: Somiya S, Mitomo M, Yoshimura M
(eds) Silicon Nitride I. Elsevier, London, p 93
210. Friedrich M, Mohr R, Drost H, Mach R, Gey E (1997) Silicates Industriels 11: 1±2
211. Drost H, Friedrich M, Mohr R, Gey E (1997) Nucl Instr and Methods in Physics Res
122: 598
212. Borsella E, Caneve L, Fantoni R, Piccirillo S, Basili N, Enzo S (1989) Appl Surface Sci 36:
213
213. Kroke E, Li YL, Konetschny C, Lecomte E, Fasel C, Riedel A (2000) Mat Sci Eng R26: 97
214. Merzhanov AG (1995) Ceram Int 21: 371
215. Temer MR, Swenser SP, Cheng YB (2001) Characterization of Multi-Cation stabilized
Alpha-SiAlON Materials. In: Heinrich JG, Aldinger F (eds) Ceramic Materials and
Components for Engines, 7th International Symposium. Wiley VCH, Weinheim, p 447
216. Hirata T, Akiyama K, Morimoto T (2000). J Eur Ceram Soc 20: 1191
217. Bermudo VJ Osendi MI (1999) Ceram Intern 25: 607
218. Wu Y, Zhuang H, Wu F, Dollimore D, Zhang B, Chen-Li W (1998) J Mat Res 13: 166
219. Herrmann M, Schulz I, Hermel W, Schubert C, Wendt A (2001) Z Metallkde 92: 788
220. Arakawa T (1987) State of the art of silicon nitride powders obtained by thermal
decomposition of Si(NH)2 and the injection molding thereof. In: Somiya S, Mitomo M,
Yoshimura M (eds) Silicon Nitride I. Elsevier, New York, p 81
221. Brink R, Lange H (1994) Investigations on the Synthesis of ®ne-grained, high-purity
b-Si3N4 Powder by Crystallization of amorphous Precursors. In: Hoffmann MJ, Becher
PF, Petzow G (eds) Silicon Nitride 93. Key Eng Mat 89±91. Trans Tech Publications Ltd,
Switzerland, p 73
222. Borsella E, Botti S, Martelli S, Alexandrescu R, Cesile MC, Nesterenko A, Giorgi R,
Turtu S, Zappa G (1997) Silicates Industriels 3: 1±2
223. Niihara K (1991) J Ceram Soc Jpn Int Edition 99: 945
224. Petzow G, Sersale R (1987) Pure & Appl Chem 59: 1673
225. Samsonov GV, Kulik OP, Poluschschick VS (1978) Poluschenie i Methodij analisa
Nitridov. Naukowa Dumka, Kiev, p 273
226. Rabe T, Sontag E, Kranz G, RoÈhl K, Linke D (1990) Analytische Bestimmung der in
Siliciumnitrid vorliegenden Phasen. In: Kriegesmann J (ed) Keramische Werkstoffe,
Deutscher Wirtschaftsdienst, KoÈln, p 6.1.6.1
227. Haûler I, FoÈrster O, Schwetz KA (2000) c®/Ber DKG 77/7: D11
228. Peplinski B, Schultze D, Wenzel J (2001) Interlaboratory comparison (round robin) on
the application of the Rietveld method to quantitative phase analysis by X-ray and
neutron diffraction. In: Delhez R, Mittemeijer EJ (eds) Proc 7th European powder
diffraction conference (EPDIC-7), Barcelona, Spain, 20th 23rd May 2000, Trans Tech
Publications Ltd, Switzerland, p 124
229. Matern N, Riedel A, Wassermann A (1993) Mat Sci Forum 133: 39
230. Gazzara CP, Messier DR (1977) Ceram. Bull 56: 777
231. KaÈll PO, EkstroÈm T (1990) J Eur Ceram Soc 6: 119l
232. Goeb O (2001) PhD Thesis, Technical University of Dresden
233. Bermudo J, Osendi MI, Fierro JLG (2000) Ceram Intern 26: 141
234. Hirao K, Tsuge A, Toriyama M, Kanzaki S (1999) J Am Ceram Soc 82: 3263
235. Kaiser G, Schubert H (1993) J Eur Ceram Soc 11: 253
236. SunderkoÈtter JD, Grallath E, Jenett M (1993) Fresenius J Anal Chem 346: 237
237. Peuckert M, Greil P (1987) J Mater Sci 22: 3717
238. Richter HJ, Herrmann M (1991) J Mat Sci Lett 10: 783
239. Berger LM (1993) Application of Adsorption for the Characterization of Solids in

Ceramic and Related Technologies. In: Aldinger F (ed) PTM `93, DGM Information-
sgesellschaft, Oberursel, p 677
240. Scarlett B (1996) Characterisation of Particles and powders. In: Brook RI (ed)
Processing of Ceramics, Part I.VCH, Weinheim, p 99
241. Boden G, Nitsche R (1993) Sprechsaal 126: 220
242. Sonnefeld J (1996) Physicochem Eng Aspects 108: 27
243. Petzow G (1988) Pract Met 25: 53
244. Reed JS (1995) Principles of Ceramic Processing, John Wiley & Sons Inc, New York,
p 137
245. Riedel G, KruÈner H (1993) Si3N4 material with high strength (1400 MPa). In: Duran P,
Fernandez JF (eds) Third Euro-Ceramics 3. Faenza Editrice Iberica, San Vicente, p 453
246. Yoshimura M, Nishioka T, Yamakawa A, Miyake M (1995) J Ceram Soc Jpn 103: 407
247. Brook RJ (ed) (1990) Processing of Ceramics, Part I and II, Wiley-VCH, Weinheim
248. Riedel G, Schubert J (1996) Keram Zeit 48: 294 and 396
249. Riedel G, Krieger S (1996) Keram Zeit 48: 192
250. Kosmac T, Novak S, Sajko M (1997) J Eur Ceram Soc 17: 427
251. Hackley VA (1997) J Am Ceram Soc 80: 2315
252. Schwelm M (1992) PhD Thesis, University of Stuttgart
253. BergstroÈm L (1996) J Am Ceram Soc 79: 3033
254. Lange FF, Luther EP (1994) Colloidal Processing of Structurally Reliable Si3N4.
In: Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical Properties of Si3N4
Ceramics. NATO ASI Series E. Vol 276, Kluwer Acad Publ Dordrecht, p 3
255. Vieth S, Mitzner E, Finke B, Linke D (2000) c® Ber DKG 77(10): E63
256. Kuzjukevies A Ishizaki K (1993) J Am Ceram Soc 76: 2373
257. Pugh RI, BergstroÈm L (1994) Surface and Colloid Chemistry in Advanced Ceramics
Processing. Marcel Decker Inc, New York
258. Wang L, Sigmund W, Aldinger F (2000) J Am Ceram Soc 83: 691 and 697
259. Negita K (1985) J Mater Sci Letters 4: 755
260. Kim J, Iseki T (1996) J Am Ceram Soc 79: 2744
261. Wang L, Roy S, Sigmund W, Aldinger F (1999) J Eur Ceram Soc 19: 61
262. Wang C, Riley FL (1992) J Eur Ceram Soc 10: 83
263. Luther EP, Lange FF, Pearson DS (1995) J Am Ceram Soc 78: 2009
264. Wang L, Sigmund W, Roy S, Aldinger F (1999) J Mater Res 14: 4562
265. Greil P (1982) PhD thesis, University of Stuttgart
266. Herrmann M, Putzky G, Siegel S, Hermel W (1992) c® Ber DKG 69: 375
267. Yokoyama K, Wada S (2000) J Ceram Soc Jpn 108: 6
268. Neidhardt U (1993) PhD Thesis, University of Stuttgart
269. Riedel G, Bestgen H, Herrmann M (1998) c®, Ber DKG, 75/12: 30
270. Grechkovich C, Prochazka S (1981) J Am Ceram Soc 64: C96
271. Messier DR, Patel PJ (1993) Chemically induced defects in oxinitride glasses. In: Chen
IW, Becher PF, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics, Mat Res
Soc Symp 287; Mat Res Soc, Pittsburgh, p 365
272. Herrmann M, Goeb O (2001) J Eur Ceram Soc 21: 304 and 461
273. Yokoyama K, Wada S (1999) Effect of Sintering Atmosphere on the a/b-Transforma-
tion of Si3N4. In: Suzuki H, Komeya K, Uematsu K (eds) Novel Synthesis and
Processing of Ceramics, Key Eng Mat 159±160. Trans Tech Publications, Switzerland,
p 209
274. Yokoyama K, Wada S (2000) J Ceram Soc Jpn 108: 230 and 357
275. Herrmann M, Schubert C, Rendtel A, HuÈbner H (1998) J Am Ceram Soc 81: 1095
276. Pompe R, Carlson R (1983) Sintering of Si3N4 based Materials using the Powder Bed
Technique. In: Riley FL (ed) Progress in Nitrogen Ceramics, NATO ASI Ser E. 65.
Kluwer Academic Publishers, Dordrecht, p 219
277. Giachello A, Martinengo PC, Tommasini G, Popper P (1979) J Mat Sci 14: 2855
278. Warnecke G (ed) (2000) ZuverlaÈssige Hochleistungskeramik. UniversitaÈt Kaiserslau-

tern, Kaiserslautern
279. Richerson DW (1992) Modern Ceramic Engineering. Marcel Decker Inc, New York
280. Pattimore J, Nishio A, Dewitte C, Unno Y, Masuda M (1997) Optimisation of Grinding
for Cylindrical Silicon Nitride Components for Mass Production. In: Niihara K, Hirano
S, Kanzaki S, Komeya K, Moriuaga K (eds) Ceramic Materials and Component for
Engines. Japan Fine Ceramics Association, Tokyo, p 406
281. Greil P (1997) Near Net Shape Manufacturing of Polymer Derived Ceramics In: Baxter
J, Cot L, Fordham R Gabis V, Hellot Y (eds) Euro Ceramics 5 Part 3- Key Eng Mat
132±136 ; Trans Tech Publications, Switzerland, p 1981
283. KraÈmer M, Hoffmann MJ, Petzow G (1993) Acta metall mater 41: 2939
284. KraÈmer M, Hoffmann MJ, Petzow G (1993) J Am Ceram Soc 76: 2778
285. Dressler W (1993) PhD thesis, University of Stuttgart
286. Chen IW, Hwang SL (1993) Superplastic SiAlON±A Birds Eye View of Silicon Nitride
Ceramics. In: Chen IW, Becher PF, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride
Ceramics, Mat Res Soc Symp 287; Mat Res Soc, Pittsburgh, p 209
287. Bestgen H, Boberski C (1998) HTEM Analyse der Mikrostrukturentwicklung waÈhrend
des FluÈssigphasensinterns eines Si3N4/HfO2 Composites In: Heinrich J, Ziegler G,
Hermel W, Riedel H (eds) Werkstoffwoche'98 Band VII, Symposium 9, Keramik,
Wiley-VCH, Weinheim, p 401
288. Hwang SL, Chen IW (1994) J Am Ceram Soc 77: 1711
289. Mitomo M, Hirotsuru H, Suematsu H, Nishimura T (1995) J Am Ceram Soc 78: 211
290. Herrmann M, Schulz I, Schubert C, Zalite I, Ziegler G (1998) c®/Ber DKG 75: 38
291. Okamoto Y, Hirosaki N, AkimuÈne Y, Mitomo M (1997) J Ceram Soc Jpn 105: 476
292. Kitayama M, Hirao K, Toriyama M, Kanzaki S (1998) Anisotropie Ostwald Ripening in
b-Si3N4 with different Lanthanide Additives. In: Messing GL, Hirano S, Lange FF (eds)
Ceramic Processing Science. Ceram Transactions Ser 83: 517; (1999) J Ceram Soc Jpn
107: 930 and 107: 995
293. Becher PF, Sun EY, Plucknett KP, Alexander KB, Hsuch CH, Lin HAT, Waters SB,
Westmoreland CG (1998) J Am Ceram Soc 81: 2821
294. Woetting G, Feuer H, Gugel E (1993) The In¯uence of Powders and Processing
Methods on Microstructure and Properties of Dense Silicon Nitride. In: Chen IW,
Becher PF, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics, Mat Res Soc
Symp Proc 287. Mat Res Soc, Pittsburgh, p 133
295. Dressler W, Kleebe HJ, Hoffmann MJ, RuÈuÈhle M, Petzow G (1996) J Eur Ceram Soc 16: 3
296. Kessler S, Herrmann M, Schubert C (1991) Mat Sci Forum 94±96: 821
297. Tien TY (1993) Silicon Nitride Ceramics±Alloy Design. In: Chen IW, Becher PF,
Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics. Mat Res Soc Symp 287.
Mat Res Soc, Pittsburgh, p 51
298. Wang LL, Tien TY, Chen IW (1998) J Am Ceram Soc 81: 2677
299. Emoto H, Mitomo M (1997) J Eur Ceram Soc 17: 797
300. Kanamaru M (1994) PhD Thesis, University of Stuttgart
301. Riedel G, Bestgen H, Herrmann M (1999) c®/Ber DKG 71 No 1: 24
302. BjoÈrklund H, Falk LKL, Rundgren K, Wasen J (1997) J Eur Ceram Soc 17: 1285
303. Pyzik AJ, Beaman DR (1993) J Am Ceram Soc 76: 2737
304. Pyzik AJ, Carrol DF (1994) Annu Rev Mater Sci 24: 168
305. Ziegler G, Lehner W, Kleebe HJ (1999) Br Ceram Proc 60: 5
306. BraÈnvall P, Rundgren K (1999) Br Ceram Proc 60: 7
307. Kim HD, Han BD, Park DS (2001) Process to obtain Bimodal Microstructure in Silicon
Nitride. Paper presented at 25th Annual Conference on Composites, Advanced
Ceramics, Materials, and Structures. Am Ceram Soc, Cocoa Beach, Fl
308. Lai KR, Tien TY (1993) J Am Ceram Soc 76: 91
309. Mitomo M, Uenosono J, (1992) J Am Ceram Soc 75: 103
310. Mitomo M (1999) In-Situ Microstructural Control in Engineering Ceramics. In: Niihara
K, Sekino T, Yasuda E, Sasa T (eds) The Science of Engineering Ceramics II. Key Eng
Mat 161±163. Trans Tech Publications, Switzerland, p 53
311. Park DS, Lee SY, Kim HD, Yoo BJ, Kim BA (1998) J Am Ceram Soc 81: 1876
312. Stemmer S, Roebben G, Van der Biest O (1998) Acta mater 46: 5599
313. Cinibulk MK, Kleebe HJ Schneider GA, RuÈhle M (1993) J Am Ceram Soc 76: 2081
314. Wang CM, Pan X, Hoffmann MJ, Cannon RM, RuÈhle M (1996) J Am Ceram Soc 79: 788
315. Lee CJ, Kim DJ, Kang ES (1999) J Am Ceram Soc 82: 753
316. Bando Y, Mitomo M, Kurashima K (1998) J Mater Synthesis and Processing 6: 359
317. Kleebe HJ (1997) J Ceram Soc Jpn 105: 453
318. Clarke DR (1987) J Am Ceram Soc 70: 15
319. Raj R, Lange FF (1981) Acta Met 29: 1993
320. Yoshiya M, Tanaka I, Adachi H (2000) Atomic Structure and Chemical Bonding of
Intergranular ®lm in Si3N4-SiO2 Ceramics. In: Sajgalik P, Lences Z (eds) Engineering
Ceramics: Multifunctional Properties-New Perspectives. Trans Tech Publications Ltd,
Switzerland, p 107
321. Lewis MH (1994) Crystallisation of Grain Boundary Phases in Silicon Nitride and
SiAlON Ceramics. In: Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical Properties
of Si3N4 Ceramics. NATO ASi Ser E 276, KluÈwer Academic Publishers, Dordrecht, p 217
322. Kleebe HJ, Hoffmann MJ, RuÈhle M (1992) Z Metallkde 83: 610
323. Okamoto K, Kleebe HJ, Ota K, Pezzotti G (1999) J Jap Inst Mat 63: 1479
324. Tanaka J, Kleebe HJ, Cinibulk MK, Bruley J, Clarke DR, RuÈhle M (1994) J Am Ceram
Soc 77: 911
325. Kleebe HJ, Cinibulk MK, Cannon RM, RuÈhle M (1993) J Am Ceram Soc 76: 1969
326. BjoÈrklund H, Falk LKL (1997) J Eur Ceram Soc 17: 1301
327. Bobeth M, Clarke DR, Pompe W (1999) J Am Ceram Soc 82: 1537
328. Ackler HD, Chaing YM (1997) J Am Ceram Soc 80: 189
329. Keblinski P, Philpert SR, Wolf D, Gleiter H (1996) Phys Rev Lett 77: 2965
330. Golczewski JA, Seifert HJ, Aldinger F (2001) Z Metallkde 92: 695
331. Menon M, Chen IW (1995) J Am Ceram Soc 78: 545 and 553
332. Mandel H, Hoffmann MJ (1999) Br Ceram Proc 60: 11
333. Chen IW, Rosen¯anz A (1997) Nature 389: 701
334. Klemm H, Herrmann M, Reich T, Schubert C, Frassek L, WoÈtting G, Gugel E, Nietfeld G
(1998) J Am Ceram Soc 81: 1141
335. Mandal H (1999) J Eur Ceram Soc 19: 2349
336. Sheu TS (1994) J Am Ceram Soc 77: 2345
337. Rosen¯anz A (1999) Current Opinion in Sol State & Mat Sci 4: 453
338. Kim J, Rosen¯anz A, Chen IW (2000) J Am Ceram Soc 83: 1819
339. Wang CM, Hirosaki N, Mitomo M (2000) J Ceram Soc Jpn 108: 298
340. Herrmann M (1998) annual report 1998 Fraunhofer IKTS, Dresden
341. Kurama S, Herrmann M, Mandal H (2001) J Eur Ceram Soc accepted for publication
342. Rosen¯anz A, Chen IW (1999) J Eur Ceram Soc 19: 2337
343. Mandal H. Thompsson DP, EkstroÈm T (1993) J Eur Ceram Soc 12: 421
344. Cao GZ, Melselaar R, Ziegler G (1992) Microstructure and properties of mixed a¢+b¢-
Sialons. In: Carlsson R, Johansson T, Kahlman L (eds) Ceramics Materials and
Components for Engines. Elsevier, London, p 188
345. TaÈffner U, Carle V, SchaÈfer U, Predel F, Petzow G (1991) Pract Met 28: 592
346. MuÈcklich F, Ohser J, Hartmann S, Hoffmann MJ, Petzow G (1994) 3D-Characterisation
of Sintered Microstructures with Prismatic Grains±A Precondition for Microstructural
Modelling of Si3N4 Ceramics. In: Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical
Properties of Si3N4 Ceramics. NATO ASI Ser E 276. Kluwer Academic Publishers,
Dordrecht, p 73
347. Obenaus P, Herrmann M (1990) Pract Met 27: 503
348. Kawashima T, Okamoto H, Yamamoto H, Kitamura A (1991) J Ceram Soc Jpn 99: 1
349. Becher PF, Hwang SL, Lin HT, Ticgs TN (1994) Microstructural contributions to the
fracture resistance of silicon nitride ceramics. In: Hoffmann MJ, Petzow G (eds)
Tailoring of Mechanical Properties of Si3N4 Ceramics, NATO ASI Ser E Vol. 276,
350. Schneider JA, Mukherjee AK (1999) J Am Ceram Soc 82: 761
351. Tanaka I, Nasu S, Adachi H, Miyamoto Y, Niihara K (1992) Acta met mater 40: 1995
352. Wakai F (1994) Acta met 42: 1163
353. Wiederhorn SM, Hockey BJ, Crammer DC, Yeekley R (1993) J Mater Sci 28: 445
354. Chadwick MM, Jupp RS, Wilkinson DS (1993) J Am Ceram Soc 76: 385
355. Menon MM, Fang HAT, Wu DC, Jenkins MG, Ferber MK (1994) J Am Ceram Soc 77:
1228
356. Gogotsi YG, Grathwohl G, ThuÈmmler F, Yaroshenko VP, Herrmann M, Taut C (1993)
J Eur Ceram Soc 11 : 375
357. Bonnell DA (1986) PhD thesis, University of Michigan
358. Mandal H, Hoffmann MJ (2000) Hard and Tough a-SiALON Ceramics. In: Hampshire
S, Pomeroy MJ (eds) Nitrides and Oxynitrides. Trans Tech Publications Ltd,
Switzerland, p 219
359. Becher PF, Lin HAT, Hwang SL, Hoffmann MJ, Chen IW (1993) The in¯uence of
microstructure on the mechanical behavior of silicon nitride ceramics. In: Chen IW,
Becher P, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics±Scienti®c and
Technological Advances. MRS Symposium Proc 287, MRS Pittsburgh, p 147
360. Hoffmann MJ, Schneider GA, Petzow G (1993) The potential of Si3N4 for thermal shock
applications. In: Schneider GA, Petzow G (eds) Thermal Shock and Thermal Fatigue
Behaviour of Advanced Ceramics. NATO ASI Ser E 241, KluÈwer Academic Publishers,
Dordrecht p 49
361. Hoffmann MJ, Petzow G (1994) Pure & Appl. Chem 66: 1807
362. Chen L, Kuy E, Groboth G (1998) Surface and Coatings Technology 320: 100±101
363. Franchini C (1981) Ceramurgia 11: 140
364. Mitomo M, Uemura Y (1981) J Mat Sci Let 16: 5527
365. Thorp JS, Sharif I (1976) J Mat Sci 11: 1494
366. Niihara K (2001) Nanomaterials and Nanocomposites with Multifunctionality. Paper
presented at 25th Annual Conference on Composites, Advanced Ceramics, Materials,
and Structures. Am Ceram Soc, Cocoa Beach, Fl
367. Gnesin GG, Kirilenko VM, Petrovskii VA, Gervits EI, Chernavskii YA (1982) Poroshk
Metall 21: 53
368. Katz NR (1993) Applications of Silicon Nitride Based Ceramics in the U.S. In: Chen IW,
Becher PF Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics± Scienti®c and
Technological Advances. Mat Res Soc Symp Proc 287, MRS, Pittsburgh, p 197
369. Subirats M, Iskander MF, White MJ, Kiggans JO (1997) J of Microwave Power and
Electromagnetic Energy 32: 1997
370. Willert-Porada M, Dhupia G, MuÈller G, Nagel A (1999) Material and Technology
Development for Microwave Sintering of High Performance. In: MuÈller G (ed)
Ceramics-Ceramics-Processing, Reliability, Tribology and Wear, EUROMAT-Vol 12.
371. Watari K Seki Y, Ishizaki K (1989) J Ceram Soc Jpn Int Ed 97: 170
372. Kitayama M, Hirao K, Toriyama M, Kanzaki S (1999) J Am Ceram Soc 82: 3105
373. Watari K, Hirao K, Toriyama M, Ishizaki K (1999) J Am Ceram Soc 82: 777
374. Kitayama M, Hirao K, Watari K, Toriyama M, Kanzaki S (2001) J Am Ceram Soc 84:
353
375. Li B, Pottier L, Roger JP, Fournier D, Watari K, Hirao K (1999) J Eur Ceram Soc 19:
1631
376. Kitayama M, Hirao K, Tsuge A, Watari K, Toriyama M, Kanzaki S (2000) J Am Ceram
Soc 83: 1985
377. Hirao K, Imamura H, Watari K, Brito ME, Toriyama M, Kanzaki S (1999) Seeded
Silicon Nitride: Microstructure and Performance. In: Niihara K, Sekino T, Yasuda E,
Sasa T (eds) The Science of Engineering Ceramics II, Key Eng Mater 161±163. Trans
Tech Publications, Switzerland, p 469
378. Mitomo M, Izumi F, Greil P, Petzow G ( 1984) Bull Am Ceram Soc 65: 84
379. Herrmann M, Janecke-RoÈûler K (1999) Lab Rep, IKTS Dresden
380. Shermann D, Brandon D (2000) Mechanical Properties and their Relation to
Microstructure. In: Riedel R (ed) Handbook of Ceramic Hard Materials. Wiley-VCH,
Weinheim, p 66
381. Munz D, Fett T (1999) Ceramics Mechanical Properties, Failure Behaviour, Materials
Selection. Springer, Berlin, Heidelberg, New York
382. Quinn GD (1990) J Mater Sci 25: 4361
383. Wachtmann JB (1996) Mechanical Properties of Ceramics, John Wiley & Sons Inc, New
York
384. Kondo N, Yozuuki Y, Ohji T (1999) J Am Ceram Soc 82: 1067
385. Tajima Y, Urashima K (1994) Improvement of Strength and Toughness of Silicon
Nitride Ceramics. In: Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical Properties
of Si3N4 Ceramics, NATO ASI Series, Series E Vol 276. Kluwer Academic Publishers,
Dordrecht, p 101
386. Tseng WJ, Kita H, (2000) Ceram Intern 26: 197
387. Matsumaru K, Ishizaki K (2001) Fabrication of Porous Materials with high Fracture
Strength. In: Singh M, Jessen T (eds) 25th Annual Conference on Composites,
Advanced Ceramics, Materials, and Structures: B, (Ceram Eng Sci Proc 22). Am Ceram
Soc, Westerville, OH, p 197
388. Sajgalik P, Dusza J, Hoffmann MJ (1995) J Am Ceram Soc 78: 2619
389. Evans AG (1990) J Am Ceram Soc 73: 187
390. RoÈdel J (1992) J Eur Ceram Soc 10: 143
391. Kleebe HJ, Pezzotti G, Ziegler G, (1999) J Am Ceram Soc 82: 1857
392. Rice RW (2000) Mechanical Properties of Ceramics and Composites, Marcel Dekker,
New York, p 245
393. Faber KT, Evans AG(1983) Acta Metall 31: 565 and 577
394. Emoto H, Hirotsuri H (1999) Microstructure Control of Silicon Nitride Ceramics
Fabricated from a Powder Containing Fine b-Nuclei. In: Niihara K, Sekino T, Yasuda E,
Sasa T (eds) The Science of Engineering Ceramics II Key Eng Mat 161±163. Trans Tech
Publications, Switzerland, p 209
395. Walker T (1996) PhD Theses, Technical University of Dresden
396. Li CW, Gasdaska CJ, Goldacker J, Lui SC (1993) Damade Resistance of In Situ
Reinforced Silicon Nitride. In: Chen IW, Becher PF, Mitomo M, Petzow G, Yen TS (eds)
Silicon Nitride Ceramics ± Scienti®c and Technological Advances. Mat Res Soc Symp
Proc 287. Materials Research Soc, Pittsburgh, p 473
397. Urashima K, Ikea Y, IRAs S (1998) Features of Superior Strength Si3N4 Ceramic and its
Application. In: Niihara K, Hirano S, Kanzaki S, Komeya K, Miring K (eds) 6th Int Symp
Ceramic Materials and Components for Engines. Jap Fine Ceramic Ass, Tokyo, p 167
398. Silver RF, Vienna JM (1995) J Mat Sci 30: 5531
399. Peterson I, Tien T (1995) J Am Ceram Soc 78: 2345
400. Tanaka I, Pezzotti G, Okamoto T, Miyamoto Y, Koizumi M (1992) J Am Ceram Soc 74 :
752
401. Becher PF, Sun EY, Hsueh CH, Painter GS, More KL (2000) Role of Intergranular Films
in Toughened Ceramics. In: Sajgalic P, Luences Z (eds) Engineering Ceramics
Multifunctional Properties ± New Perspectives, Key Eng Mat 175±176. Trans Tech
402. Speicher R, Schneider GA, Dreûler W, Lindemann G, BoÈder H, Knoblauch V (1999)
Reliability of Ceramic Valve Plates for Common Rail Injection Pumps. In: MuÈller G (ed)
Ceramic Processing, Reliability, Tribology and Wear, EUROMAT99-Vol 12, Wiley-
VCH, Weinheim, p 333
403. Ogasawara T, Mabuchi Y (1993) J Ceram Soc Jpn Int Edition 101: 1122
404. Jacobs DS, Chen IW (1994) J Am Ceram Soc 77: 1153
405. Bhatnagar A, Hoffmann MJ, Dauskardt RH (2000) J Am Ceram Soc 83: 585
406. Andrievski RA (1994) High Temp±High Press 26: 451
407. Miao H, Qi L, Cui G (1995) Key Engin Mater 114: 135
408. Rice RW, Wu CC, Borchelt F (1994) J Am Ceram Soc 77: 2539
409. Petzow G, Hoffmann MJ (1993) Mat Sci Forum 91: 113±115
410. Klemm H, Tangermann K, Schubert C, Hermel W (1996) J Am Ceram Soc 79: 429
411. Tanaka I, Pezzotti G, Matsushita K, Miyamoto Y, Okamoto T (1992) J Am Ceram Soc
74: 752
412. Luecke WE, Wiederhorn SM (1999) J Am Ceram Soc 82: 2769
413. Lofaj F, Wiederhorn SM, Long GG, Jemian PR (2001) Tensile Creep in the next
Generation Silicon Nitride. In: Singh M, Jessen T (eds) 25th Annual Conference on
Composites, Advanced Ceramics, Materials, and Structures: A (Ceram Eng Sci Proc 22).
Am Ceram Soc, Westerville, OH, p 167
414. Ohji T (2001) Long Term Tensile at Creep Behaviours of Highly Heat-Resistant Silicon
Nitride Ceramics Gas Turbines. In: Singh M, Jessen T (eds) 25th Annual Conference on
Composites, Advanced Ceramics, Materials, and Structures: A (Ceram Eng Sci Proc 22).
Am Ceram Soc, Westerville, OH, p 159
415. Yoshida M, Tanaka K, Tsuruzono S, Tatsuki T (1999) Development of Ceramic
Components for Ceramic Gas Turbine Engine (CGT302). In: Vincentini P (ed)
Ceramics: Getting Into the 2000's, Part D, Elsevier, London, p 253
416. Luecke WE, Wiederhorn SM, Hockey BJ, Krause RF, Long GG (1995) J Am Ceram Soc
78: 2085
417. Raj R, Chyung CK (1981) Acta Metall 29, 159
418. Thouless MD, Evans AG (1984) J Am Ceram Soc 67: 721
419. Ohji T, Yamauchi Y (1993) J Am Ceram Soc 76: 3105
420. Ohji T (1994) Tensile Creep Rupture And Subcritical Crack Growth of Silicon Nitride.
In: Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical Properties of Si3N4
Ceramics. Kluwer Academic Publishers, Netherlands, p 339
421. Marion JE, Evans AG, Drory MD, Clarke DR (1983) Acta Metall 31: 1445
422. Dryden JR, Kucerovsky D, Wilkinson DS, Watt DF (1989) Acta Metall 37: 2007
423. Klemm H, Herrmann M, Schubert C (1998) High Temperature Oxidation of Silicon
Nitride Based Ceramics Materials. In: Niihara K, Hirano S, Kanzaki S, Komeya K,
Morinaga K (eds) 6th Int Symp Ceramic Materials and Components for Engines. Jap
Fine Ceramic Ass, Tokyo, p 576
424. Pezzotti G, Ota K, Kleebe HJ (1997) J Am Ceram Soc 80: 2341
425. Klemm H, Pezzotti G (1994) J Am Ceram Soc 77: 553
426. Xie RJ, Mitomo M, Zhan GD (2000) Acta mater 48: 2049
427. Shigegaki Y, Inamura T, Suzuki A, Sasa T (1993) High Temperature Fatigue Properties
of Silicon Nitride in Nitrogen Atmosphere. In: Chen IW, Becher PF Mitomo M, Petzow
G, Yen TS (eds) Silicon Nitride Ceramics - Scienti®c and Technological Advances.
Symp Proc 287, Mat Res Soc, Pittsburgh, p 461
428. Lin CKJ, Jenkins MG, Ferber MK (1993) Evaluation of Tensile Static, Dynamic and
Cyclic Fatigue Behaviour for a HIPed Silicon Nitride at Elevated Temperatures. In:
Chen IW, Becher PF, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics ±
Scienti®c and Technological Advances. Symp Proc 287, Mat Res Soc, Pittsburgh,
p 455
429. Herrmann M, Klemm H, Schubert C, Hermel W (1997) Long Term Behaviour of
SiC/Si3N4-Nanocomposites at 1400±1500 °C. In: Baxter J, Fordham LCR, Gabis V,
Hellot Y, Lefebvre M, Le Doussal H, Sech AL (eds) Euro Ceramics V Key Eng Mat
132±136. Trans Tech Publications, Switzerland, p 1997
430. Jack KH (1993) SIALON CERAMICS: Retrospect and Prospect. In: Chen IW, Becher PF,
Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride Ceramics ± Scienti®c and
Technological Advances. Mat Res Soc Symp Proc 287. Materials Research Soc,
Pittsburgh, p 15
431. Opila EJ, Jacobson NS (2000) Corrosion of ceramics. In: SchuÈtze M (ed) Corrosion and
Environmental Degradation. Wiley-VCH, Weinheim, p 329
432. Gogotsi YG, Lavrenko VA (1992) Corrosion of High Performance Ceramics. Springer,
Berlin, Heidelberg, New York
433. Jacobson NS (1993) J Am Ceram Soc 76: 3
434. Nickel KG (ed) (1994) Corrosion of Advanced Ceramics ± Measurement and Modelling.
Kluwer Academic Publishers, Dordrecht
435. Fordham RJ, Baxter DJ, Graziani T (eds) (1996) Corrosion of Advanced Ceramics, Key
Eng Mater 113. Trans Tech Publications, Switzerland
436. Nickel KG, Gogotsi YG (2000) Corrosion of Hard Materials. In: Riedel R (ed) Handbook
of Ceramic Hard Materials. Wiley-VCH, Weinheim, p 140
437. Telle R, Quirnbach P (eds) (1994) Korrosion und Verschleiû von Keramischen
Werkstoffen: Kombinierte BelastungsfaÈlle als Anwendungsgrenze. DGM Information-
sgesellschaft, Oberursel
438. Richter HJ, Kaiser A, Taut C, Herrmann M (1994) AnwendungsmoÈglichkeiten und
Grenzen von Thermodynamischen Berechnungen bei der Interpretation und Voraus-
sage von Korrosionsprozessen. In: Telle R, Quirnbach P (eds) Korrosion und
Verschleiû von Keramischen Werkstoffen: Kombinierte BelastungsfaÈlle als
Anwendungsgrenze. DGM Informationsgesellschaft, Oberursel, p 322
439. Wagner C (1958) J Appl Physics 29: 1295
440. Herrmann M, Schober R (1989) Silikattechnik 40: 260
441. Opila EJ, Smialek JL, Robinson RC, Fox DS, Jacobson NS (1999) J Am Ceram Soc 82:
1826
442. Filsinger D, Schulz A, Wiitig S, Klemm H, Taut C, WoÈtting G (2000) Model Cobustor to
Assess the Oxidation Behavior of Ceramic Materials under Engine Conditions. In:
Nelson HD (ed) Internat Gas Turbine & Aeroengine Congress & Exhibition 2000:
Transactions of the ASME, ASME, New York, p 2000-GT-349
443. Newson DD, Pollinger JP, Twait DJ (2000) Internat Gas Turbine & Aeroengine
Congress & Exhibition 2000: Transactions of the ASME, New York, p 2000-GT-533
444. Yokoyama K, Wada S (2000) J Ceram Soc Jpn 108: 627
445. Klemm H, Herrmann M, Schubert C (2000) J Eng. Gas Turbines and Power 122: 13
446. Klemm H, Herrmann M, Schubert C (1996) The In¯uence of the Grain Boundary Phase
Composition on the High Temperature Properties of Silicon Nitride Materials. In:
Parilak L, Danninger H, Dusza J, Weiss B (eds) Proc Int Conf Deformation and
Fracture in Structural PM Materials, IMR SAS Kosice 2: 75
447. Andrews P, Riley FL (1991) J Eur Ceram Soc 7: 125
448. O'Meara C, SjoÈdberg J (1997) J Am Ceram Soc 80: 1491
449. Ricoult-Barkhausen M, Gogotsi YG (1996) Identi®cation of Oxidation Mechanisms in
Silicon Nitride Ceramics by TEM. In: Fordham PJ, Baxter DJ, Graziani T (eds)
Corrosion of Advanced Ceramics. Key Eng Mat 113. Trans Tech Publications,
Switzerland, p 81
450. Luthra KL (1994) Theoretical Aspects of the Oxidation of Silica-Forming Ceramics.
In: Nickel KG (ed) Corrosion of Advanced Ceramics ± Measurement and Modelling.
451. Luthra KL (1991) J Am Ceram Soc 74: 1095
452. Ogbuji LUJT (1994) The Oxidation Process in Silicon Nitride. In: Nickel KG (ed)
Corrosion of Advanced Ceramics ± Measurement and Modelling. Kluwer Academic
Publishers, Dordrecht, p 117
453. Ogbuji LUJT, Bryan SR (1995) J Am Ceram Soc 78: 1272
454. Ogbuji LUJT, Opila EJ (1995) J Electrochem Soc 142: 925
455. Taut C (1994) PhD thesis, University of Dresden
456. Tressler RE (1990) Environmental Effects on Long Therm Reliability of SiC and Si3N4.
In: Tressler RE, McNallan M (eds) Ceramic Transactions 10, Am Ceram Soc,
Westerville, OH, p 99
457. Van der Biest O, Weber C (1994) In¯uence of Oxidation on Long Term Reliability of
Silicon Nitride. In: Nickel KG (ed) Corrosion of Advanced Ceramics ± Measurement
and Modelling. Kluwer Academic Publishers, Dordrecht, p 453
458. Nishimura N, Masuo E, Takita K (1991) Effect of Microstructural Oxidation on the
Strength of Silicon Nitride after High Temperature Exposure. In: Carlsson T, Johansson
T, Kahlman T (eds) Proc 4th Int Symp Ceram Mater & Components for Engines.
Elsevier, London, p 1139
459. Mukundhan P, Wu P, Du HH (1999) J Am Ceram Soc 82: 226
460. Herrmann M, Klemm H, GoÈbel B, Schubert C, Hermel W (1999) SiC/Si3N4
Nanocomposites with Excellent High Temperature Long-Term Behaviour. In: Niihara,
Sekino, Yasuda E, Sasa T (eds) The Science of Engineering Ceramics II, Key Eng Mat
161±163. Trans Tech Publications, Switzerland, p 377
461. Watanabe M, Shimamori T, Noda Y (1992) The High Temperature Properties of Sc2O3
doped Si3N4. In: Ishizaki K (ed) Grain Boundary Controlled Properties of Fine
Ceramics, Elsevier, London, p 199
462. Nordberg LO, KaÈll PO, Nygren M (1996) A Mathematical Analysis of the Non-Parabolic
Oxidation Behaviour of b-SiALON Matrices and Composites. In: Fordham PJ, Baxter
DJ, Graziani T (eds) Corrosion of Advanced Ceramics, Key Eng Mat 113. Trans Tech
463. Patel JK, Thompson DP (1988) Br Ceram Trans J. 87: 70
464. Baxter DJ, Graziani T, Wang HM, McCauley RA ( 1998) J Eur Ceram Soc 18: 2323
465. Ernstberger U (1985) PhD Thesis, University of Karlsruhe
466. Ernstberger U, Grathwohl G, ThuÈmmler F (1987) Int J High Techn Ceram 3: 43
467. Babini GN, Belosi A, Vincenzini P (1983) J Mater Sci 18: 231
468. Loehman RE (1989) Am Ceram Soc Bull 68: 891
469. Loehman RE (1999) Key Engin Mat 657: 161±163
470. ToÈnshoff HK, Denkena B (1991) J Society of Tribologists and Lubrication Engineers 47:
772
471. Miao H, Qi, L, Cui G (1995) Silicon Nitride Ceramic Cutting-Tools and their
Applications. In: Low IM, Li XS (eds) Advanced Ceramic Tools for Machining
Application ± II Key Engineering Materials 114, Trans Tech Publications Ltd,
Switzerland, p 135
472. Voitovich RF (1971) Tugoplavkie Soedineniya, termoddinamicheskie Kharakteriustiki,
Naukova Duma, Kiev
473. Samsonov GV, Vinitskii IM (1980) Handbook of Refractory Compounds, Plenum, New
York, p 400
474. Feld G (1969) Sprechsaal 102: 1098
475. Watanabe M, Usami T, Takasu S, Matsuo S, Toji E (1983) Jap J Appl Phys 1 22: 185
476. Li JG, Hausner H (1992) J Eur Ceram Soc 9: 101
477. Li JG, Hausner H (1990) Benetzung von anorganischen Materialien durch Siliciumsch-
melzen, Abschluûbericht, TU Berlin, p 37
478. Swartz JC (1976) J Am Ceram Soc 59: 272
479. Singh RN, Tuohig WD (1975) J Am Ceram Soc 58: 70
480. Smith PL, White J (1983) Trans J Br Ceram Soc 82: 23
481. Belyi VI (1988) Mater Sci Mnogr 34: 126
482. David J (1972) Rev Chim Miner 9: 717
483. Leimer G, Gugel E (1975) Z Metallkd 66: 570
484. Panacjuk AD, Fomenko VS, Glebova GG (1986) Stoikost nemetallischeskich materialov
b rasplavach. Naukova Dumka, Kiev, p 175
485. Jelacic C, Dervisbegovic H (1974) Bull Soc Fr Ceram 105: 17
486. Felten RP (1974) Sprechsaal 107: 92, 101
487. Naka M (1991) Ultramicroscopy 39: 128
488. Naka M, Kubo M, Mori H, Okamoto I (1990) Koon Gakkaishi 16: 225
489. Gilde W (1953) Metall Giessereitech 3: 324
490. Schwabe U, Wolff L, Ziegler G (1988) Klei/Glas/Keram 9: 231
491. Feld H, Gugel E, Nitzche HG (1969) Werkst Korros 20: 571

492. Ljungberg L, Warren R (1989) Ceram Eng Sci Proc 10: 1655
493. Brunken R (1970) Werkstatt Betr 103(1): 65
494. Schuster JC, Weitzer F, Bauer J, Nowotny H (1988) Mater Sci Eng A 105/106: 201
495. Weitzer F, Schuster JC, Bauer J, Jounel B (1991) J Mater Sci 26: 2076
496. Naidich YV, Zhuravlev VS, Frumina NI, Kostyuk BD, Krasovskaya NA Ostrovskii VG
(1988) Poroshk Metall 11: 58
497. Kunz KP, Sarin VK, Davis RF Bryan SR (1988) Mater Sci Eng A 105/106: 47
498. Dummler W, Weber S, Tete C, Scherrer S (1999) J Mater Sci Lett 18: 193
499. Akselsen OM (1992) J Mat Sci 27: 1989
500. Raic KT (1999) Ceram Int 26/1: 19
501. Shimoo T, Okamura K, Shibata D (2000) J Mater Sci 35/21: 5485
502. Schuster JC (1988) J Mat Sci 23: 2792
503. Glemser O, Beltz K, Naumann P (1957) Z Anorg Allg Chem 291: 51
504. Shimoo T, Okamura K, Yamasaki T (1999) J Mater Sci 34: 5525
505. Weitzer F, Schuster JC (1987) J Solid State Chem 70: 178
506. Komeya K, Meguro T, Atago S, Lin CH, Abe Y, Komatsu M (1999) Corrosion Resistance
of Silicon Nitride Ceramics. In: Niihara K, Sekino T, Yasuda E, Sasa T (eds) The Science
of Engineering Ceramics II. Key Eng Mat 161±163, Trans Tech Publications,
Switzerland, p 235
507. Hollstein T, Graas T, Bundschuh K, SchuÈtze M (1998) Keram Zeitschrift 50: 416
508. Sato T, Tokunaga Y, Endo T, Shimada M, Komeya K (1988) J Mat Sci 23: 3440
509. Iio S, Okada A, Akira A, Tetsuo A, Masahiro Y (1992) J Ceram Soc Jpn Int Edition 100:
954
510. Herrmann M, Schubert C, Michael G (1999) Korrosionsstabile keramische Werkstoffe
fuÈr Anwendungen in WaÈlzlagern und im Anlagenbau. In: c® DKG Berichte 14: 130
511. Herrmann M, Michael G (1999) Br Ceram Proc 60/1: 455
512. Okada A, Yoshimura M (1996) Mechanical Degradation of Silicon Nitride Ceramics in
Corrosive Solutions of Boiling Sulphuric Acid. In: Fordham RJ, Baxter DJ, Graziani T
(eds) Corrosion of Advanced Ceramics, Key Eng Mater 113. Trans Tech Publications,
Switzerland, p 227
513. Gogozi YG, Lavrenko VA (1992) Springer Verlag, Berlin, p 76
514. Shimada M, Sato T (1989) Ceram Trans 10: 355
515. Fang Q, Sidky PS, Hocking MG (1997) Corr Sci 39: 511
516. Kanbara K, Uchida N, Uematsu K, Kurita T, Yoshimoto K, Suzuki Y (1993) Corrosion
of Silicon Nitride Ceramics by Nitric Acid. In: Chen IW, Becher PF, Mitomo M, Petzow
G, Yen TS (eds) Silicon Nitride Ceramics, Mat Res Soc Symp Proc 287. Mat Res Soc,
Pittsburgh, p 533
517. WoÈtting G, Herrmann M, Michael G, Siegel S, Frassek L (1999) WO99/20579
518. Sharkawy SW, El-Aslabi AM (1998) Corros Sci 40: 1119
519. Sato T, Tokunaga Y, Endo T, Shimada M, Komeya K, Nishida K, Komatsu M, Kameda T
(1988) J Mat Sci 23: 3440
520. Sato T, Murakami T, Shimada E, Komeya K, Komeda T, Komatsu M (1991) J Mat Sci
26: 1749
521. Yoshimura M, Yamamoto S (1995) Seramikkusu 30: 995
522. Oda K, Yoshio T, Miyamoto Y, Koizumi M (1993) J Am Ceram Soc 76: 1365
523. Hollstein T (1999) Adv Sci Technol 13: 433
524. Sato T, Sato S, Tamura K, Okuwaki A (1992) J Br Ceram Trans 91: 117
525. Herrmann M, Michael G, Schubert C, Adler J, Krell A (1998) Final report MATECH
project 03N2002 F
526. Ye J, Furuya K, Misono Y, Matsuo K, Munakata F, Ishikawa I, Akimune Y (2000)
J Luminescence 574: 87±89
527. Shen Z, Nygren M, Halenius U (1997) J Mat Sci Lett 16: 263
528. Pechenik A, Piermarini GJ (1992) J Am Ceram Soc 75: 3283
529. Rouxel T, Piriou B (1996) J Appl Phys 79: 9074
530. Keûler S (1993) PhD Thesis, University of Stuttgart

531. Clarke DR (1989) Mat Sci. Forum 47: 110
532. Yamakawa A, Miyake M, Ishizaki K (1994) J Ceram Soc Jpn 102: 339
533. Lewis HM (1993) Sialons and Silicon Nitrides; Microstructural Design and Perfor-
mance. In: Chen IW, Becher PF, Mitomo M, Petzow G, Yen TS (eds) Silicon Nitride
Ceramics ± Scienti®c and Technological Advances, Mat Res Soc Symp Proc 287, MRS,
Pittsburgh, p 159
534. Kessler H, Herrmann M, Pompe W (1995) Acta metall mater 43: 2789
535. Keûler H (1995) PhD thesis, Technical University of Dresden
536. Carborundum Co. (1952) US Patent No 2618565
537. Moulson AJ (1979) J Mat Sci 14: 1017.
538. Heinrich JG, KruÈner H (1994) Silicon Nitride Materials for Engine Applications. In:
Hoffmann MJ, Petzow G (eds) Tailoring of Mechanical Properties of Si3N4 Ceramics.
NATO ASI Series E, Vol 276, Kluwer Academic Publishers, Dordrecht, p 19
539. Ziegler G (1985) Z Werkstofftechnik 16: 12, 45, 81
540. Mikijeli B, Mangels J (2001) SRBSN Material Development For Automotive Applica-
tions. In: Heinrich JG, Aldinger F (eds) Ceramic Materials Components for Engines,
541. Riley FF (1983) Silicon Nitridation. In: Riley FF (ed) Progress in Nitrogen Ceramics.
Martinus Niihoff Publishers, Dordrecht, p 121
542. Jennings HM, Richmann MH (1976) J Mat Sci 11: 285
543. Ziegler G, Heinrich JG, WoÈtting G (1987) J Mat Sci 22: 3041
544. Rahaman MN, Moulson AJ (1984) J Mat Sci 19: 189
545. Jennings M (1983) J Mat Sci 18: 951
546. Atkinson A, Moulson AJ, Roberts EW (1976) J Am Ceram Soc 59: 285
547. Rossetty GA, Denkewicz RP (1989) J Mat Sci 24: 3081
548. Li WB, Lei BQ, Lindback T (1997) J Eur Ceram Soc 17: 1119
549. Campos-Loriz D, Riley FL (1978) J Mat Sci 13: 1125
550. Sheldon BW, Rankin J, Haggerty JS (1995) J Am Ceram Soc 78: 1624
551. Mangels JA, Tennenhouse GJ (1980) Am Ceram Soc Bull 59: 1216
552. Herrmann M, Heû S, Pabst S, Richter HJ, Hermel W (1990) Keram Zeitschr 42: 250
553. WoÈtting G (1999) Final report BMBF- project 03N2002B
554. Liu H, Hsu SM (1996) J Am Ceram Soc 79: 2452
555. Schober R, Richter HJ (1999) Composite ceramics based on silicon nitride or
aluminium nitride. In: Krams J (ed) Proceedings Biennial Worldwide Congress
UNITECR `99, Verlag Stahleisen, DuÈsseldorf, p 198
556. Isomura K, (1990) J Am Ceram Soc 73: 624
557. Petrovsky VY, Rak ZS (2001) J Eur Ceram Soc 21: 219
558. Petrovsky VY, Rak ZS (2001) J Eur Ceram Soc 21: 237
559. Gogotsi YG (1994) J Mat Sci 29: 2541
560. Sigulinski F, Boskovic S (1999) Ceram Intern 25: 41
561. Liu CC (2000) J Ceram Soc Jpn 108: 46
562. Yamada K, Kamiya N (1999) Mat Sci Eng 261: 270
563. Herrmann M, Bales A, Bach E, Leuteritz U (unpublished)
564. Kawamura H (1999) New Perspectives in Engine Applications of Engineering Ceramics.
In: Niihara K, Sekino T, Yasuda E, Sasa T (eds) The Science of Engineering Ceramics II,
Key Eng Mat 161±163. Trans Tech Publications, Switzerland, p 9
565. Veprek S (2000) Nanostructured Superhard Materials. In: Riedel R (ed) Handbook of
Ceramic Hard Materials. Wiley VCH, Weinheim, p 104
566. Pezzotti G, Tanaka I, Ikuhara Y, Sakai M, Nishida T (1994) Scripta Met Et Mat 31: 403
567. Seifert HJ (1993) PhD Thesis, University of Stuttgart
568. Herrmann M, Balzer B, Schubert C, Hermel W(1993) J Eur Ceram Soc 12: 287
569. Huang JL, Chiu HL, Lee MT (1994) J Am Ceram Soc 77: 705
570. Herrmann M, Schubert C, Hermel W, Meiûner E, Ziegler G (1996) c®/Ber DKG 73: 434
571. Fricke M, Nonninger R, Schmidt H (2000) Adv Eng Mat 2(10): 647
572. Lee BT, Yoon YJ, Lee KH (2001) Mater Lett 47: 71
573. Wang CM (1995) J Mater Sci 30: 3222
574. Iwamoto Y, Kikuta K, Hirano S (2000) J Ceram Soc Jpn 108(4): 350 and 1072
575. Nagaoka T, Yasuoka M, Hirao K, Kanzaki S (1992) J Ceram Soc Jpn 100: 612
576. Bellosi A, Guicciardi S, Tampieri A (1992) J Eur Ceram Soc 9: 83
577. Liu CC, Huang JL (2000) Brit Ceram Trans 99: 149
578. Akimune Y, Munakata F, Hirosaki N, Okamoto Y (1998) J Ceram Soc Jpn 106: 75
579. Opsommer A, Gomez E, Castro F (1998) J Mater Sci 33: 2583
580. Woydt M, Skopp A, Habig KH (1991) Wear 148: 377
581. Skopp A (1993) PhD Thesis, University of Berlin
582. Xu J, Kato K (1997) Wear 202: 165
583. Popp M, Sternagel R, Pfeifer W, Blug B, Meier S, WoÈtting G, Frasseck L (2000) Hybrid-
and Ceramic Rolling Bearings with Modi®ed Surface and Low friction Rolling Contact.
In: MuÈller G (ed) Ceramics ± Processing, Reliability, Triboloy and Wear, Euromat 99
Vol 12. Wiley VCH, Weinheim, p 449
584. Lences Z, Sajgalik P, Toriyama M (2000) J Eur Ceram Soc 20: 347
585. Yeh CH, Hon MH (1996) Ceram Intern 23: 361
587. Herrmann M, Schubert C, Klemm H (1996) Nanocompositkeramik: U È bersicht uÈber
Werkstoffkonzepte, Herstellungstechnologien und spezi®sche Eigenschaften. In:
Kriegesmann J (ed) Technische Keramische Werkstoffe, Deutscher Wirtschaftsdienst,
KoÈln, Chap 4.4.3.0.
588. Greil P, Petzow G, Tanaka H (1987) Ceram Int 13: 19
589. Hockey BJ, Widerhorn SM, Liu W, Balfdoni G, Buljan ST (1991) J Mater Sci 26: 3931
590. Kodama H, Sakamoto H, Miyoshi T (1989) J Am Ceram Soc 72: 551
591. Laughner JW, Bhatt RT (1989) J Am Ceram Soc 72: 2017
592. Becher PF, Hsueh CH, Angelini P, Tiegs TN (1988) J Am Ceram Soc 71: 1050
593. Campbell GH, RuÈhle M, Dalgleish BJ, Evans AG (1990) J Am Ceram Soc 73: 521
594. Birchall JD, Stanley DR, Mockford MJ, Pigott GH, Pinto PJ (1988) J Mat Sci Lett 7: 350
595. Pezzoti G (1993) J Am Ceram Soc 76: 1313
596. Rendtel A, HuÈbner H, Schubert C (1995) Silicates Industriels 60: 305
597. Niihara K, Izaki K, Kawakami T (1990) J Mat Sci Lett 10: 112
598. Sternitzke M (1997) J Eur Ceram Soc 17: 1061
599. Izaki K, Hakkei K, Ando K, Kawakami KT, Niihara K (1988) Fabrication and
Mechanical Properties of Si3N4/SiC Composites from Fine, Amorphous Si-C-N- Powder
Precursors. In: Mackenzie D (ed) Ultrastructural Processing of Advanced Ceramics,
John Wiley & Sons Inc, New York, p 891
600. Hirano T, Niihara K (1995) Mater Lett 22: 249; (1996) 26: 285
601. Lavedrine A, Bahloul D, Coursat P (1991) J Eur Ceram Soc 8: 221
602. Palchevskis E, Grabis J, Millers T (1993) Fine Si3N4-SiC powders; Obtaining and
Characteristics. In: Aldinger F (ed) PTM `93, DGM Informationsgesellschaft, Oberursel,
p 657
603. Cauchetier M, Croix O, Luce M (1991) J Eur Ceram Soc 8: 215
604. Westerheide R, WoÈting G, Schmitz HW (1998) SiC-Si3N4 Nanocomposite- Potential
und wirtschaftliche Realisierung. In: Kriegesmann I (ed) Keramische Werkstoffe 84erg.
-lfg. Deutscher Wirtschaftsdienst, KoÈln, Chap 4.4.3.2
605. WoÈtting G, Caspers B, Gugel E, Westerheide R (2000) Transactions of the ASME 122: 8
606. Ishizaki K, Yanai T (1995) Silicates Industriels 60: 215
607. Sajgalik P, Hnatko M, Lences Z, Warbichler P, Hofer F (2001) Z Metallkd 92: 937
608. Riedel R, Klebe HJ, SchoÈnfelder H, Aldinger F (1995) Nature 374: 526
609. Riedel R, Seher M, Mayer J, Szabo DV (1995) J Eur Ceram Soc 15: 703 and 717
610. Haluschka C, Kleebe HJ, Franke R, Riedel R (2000) J Eur Ceram Soc 20: 1355
611. Haluschka C, Engel C, Riedel R (2000) J Eur Ceram Soc 20: 1365
612. Sajgalik P, Hnatko M, Lofai F, Hvizdos P, Dusza J, Warbichler P, Hofer F, Riedel R,
Lecomte E, Hoffmann MJ (2000) J Eur Ceram Soc 20: 453
613. Ukyo Y, Kandori T, Wada S (1993) J Ceram Soc Jpn Int Ed 101: 1398
614. Niihara K, Suganuma K, Nikahira K, Izaki K (1990) J Mater Sci Lett 9: 598
615. Pan X, Mayer J, RuÈhle M, Niihara K (1994) J Am Ceram Soc 77: 3039
616. Cheong DS, Hwang KT, Kim CS (1999) J Am Ceram Soc 82: 981
617. Rouxel T, Wakai F (1992) J Am Ceram Soc 75: 2363
618. Rendtel A, HuÈbner H, Herrmann M, Schubert C (1998) J Am Ceram Soc 81: 1109
619. Ramoul-Badabache K, Lancin M (1992) J Eur Ceram Soc 10: 369
620. Klemm H, Tangermann K, Reich T, Herrmann M, Schubert C, Hermel W (1995) High-
Temperature Properties of Silicon Nitride Molybdenum Silicide Composites. In: Bellosi
A (ed) Fourth Europ Ceramics, Vol 4, Grupp Editoriale faenca Editrice, Faenza, p 233
621. Pezotti G, Sakai M (1994) J Am Ceram Soc 77: 3039
622. Rendtel A, HuÈbner H, Herrmann M (1995) Creep Behaviour of Si3N4/SiC-Nanocom-
posite Materials. In: Bellosi A (ed) Fourth Europ Ceramics Vol 4, Grupp Editoriale
faenca Editrice S p.A, Faenza, p 225
623. Klebe HJ, Pezzotti G, RuÈhle M (1998) J Ceram Soc Jpn 106: 17
624. Rendtel P, Rendtel A HuÈbner H, Herrmann M (1999) J Eur Ceram Soc 19: 217
625. Gommes JR, Osendi MI, Miranzo P, Oliveira FJ, Silva RF (1999) Wear 222: 233±235
626. Sawagushi A, Toda K, Niihara K (1991) J Am Ceram Soc 74: 1142
627. Kanzaki S, Brito ME, Valecillos MC, Hirao K, Toriyama M (1997) J Eur Ceram Soc 17:
1841
628. Akimune Y, Munakata F, Matsuo K, Okamoto Y, Hirosaki N, Satoh C (1999) J Ceram
Soc Jpn 107: 1180
629. Imamura H, Hirao K, Brito ME, Toriyama M, Kanzaki S (2000) J Am Ceram Soc 83: 495
630. Hirao K, Nagaoka T, Brito ME, Kanzaki S (1994) J Am Ceram Soc 77: 1857
631. Teshima H, Hirao K, Toriyama M, Kanzaki S (1999) J Ceram Soc Jpn 107: 1216
632. Hirao K, Ohashi M, Brito ME, Kanzaki S (1995) J Am Ceram Soc 78: 1687
633. Lee SY, Amoako-Appiagyei K, Kim HD (1999) J Mater Res 14: 178
634. Hirosaki N, Okamoto Y, Munakata F, Akimune Y (1999) J Eur Ceram Soc 19: 2183
635. Kitayama M, Hirao K, Toriyama M, Kanzaki S (2000) J Ceram Soc Jpn 108: 646
636. Jack KH (2000) Nitrogen Ceramics for Engine Applications. In: Hampshire S, Pomoroy
MJ (eds) Nitrides and Oxynitrides, Mat Sci Forum 325±326. Trans Tech Publications,
Switzerland, p 255
637. Katz NR (1997) Industrial Ceram 17: 158
638. Savitz M (1999) Am Bull Ceram Soc 78 No. 1: 53
639. Brandt G (2001) Ceramic Cutting Tools. In: Heinrich JG, Aldinger F (eds) Ceramic
Materials and Components for Engines. Wiley-VCH, Weinheim, p 21
640. Bayer O, Streit E (2000) Keram Zeit 52: 1092
641. Cundill RT (1997) Impact Resistance of Silicon Nitride Balls. In: Niihara K, Hirano S,
Kanzaki S, Komeya K, Morinaga K (eds) Ceramic Materials and Components for
Engines. Jap Fine Ceramic Assoc, Tokyo, p 556
642. Wagemann A (2000) Keram Zeit 52: 503
643. Guangchuan L, Jinsheng L, Xingua G (1999) Industrial Ceram 19: 17
644. Gugel E, WoÈtting G (1999) Industrial Ceram 19: 196
645. Woetting G, Frassek L, Leimer G, SchoÈnfelder L (1993) c®/Ber DKG 70: 287
646. Mandler WF (2001) Commercial Applications for Advanced Ceramics in Diesel
Engines. In: Singh M, Jessen T (eds) 25th Annual Conference on Composites, Advanced
Ceramics, Materials, and Structures: A (Ceram Eng Sci Proc 22). Am Ceram Soc,
Westerville, OH, p 3
647. Mandler WF (1997) Ceramic Successes in Diesel Engines. In: Niihara K, Hirano S,
Kanzaki S, Komeya K, Morinaga K (eds) Ceramic Materials and Components for
Engines. Jap Fine Ceramic Assoc, Tokyo, p 137
648. Kamo R, Mavinahally NS, Kamo L, Bryzik W, Reid M (1997) Experimental Heat Release
of Insulated Turbocharged Diesel Engine. In: Niihara K, Hirano S, Kanzaki S, Komeya
K, Morinaga K (eds) Ceramic Materials and Components for Engines. Jap Fine Ceramic
Assoc, Tokyo, p 146
649. WoÈtting G, Hennicke J, Feuer H, Thiemann KH, Vollmer D, Fechter E, Sticher F, Geyer
A (2001) Reliability and Reproducibility of Silicon Nitride Valves: Experiences of a
Field Test. In: Heinrich JG, Aldinger F (eds) Ceramic Materials and Components for
Engines. Wiley VCH, Weinheim, p 181
650. Savitz M (1999) Bull Am Ceram Soc 78 No. 3: 52
651. Meiser M (2001) Current Status of Structural Ceramics. Paper presented at the 25th
Annual Conference on Composites, Advanced Ceramics, Materials, and Structures. Am
Ceram Soc, Cocoa Beach, Fl
652. Matsuura T, Kawai C, Yamakawa A (1997) Sumitomo Electric Technical Review 43: 77
653. Kawai C, Matsuura T, Yamakawa A (1999) J Mater Sci 43: 893
654. Kawamura H (2001) Practical use of Ceramic Components and Ceramic Engines.
In: Heinrich JG, Aldinger F (eds) Ceramic Materials and Components for Engines.
655. Yoshida M, Tanaka K, Tsuruzono S, Tatsumi T (1999) Industrial Ceram 19: 188
656. Der Kochherd der Zukunft arbeitet mit Si3N4-Platten (1999) In: c®/Ber DKG, 76, No.4:
D18-20
657. Aberle AG (2001) Solar Energy Materials & Solar Cells 65: 239
658. Tsukumura K, Kamo K (2000) J Ceram Soc Jpn 108: 882
Subject Index
Acid, boric 15 Densification 51,79-96,101-102,136-141,

Acid corrosion, silicon nitride ceramics 147
125-129 Devitrification 83,87,96-101,105-106,
114-115,132,141
Dewaxing (burnout) 80,82
Battery 22 Diamond 4,7,11,27,28,32-38
Bias 28,31-33 Diimide synthesis 73-76
Binder 33-37 DTA 9,10
Borazon (c-BN) 7
Boric acid 15
Boron nitride, cubic (c-BN) 7,14ff Elastic constants, Si3N4 modifications 55
- -, hexagonal (h-BN) 6,14ff Elbor 7,35
- -, polycrystalline c-BN (PcBN) 24,34, Electrical properties, Si3N4 ceramics
35 106-107,137-139,142
- -, Wurzite (w-BN) 7,14ff Enthalpy of formation, Si3N4 modifications
Boron nitrides, applications 4,15-22,25, 55
33,34 Equivalent percent 62,65,67
- -, -, electronic 18,22,38 Eutectic 24,25
- - , - , high temperature 14,20 Exaggerated grain growth 94-95,141-143
- -, -, wear 34
Break rings 18,19 Fibers 17,21,22
Fuel cell 22
Carbothermal nitridation 72-75
Catalysts 10,16,22-25,34 GPSN 52,83-85,90,113,132,136-137,
Coatings 4,17,20-23,28-38 141,147
Conductivity, electrical 20,22 Grinding wheels 37,38
-.thermal 7,18,20,22,34,38
Continuous casting 18,19
Conversion 8,9,36 Hardmetal 33,35,36
-, catalytic 24 Heterodiamond 27
-, direct 23,34 High-pressure high temperature 4,7,23,
-, dynamic 27 27,35
HIP-BN 18
Conversion temperature, DTA 9,10
HIPSN 52,83-85,90,113,119,132,147
Creep 105-119,136,141,142,147
HP 18
Crucibles, evaporator 18,20-22
HPSN 51-52,83-85,90,110,113,119,132,
Crystal structure, ass (a-SiAlON)
140,143,147
69-70
- -, Si3N4 53,55
Cubonite 7 Insulator 6,7,14,18,19,38
CVD (chemical vapour deposition) 10,16, Ion-bombardment 9,29,30
17,21,22,27-32 IR 9,12-14
170 Subject Index
Laser-induced synthesis 72,75 , /?ss (0-SiAlON) 52,58-71,83,88,92,

Lattice parameter, Si3N4 solid solutions 98-108,115,138,147
66,70 , yss (c-Si3N4) 65
Li-catalyst 10 , colour 129-132
Liquid phase sintering 80-96,101-102, .composites 73,91,114-120,
130-132,136 137-142
Lubricant 4,14-18 - - -, corrosion in acids/bases 125-129
.creep 105-119,136,141,142,147
.decomposition 60-61,88-89,109,
Metallization 20 137,140
Microstructural evaluation 103-104 - - -, electrical properties 106-107,
Mold 18,21,37 137-142
- - -, fracture toughness 84,102-114,
136-143
Nano-cBN 17,27-33 .grain growth 93-96,102,
Nanocomposites 73,91,115,140-141 141-142
Neutron absorbers 18 .hardness 90,102-112,133-137,148
Nitridation, direct 72,74,76 - - -, hydrothermal conditions 126,
Nozzles 19 127,130
Nuclear fuel pellets 17 , microstructure 72,77-82,90-91,
104,110,112,133-139,147
.oxidation 105-121,136-142
Oxide nitride 53,56,60-69,83-102,108, - - -, phase diagrams 57-70
112-121,130-140 - - -, phase transformation 71,83,
92-93,107
, seeded materials 94,102,110,
Phase diagrams 8,11,57-71 137-138,142-143
- -, BN 8,11 .sintering 51,61,65,78,80-121,
Phase transformation 57,71,83,92-93,107 126-142,147
Plasma CVD 14,23,32 - - -.snowflakes 132
Plasmachemical synthesis 72,74,76 , solid solutions 52,58-92,98-108,
Polycrystalline c-BN (PcBN) 24,34,35 115,138,147
Precursor derived ceramics 91,137,140 , subcritical crack growth 108-115,
PVD 14,17,23,27,28,30 126
Pyrolysis, silicon organic compounds 72, , thermal conductivity 107-108
75,91,140 - - -, thermal expansion 108
.types 51-52,83-85,90
Silicon nitride powder 71-82,92-96,107,
R-curve 106,110,112 134-135
Raman spectroscopy 12-14,27 Sliding part 20,21
RBSN 52,84-85,133-136,147 Snow flakes 132
Rubber, BN 18,22 Sol-gel technique 82
Solution-reprecipitation 71,83,86,92-94,
113
Selective etching 27,28,33 Source holder 19
Self-propagating high temperature synthesis Specific heat, Si3N4 ceramics 107
72,74,76 SRBSN 52,84-85,132,136-137,147
Si3N4 ceramics, applications 51-52,63, SSN 52,83-85,90,136,147
79-95,90,108-148 Stable phase 10
SiAlON 19 Steric hindrance 92-95
Silicon nitride 50ff
Silicon nitride ceramics, aspect ratio
92-106,110,142 Temperature gradient 26,27
, ass (a-SiAlON) 52,58-71,83,88, Thermal conductivity, Si3N4 55,107
101-108,115,147 Thermal diffusivity, Si3N4 ceramics 107
Subject Index 171
Thermal expansion, Si3N4 55,108 Welding tip 19

Thermal shock resistance 106,112,138 Whisker 50,92,109,110,134,142
Toner 16 White graphite 6
Van der Waals bonding 6 X-ray diffraction 12,14,24

Vapour phase synthesis 72-75
Author Index Volumes 101-102
Aldinger F, see Seifert HJ (2002) 101:1-58
Fruhauf S, see Roewer G (2002) 10J:59-136
Haubner R, Wilhelm M, Weissenbacher R, Lux B (2002) Boron Nitrides - Properties,

Synthesis and Applications. 102:1-46
Herrmann M, see Petzow G (2002) 102:47-166
Herzog U, see Roewer G (2002) 101:59-136
Jansen M, Jaschke B, Jaschke T (2002) Amorphous Multinary Ceramics in the Si-B-N-C

System. 101:137-192
Jaschke B, see Jansen M (2002) 101:137-192
Jaschke T, see Jansen M (2002) 101:137-192
Lux B, see Haubner R (2002) 102:1-46
Muller E, see Roewer G (2002) 101:59-136
Petzow G, Hermann M (2002) Silicon Nitride Ceramics. 102:47-166
Roewer G, Herzog U, Trommer K, Muller E, Fruhauf S (2002) Silicon Carbide - A Survey

of Synthetic Approaches, Properties and Applications. 101:59-136
Seifert HJ, Aldinger F (2002) Phase Equilibria in the Si-B-C-N System. 101:1-58
Trommer K, see Roewer G (2002) 101:59-136
Weissenbacher R, see Haubner R (2002) 102:1-46

Wilhelm M, see Haubner R (2002) 102:1-46

High Performance Non-Oxide Ceramics II

Uploaded by

Copyright:

Available Formats

High Performance Non-Oxide Ceramics II

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

High Performance Non-Oxide Ceramics II

Uploaded by

Copyright:

Available Formats

102

Structure and Bonding

Volume Editor: M. Jansen

Springer WWW home page: http://www.springer.de

CIP Data applied for

Springer-Verlag Berlin Heidelberg New York a member of BertelsmannSpringer

Typesetting: Scientific Publishing Services (P) Ltd, Madras

Professor D. Michael P. Mingos Professor Dr. Martin Jansen

Prof. Allen J. Bard Prof. Ian G. Dance

Prof. Peter Day, FRS Prof. James A. Ibers

Boron Nitrides - Properties, Synthesis and Applications

Silicon Nitride Ceramics

Author Index Volumes 101-102 169

Subject Index 171

High Performance Non-Oxide Ceramics I

Phase Equilibria in the Si-B-C-N System

Silicon Carbide - A Survey of Synthetic Approaches, Properties

Amorphous Multinary Ceramics in the Si-B-N-C System

Keywords: Boron nitride, Hexagonal-BN, Cubic-BN, High-pressure high-temperature

2 Properties of the Various BN Phases . . . . . . . . . . . . . . . . . . 5

3 Hexagonal Boron Nitride (h-BN) . . . . . . . . . . . . . . . . . . . . . 14

Structure and Bonding, Vol. 102

3.1.1 Boric Acid with Carrier Substances . . . . . . . . . . . . . . . . . . . . 15

4.2.4.2 ICP CVD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

List of Abbreviations and Symbols

MISFET metal insulator semiconductor ®eld effect transistor

Similar to the carbon system, BN exists in a soft hexagonal (h-BN)

Table 1. Properties of h-BN and c-BN

Property Hexagonal BN Cubic BN

Fig. 1. Crystal structures of c-BN, w-BN, and h-BN

Hexagonal BN is mostly named h-BN, but also a-BN or g-BN (graphitic-BN)

Fig. 2. Pictures of h-BN powders. (Courtesy of ESK-Kempten, a company of WACKER-

Lattice parameters for w-BN are: a = 2.55  0.01 A Ê ; c = 4.20  0.01 A

Amorphous boron nitride (a-BN) can be synthesized by decomposition of

E-BN (E = explosion) is described as high pressure phase by a few scientists.

Fig. 4. Crystal structures of h-BN, r-BN, and t-BN

The transformation of c-BN into h-BN was investigated during DTA-analysis

c-BNDHr 298:15 K  869:2  2:0 kJ/mol 43

Enthalpy DHr° (298.15 K) [kJ/mol] )16.3  2.7 )16.5

Reference [48] [45] [49] [46] [50]

Due to the complexity of BN-structures and atomic bonding situations, the

Some data for analytical characterization are summarized in the following

Table 4. IR-peaks of c-BN and h-BN samples synthesized by different methods

c-BN h-BN Reference Method

± 810 1310 [57] CVD

This combination of excellent properties of hexagonal BN (h-BN) opens a

Balmain ®rst synthesized BN in 1842 by reaction of molten boric acid with

Boric acid with ammonia reacts in the presence of carrier substances

± In carbothermal reactions BN is synthesized by reduction of boric acid or

Using thermal CVD methods with B2H6±NH3±H2 [98, 99] or BCl3±NH3±H2

Hexagonal boron nitride is an interesting construction material because of its

Graphitic BN (h-BN) is used as lubricant with low friction in numerous

mixtures [107, 108]. Stabilizers, thickeners, oxidation-preventing additives are

To produce solid based lubricants, h-BN is added to plastics, rubbers, and

Dense shapes of h-BN are made exclusively by hot-pressing (HP) or hot-

components for high-temperature electric furnaces (Fig. 9) [120]. Other

SiAlON (Si3N4-Al2O3) ceramics reinforced with BN powder are used as break

TiN-based composites with additions of Al2O3 and BN show excellent

BN is used for the production of TiB2-BN and TiB2-AlN-BN composites. High

Lattice parameters for w-BN are: a = 2.55 0.01 A Ê ; c = 4.20 0.01 A

c-BNDHr 298:15 K 869:2 2:0 kJ/mol 43

Enthalpy DHr° (298.15 K) [kJ/mol] )16.3 2.7 )16.5

Si6 z Alz Oz N8 z 5