(Materials and Processing Technology) Rointan F. Bunshah, Rointan F. Bunshah - Handbook of Hard Coatings - Deposition Technologies, Properties and Applications-William Andrew (2002)

HANDBOOK OF HARD COATINGS
Deposition Technologies, Properties and

Applications
HANDBOOK OF HARD COATINGS
Deposition Technologies, Properties and
Applications
Edited by
Rointan F. Bunshah
University of California, Los Angeles
Los Angeles, California
NOYES PUBLICATIONS
Park Ridge, New Jersey, U.S.A.
WILLIAM ANDREW PUBLISHING, LLC
Norwich, New York, U.S.A.
Copyright © 2001 by Noyes Publications
No part of this book may be reproduced or
utilized in any form or by any means, electronic
or mechanical, including photocopying, recording
or by any information storage and retrieval
system, without permission in writing from the
Publisher.
Library of Congress Catalog Card Number: 01-27318
ISBN: 0-8155-1438-7
Printed in the United States
Published in the United States of America by

Noyes Publications / William Andrew Publishing, LLC
Norwich, New York, U.S.A.
10 9 8 7 6 5 4 3 2 1
Library of Congress Cataloging-in-Publication Data
Handbook of hard coatings / edited by Rointan F. Bunshah

p. cm.
Includes bibliographical references.
ISBN 0-8155-1438-7
1. Protective coatings. 2. Hard materials. 3. Coating processes.
4. Surface hardening. I. Bunshah. R. F. (Rointan Framroze)
TA418.76.H363 2001
667'.9--dc21 01-27318
CIP
MATERIALS SCIENCE AND PROCESS TECHNOLOGY SERIES
Series Editors
Gary E. McGuire, Microelectronics Center of North Carolina
Stephen M. Rossnagel, IBM Thomas J. Watson Research Center
Rointan F. Bunshah, University of California, Los Angeles (1927-1999), founding editor
Electronic Materials and Process Technology
CHARACTERIZATION OF SEMICONDUCTOR MATERIALS, Volume 1: edited by Gary
E. McGuire
CHEMICAL VAPOR DEPOSITION FOR MICROELECTRONICS: by Arthur Sherman
CHEMICAL VAPOR DEPOSITION OF TUNGSTEN AND TUNGSTEN SILICIDES: by
John E. J. Schmitz
CHEMISTRY OF SUPERCONDUCTOR MATERIALS: edited by Terrell A. Vanderah
CONTACTS TO SEMICONDUCTORS: edited by Leonard J. Brillson
DIAMOND CHEMICAL VAPOR DEPOSITION: by Huimin Liu and David S. Dandy
DIAMOND FILMS AND COATINGS: edited by Robert F. Davis
DIFFUSION PHENOMENA IN THIN FILMS AND MICROELECTRONIC MATERIALS:
edited by Devendra Gupta and Paul S. Ho
ELECTROCHEMISTRY OF SEMICONDUCTORS AND ELECTRONICS: edited by John
McHardy and Frank Ludwig
ELECTRODEPOSITION: by Jack W. Dini
HANDBOOK OF CARBON, GRAPHITE, DIAMONDS AND FULLERENES: by Hugh O. Pierson
HANDBOOK OF CHEMICAL VAPOR DEPOSITION, Second Edition: by Hugh O. Pierson
HANDBOOK OF COMPOUND SEMICONDUCTORS: edited by Paul H. Holloway and
Gary E. McGuire
HANDBOOK OF CONTAMINATION CONTROL IN MICROELECTRONICS: edited by
Donald L. Tolliver
HANDBOOK OF DEPOSITION TECHNOLOGIES FOR FILMS AND COATINGS, Sec-
ond Edition: edited by Rointan F. Bunshah
HANDBOOK OF HARD COATINGS: by Rointan F. Bunshah
HANDBOOK OF HYDROTHERMAL TECHNOLOGY: by K. Byrappa and Masahiro
Yoshimura
HANDBOOK OF ION BEAM PROCESSING TECHNOLOGY: edited by Jerome J.
Cuomo, Stephen M. Rossnagel, and Harold R. Kaufman
HANDBOOK OF MAGNETO-OPTICAL DATA RECORDING: edited by Terry McDaniel
and Randall H. Victora
HANDBOOK OF MULTILEVEL METALLIZATION FOR INTEGRATED CIRCUITS: ed-
ited by Syd R. Wilson, Clarence J. Tracy, and John L. Freeman, Jr.
HANDBOOK OF PLASMA PROCESSING TECHNOLOGY: edited by Stephen M.
Rossnagel, Jerome J. Cuomo, and William D. Westwood
HANDBOOK OF POLYMER COATINGS FOR ELECTRONICS, 2nd Edition: by James
Licari and Laura A. Hughes
HANDBOOK OF REFRACTORY CARBIDES AND NITRIDES: by Hugh O. Pierson
HANDBOOK OF SEMICONDUCTOR SILICON TECHNOLOGY: edited by William C.
O’Mara, Robert B. Herring, and Lee P. Hunt
HANDBOOK OF SEMICONDUCTOR WAFER CLEANING TECHNOLOGY: edited by
Werner Kern
v
vi Series
HANDBOOK OF SPUTTER DEPOSITION TECHNOLOGY: by Kiyotaka Wasa and

Shigeru Hayakawa
HANDBOOK OF THIN FILM DEPOSITION PROCESSES AND TECHNIQUES: edited by
Klaus K. Schuegraf
HANDBOOK OF VACUUM ARC SCIENCE AND TECHNOLOGY: edited by Raymond L.
Boxman, Philip J. Martin, and David M. Sanders
HANDBOOK OF VLSI MICROLITHOGRAPHY: edited by William B. Glendinning and
John N. Helbert
HIGH DENSITY PLASMA SOURCES: edited by Oleg A. Popov
HYBRID MICROCIRCUIT TECHNOLOGY HANDBOOK, Second Edition: by James J.
Licari and Leonard R. Enlow
IONIZED-CLUSTER BEAM DEPOSITION AND EPITAXY: by Toshinori Takagi
MOLECULAR BEAM EPITAXY: edited by Robin F. C. Farrow
SEMICONDUCTOR MATERIALS AND PROCESS TECHNOLOGY HANDBOOK: edited
by Gary E. McGuire
ULTRA-FINE PARTICLES: edited by Chikara Hayashi, R. Ueda and A. Tasaki
WIDE BANDGAP SEMICONDUCTORS: edited by Stephen J. Pearton
Ceramic and Other Materials—Processing and Technology
ADVANCED CERAMIC PROCESSING AND TECHNOLOGY, Volume 1: edited by Jon G.
P. Binner
CEMENTED TUNGSTEN CARBIDES: by Gopal S. Upadhyaya
CERAMIC CUTTING TOOLS: edited by E. Dow Whitney
CERAMIC FILMS AND COATINGS: edited by John B. Wachtman and Richard A. Haber
CORROSION OF GLASS, CERAMICS AND CERAMIC SUPERCONDUCTORS: edited
by David E. Clark and Bruce K. Zoitos
FIBER REINFORCED CERAMIC COMPOSITES: edited by K. S. Mazdiyasni
FRICTION AND WEAR TRANSITIONS OF MATERIALS: by Peter J. Blau
HANDBOOK OF CERAMIC GRINDING AND POLISHING: edited by Ioan D. Mavinescu,
Hans K. Tonshoff, and Ichiro Inasaki
HANDBOOK OF INDUSTRIAL REFRACTORIES TECHNOLOGY: by Stephen C. Carniglia
and Gordon L. Barna
SHOCK WAVES FOR INDUSTRIAL APPLICATIONS: edited by Lawrence E. Murr
SOL-GEL TECHNOLOGY FOR THIN FILMS, FIBERS, PREFORMS, ELECTRONICS
AND SPECIALTY SHAPES: edited by Lisa C. Klein
SOL-GEL SILICA: by Larry L. Hench
SPECIAL MELTING AND PROCESSING TECHNOLOGIES: edited by G. K. Bhat
SUPERCRITICAL FLUID CLEANING: edited by John McHardy and Samuel P. Sawan
— OTHER RELATED TITLES —
HANDBOOK OF PHYSICAL VAPOR DEPOSITION (PVD) PROCESSING: by Donald M.
Mattox
HERMETICITY OF ELECTRONIC PACKAGES: by Hal Greenhouse
SEMICONDUCTOR INDUSTRIAL HYGIENE HANDBOOK: by Michael E. Williams and
David G. Baldwin
SEMICONDUCTOR SAFETY HANDBOOK: edited by Richard A. Bolmen, Jr.
IN MEMORIAM
Professor Rointan F. (Ron) Bunshah

1927 – 1999
It is with great sadness that we note the death of Professor Rointan (Ron) F.
Bunshah. Ron passed away on October 24, 1999 after a long, gallant and quiet
battle with leukemia. This book constitutes his final project and will stand as a
memorial to his career and his many accomplishments over the years. This book
is also one of many dozens of books in the Materials Science and Process Technology
Series, originally from Noyes Publications, and now William Andrew Publishing,
that Ron helped originate and manage as Series Editor for many, many years.
Ron's career started over fifty years ago with his B.S. degree in 1948 from
the Banares Hindu University in India. After moving to the United States, he
received his M.S. and D.Sc. degrees in 1951 and 1952, respectively, from the
Carnegie Institute of Technology in Pittsburgh. He remained at Carnegie until
1954 as a research metallurgist and instructor. He then joined New York Univer-
sity as an Adjunct Professor and Research Scientist. Six years later, he moved to
the Lawrence Radiation Laboratory in Livermore, CA as a Senior Metallurgist. His
last appointment, in 1968, was to the University of California at Los Angeles, as
a Professor in the Department of Materials Science which is in the School of
Engineering and Applied Science. He took “early” retirement from UCLA in 1991,
but returned soon thereafter for another seven years, continuing to teach and lead
active research projects until the fall of 1998.
While at UCLA, Professor Bunshah was a distinguished senior faculty
member who took an active part in the University. He was presented with the
vii
viii In Memoriam
Distinguished Teaching Award of the Engineering Graduate Student Association

in 1971, and directly supervised twenty-eight M.S. and ten Ph.D. students as well
as mentoring over twenty post-doctoral researchers. His laboratory and his work
on vacuum metallurgy and thin films were internationally recognized as among the
best in the world.
Ron’s technical contributions covered a wide range, not only with metals
and alloys, but also in the realm of semiconductors, dielectrics, and diamond
coatings. His work covered the materials themselves as well as novel fabrication
and deposition techniques and technologies. The most important of these tech-
nologies is known as Activated Reactive Evaporation (ARE), which combined
evaporative deposition with a reactive plasma activation or modification of the
depositing films. This novel technology allowed researchers to alter the kinetics
of an evaporative deposition in ways which had been impossible, and led to new
discoveries in materials and processing. His research output was documented in
nearly three-hundred publications, a number of edited books, and innumerable
invited talks and presentations at conferences.
Professor Bunshah was also very generous in devoting a significant fraction
of his professional time to his colleges outside of UCLA. Much of this focused on
the American Vacuum Society, of which he was President in the early 1960's. He
was instrumental in two key accomplishments with the AVS: the formation of the
Vacuum Metallurgy Division in 1960, and the initiation of the International
Conference on Metallurgical Coatings (ICMC) in 1973. He served as the first
Chair of the Vacuum Metallurgy Division, and was Chair of the ICMC for at least
the first ten years of its operation. In recognition of this founding role, the ICMC
initiated in 1983 an “R.F. Bunshah” award, which is given in his honor each year
for the best paper at the Conference.
Ron received many honors during his career. From the AVS, he was awarded
the Gaede-Langmuir prize, AVS Fellow, and AVS Honorary Member. At the
ICMC Conference in 1998, I had the honor of presenting him with the Founder’s
Award, recognizing his singular devotion to organizing and running the Confer-
ence. Other honors included the First Life Member of the Indian Vacuum Society,
Life Membership in the Indian Institute for Metals, the Distinguished Alumnus
Award of the Benares Hindu University, and Fellowship in the American Society
for Metals. He was on the editorial board of many journals, including Applied
Surface Science, Surface Technology, Coatings Science and Technology, Thin
Solid Films, Research and Development, and International Research/Development.
His colleges within the ICMC community, the AVS, UCLA, and the
numerous students and collaborators, will now miss his leadership, his insight, and
his active drive to develop new deposition techniques and coatings, along with his
friendship which, for many, goes back several decades. We extend our deepest
sympathy to his wife, Zoreh, and his family.
Stephen Rossnagel June, 2000

Yorktown Heights, NY
Preface
This volume covers a wide range of topics relating to the fabrication,

characterization, and application of what we know as hard coatings. This book
project dates back many years to a proposed collaboration with Professor Dr.
Christian Weissmantel, who suggested a coedited volume with Professor Bunshah.
Unfortunately, Professor Weissmantel passed away and it was many years before
the project could be started.
The contributors to this volume represent the senior workers in this field
over the past twenty years, and their names will be well known to those working
in the field. It is a distinctly international list, representing the countries and labs
that have been at the forefront of hard coatings work. It should be noted that many,
if not all, of these authors are regular participants in the annual International
Conference on Metallurgical Coatings and Thin Films (ICMTF), a conference
founded by Professor Bunshah more than twenty-five years ago, and still vibrant
and growing each year.
The volume is roughly organized into three areas: fabrication, characteriza-
tion, and applications. The first chapter is a general introduction to the field of hard
coatings and wear-resistant surfaces by Professor Bunshah. In this brief chapter,
Bunshah describes some of the earliest work with these materials, and then defines
some of the terms used throughout the volume. This chapter is followed by two
general chapters on deposition technology. The first is also by Bunshah, and covers
the traditional vacuum-based thin-film techniques such as evaporation, sputter
deposition, and chemical vapor deposition. This chapter is quite useful in giving
the reader a wide overview of these topics along with the references and resources
for more detailed explanations. The second deposition chapter (Ch. 3) by Professor
Knotek describes alternate techniques used for thermal spraying of hard coatings
ix
x Preface
at high rates over large areas. These techniques and technologies are very
applicable to the commercial uses of hard coatings.
The fourth through sixth chapters introduce the science of characterizing
and measuring hard coatings or films. The first chapter in this section, by
Professors Hultman (Linkoping, Sweden) and Sundgren (Chalmers, Sweden), is
a very detailed and systematic overview of the relationships between the physical
properties of these films and composites, and the underlying structure of the
material. These two professors are without peer in the world for their fundamental
contributions to the science of hard coatings and multilayers, and the chapter is a
good introduction to the field. The second characterization chapter, by Bull
(Newcastle, UK) and Rickerby (Harwell, UK), explores some of the practical
characterization techniques used to evaluate hard materials. These include tech-
niques for hardness measurement, adhesion, and stress. The third characterization
chapter, by Bhushan and Gupta (Ohio State University), takes this a step farther,
examining the tribological properties of films. This includes hardness and adhe-
sion, as well as the measurement of friction and wear. These latter topics are key
to the applications of hard materials, which are generally used to protect underly-
ing softer materials.
The seventh and eighth chapters describe applications, and are broken into
two areas: coatings used for cutting tools, and coatings used for non-cutting
applications. In each case, the key advantage of hard coatings is that they can be
used to increase the quality and lifetime of tools of various types. These chapters
describe just some of the many applications. The ninth chapter covers a range of
materials somewhat different from the conventional hard coatings, which are often
nitrides or oxides (of metals). This chapter explores cubic boron nitride and
diamond-like films, and covers deposition, characterization, and application in the
single chapter. This materials set has some very unique properties and the field is
still rapidly evolving.
Finally, Professor Bunshah provides a summary chapter with comments on
each of the areas in the volume. In addition, he makes some suggestions and
predictions for future work and applications. In many ways, this chapter is most
valuable because it sums up a lifetime of experience and wisdom, and puts some
perspective on the field.
As mentioned in the Memoriam, this project is the last volume in the rather
prodigious output of Professor Bunshah. It is fitting, though, that this volume has
come together shortly after his death and it serves as a testament to his contribu-
tions over the many years.
Stephen Rossnagel
Yorktown Heights, NY
June, 2000
Contributors
Bharat Bhushan Otto Knotek

Ohio State University Technische Hochschule Aachen
Columbus, Ohio Aachen, Germany
Steve J. Bull G. Krämer

The University of Newcastle Technische Hochschule Aachen
Newcastle upon Tyne, England Aachen, Germany
Rointan F. Bunshah F. Löffler

University of California, Los Technische Hochschule Aachen
Angeles Aachen, Germany
Los Angeles, California, U.S.A.
David S. Rickerby
J. Wesley Cox Harwell Laboratory
Orland Park, Illinois United Kingdom Atomic Energy
Authority
Bal K. Gupta Oxfordshire, England
Ohio State University
Columbus, Ohio Chris H. Stoessel
Bodega Bay, California
Lars Hultman
Linkoping University Jan E. Sundgren
Linkoping, Sweden Linkoping University
Linkoping, Sweden
xi
NOTICE
To the best of our knowledge the information in this publication is

accurate; however the Publisher does not assume any responsibility
or liability for the accuracy or completeness of, or consequences
arising from, such information. This book is intended for informational
purposes only. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use by the Publish-
er. Final determination of the suitability of any information or
product for use contemplated by any user, and the manner of that
use, is the sole responsibility of the user. We recommend that
anyone intending to rely on any recommendation of materials or
procedures mentioned in this publication should satisfy himself as
to such suitability, and that he can meet all applicable safety and
health standards.
xii
Contents xiii
Contents
1 Introduction ............................................................................ 1
Rointan F. Bunshah
REFERENCES .................................................................................. 3
2 Vapor Deposition Technologies ............................................. 4

Rointan F. Bunshah
1.0 SCOPE .................................................................................. 4
2.0 CLASSIFICATION OF VAPOR DEPOSITION PROCESS ............... 5
3.0 PHYSICAL VAPOR DEPOSITION (PVD) PROCESSES .................. 8
3.1 Introduction .................................................................................. 8
3.2 Sputter Deposition ....................................................................... 11
3.3 Evaporation Deposition ............................................................... 22
4.0 PVD TECHNIQUES FOR DEPOSITION OF METALS,
ALLOYS, AND COMPOUNDS ................................................. 44
4.1 Single Element Deposition .......................................................... 44
4.2 Alloy Deposition .......................................................................... 44
4.3 Deposition of Compounds ........................................................... 47
4.4 Hybrid PVD Processes ................................................................. 57
xiii
xiv Contents
5.0 CHEMICAL VAPOR DEPOSITION (CVD) .................................... 60

5.1 Introduction ................................................................................ 60
5.2 Conventional CVD (Thermal CVD) ............................................ 61
5.3 Classification of CVD Reactions ................................................. 62
5.4 Rate-Limiting Steps ..................................................................... 64
5.5 Reactors ................................................................................ 64
5.6 Low Pressure CVD (LPCVD) ..................................................... 66
5.7 Plasma Assisted CVD (PACVD)/
Plasma Enhanced CVD (PECVD) ............................................. 66
5.8 Advantages and Disadvantages of PACVD ................................. 69
5.9 Advanced CVD Techniques ........................................................ 69
6.0 MATERIALS DEPOSITED BY PVD AND CVD TECHNIQUES ... 70
SUGGESTED READING .......................................................................... 71
REFERENCES ................................................................................ 72
3 Thermal Spraying and Detonation Gun Processes............ 77

Otto Knotek
1.0 HARD COATING WITH THERMAL SPRAYING PROCESSES ... 77
1.1 The Bonding Mechanisms of Thermally Sprayed Coatings ........ 81
2.0 PROCESSES ................................................................................ 84
2.1 Combustion Spraying .................................................................. 84
2.2 Combustion Flame Spraying ....................................................... 89
2.3 Plasma Spraying .......................................................................... 92
2.4 Special Processes ....................................................................... 100
3.0 COATING MATERIALS/CONSUMABLES/
FEEDSTOCK MATERIALS ..................................................... 101
3.1 Metals and Alloys ...................................................................... 102
3.2 Hard Compounds, Hard Metals ................................................. 104
3.3 Ceramics .............................................................................. 105
4.0 CONCLUSIONS .............................................................................. 107
REFERENCES .............................................................................. 107
4 Structure/Property Relationships For Hard Coatings .... 108

Lars Hultman and Jan E. Sundgren
1.0 INTRODUCTION............................................................................. 108
2.0 HARD MATERIALS........................................................................ 110
2.1 Intrinsic Hardness ...................................................................... 111
3.0 MICROSTRUCTURE AND MORPHOLOGY
OF HARD COATINGS ............................................................. 115
3.1 Characterization Methods .......................................................... 115
3.2 Influence of Deposition Parameters on Microstructure ............. 118
3.3 Substrate Effects on Microstructure .......................................... 146
Contents xv
4.0 INFLUENCE OF MICROSTRUCTURE ON COATING

PROPERTIES ............................................................................ 149
4.1 Stress .............................................................................. 149
4.2 Hardness .............................................................................. 159
5.0 ADHESION AND COATING/
SUBSTRATE INTERFACE STUDIES .................................... 166
6.0 SUMMARY .............................................................................. 171
ACKNOWLEDGMENTS ........................................................................ 172
REFERENCES .............................................................................. 173
5 Characterization of Hard Coatings................................... 181

Steve J. Bull and David S. Rickerby
1.0 INTRODUCTION............................................................................. 181
2.0 EVALUATION OF THE HARDNESS OF THIN FILMS ............... 182
2.1 Direct Coating Hardness Measurement ..................................... 183
2.2 Modeling the Hardness Behavior of Thin Films ....................... 186
3.0 MEASUREMENT OF COATING ADHESION .............................. 194
3.1 Definition of Adhesion .............................................................. 194
3.2 Requirements of an Ideal Adhesion Test ................................... 195
3.3 Measurement of Adhesion ......................................................... 196
4.0 X-RAY CHARACTERIZATION OF THIN FILMS ....................... 211
4.1 Measurement of Residual Stress by X-ray Techniques ............. 212
4.2 Line Broadening Measurements ................................................ 217
5.0 SUMMARY AND CONCLUSIONS ................................................ 221
REFERENCES .............................................................................. 223
6 Macro- and Micromechanical

and Tribological Properties ............................................... 229
Bharat Bhushan and Bal K. Gupta
1.0 INTRODUCTION............................................................................. 229
2.0 MEASUREMENT OF MECHANICAL PROPERTIES .................. 230
2.1 Apparatuses for Hardness Measurement on Macroscale ........... 233
2.2 Apparatuses for Hardness Measurement on Microscale ........... 235
2.3 Apparatuses for Hardness Measurement on Nanoscale............. 237
2.4 Mechanical Properties by Nanoindentation ............................... 249
2.5 Scratch Resistance/Adhesion Measurements ............................. 266
3.0 MEASUREMENT OF FRICTION AND WEAR............................. 279
3.1 Friction and Wear Measurements on Macroscale ...................... 279
3.2 Friction and Wear Measurements on Micro- and Nanoscale ... 287
3.3 Friction and Wear Measurements in MEMS ............................. 294
xvi Contents
4.0 MECHANICAL AND TRIBOLOGICAL PROPERTIES OF

DIAMOND AND AMORPHOUS CARBON COATINGS ...... 300
4.1 Deposition Processes ................................................................. 302
4.2 Mechanical Properties................................................................ 311
4.3 Tribological Properties .............................................................. 327
5.0 CLOSURE .............................................................................. 341
REFERENCES .............................................................................. 342
7 Applications to Cutting Tools ............................................ 370

Otto Knotek, F. Löffler, and G. Krämer
1.0 INTRODUCTION............................................................................. 370
1.1 Wear Mechanisms in Machining ............................................... 373
1.2 CVD-Coated Cutting Tools ....................................................... 377
1.3 PVD-Coated Cutting Tools ........................................................ 388
1.4 Conclusions .............................................................................. 404
REFERENCES .............................................................................. 407
8 Wear and Corrosion Resistant Hard Coatings

For Non-Cutting Tool Applications .................................. 411
J. Wesley Cox
1.0 INTRODUCTION............................................................................. 411
2.0 WEAR AND CORROSION PHENOMENA ................................... 413
2.1 Wear .............................................................................. 413
2.2 Corrosion .............................................................................. 414
3.0 SELECTION OF COATING MATERIALS
AND COATING PROCESSES ................................................. 416
4.0 APPLICATIONS OF HARD COATINGS FOR WEAR AND
CORROSION............................................................................. 420
4.1 Applications with Thermal Spray Coating (TSC) Methods ...... 421
4.2 Applications with Chemical Vapor Deposition (CVD) Methods428
4.3 Applications with Physical Vapor Deposition (PVD) Methods 440
4.4 Applications with Electro/Electroless Plating (EP/ElsP) Methods 447
5.0 DIAMOND COATINGS .................................................................. 452
REFERENCES .............................................................................. 457
9 Cubic Boron Nitride and Diamond-Related Thin Films. 466

Chris H. Stoessel, Rointan F. Bunshah
1.0 INTRODUCTION............................................................................. 466
2.0 CUBIC BORON NITRIDE FILMS .................................................. 467
2.1 Structure and Properties of B-N Compounds ............................ 467
2.2 Synthesis of c-BN Materials ...................................................... 468
2.3 Applications .............................................................................. 478
Contents xvii
3.0 DIAMOND-RELATED MATERIALS ............................................ 482

3.1 Structure and Properties of Diamond-Related Materials ........... 482
3.2 Synthesis of Diamond-Related Materials .................................. 485
3.3 Applications .............................................................................. 502
3.4 Emerging Materials.................................................................... 509
4.0 CONCLUDING REMARKS ............................................................ 512
REFERENCES .............................................................................. 513
ADDITIONAL READING ...................................................................... 515
10 Summary, Developments, and Outlook ............................ 517

Rointan F. Bunshah
1.0 CURRENT INDUSTRIAL PVD PROCESSES ............................... 517
2.0 CVD AND PACVD PROCESSES ................................................... 519
3.0 TRIBOLOGICAL COATINGS AND THEIR APPLICATIONS .... 520
4.0 ALUMINA COATINGS BY CVD, PACVD AND PVD PROCESSES
AND FIRST MACHINING STUDIES ..................................... 522
5.0 ENVIRONMENTAL CONCERNS IN MACHINING
AND FORMING ........................................................................ 525
6.0 DECORATIVE HARD COATINGS ................................................ 527
7.0 SUPERHARD COATINGS FOR CUTTING TOOLS ..................... 528
8.0 POLYCRYSTALLINE SUPERLATTICE COATINGS .................. 531
9.0 NOVEL SUPERHARD COATINGS ............................................... 533
REFERENCES .............................................................................. 533
Index .......................................................................................... 535

Introduction 1
Introduction
Rointan F. Bunshah
Mankind has used hard surfaces on various tools since antiquity.

Examples are:
1. An axe dating back to 900 B.C., possessed a Brinell Hardness
value of 444 at the edge where it had been intentionally carburized. A
chisel from 700 B.C., less satisfactorily carburized, had a Brinell Hard-
ness of 300 on its edge.[1] These are precursors of modern techniques of
carburizing and nitriding, which is a chemical vapor deposition process.
2. The ancient precursor to micro and nano laminate composites
are the “Damascus” swords, and the Roman and Viking swords made
during the latter centuries of the first millennium A.D. They consisted of
finely spaced alternating layers of high carbon, and low carbon steel
which had been forged to create intricate patterns with the layers stacks
going in different directions. The result was high strength and toughness
and, in the case of Damascus swords, a very pleasing sheen on the
surface.[2]
Traditionally, the term hard coatings refers to the property of high
hardness in the mechanical sense with good tribological properties. With
1
2 Handbook of Hard Coatings
the development of modern technology in the areas of optical, optoelec-

tronic, optical, and related defense applications, the definition of the term
hard coatings can be extended. Thus, a system which operates satisfacto-
rily, in a given environment can be said to be hard with respect to that
environment.[3] Thus hard materials can be classified as:
1. Tribologically hard—wear resistant, and low friction.
2. Optically hard—laser, and photonically inert.
3. Radiation hard—high threshold energies for energetic
particles such as gamma rays, neutrons, and beta
particles.
4. Electrically hard—high bandgap, and large electron
velocities.
Most hard coatings are ceramic compounds such as oxides, car-
bides, nitrides, ceramic alloys, cermets, metastable materials such as
diamond, and cubic boron nitride. Their properties, and environmental
resistance depend on the composition, stoichiometry, impurities, micro-
structure, imperfections, and in the case of coatings, the preferred orien-
tation (texture). Diamond, diamond-like carbon (DLC), cubic boron
nitride, and the new class of nano-layered composites are currently
referred to as superhard materials.
This volume is oriented at tribological properties. It is broken down
into three main sections.
1. Deposition Technologies. The technologies covered
are “Vapor Deposition Techniques” (PVD, and CVD),
and “Thermal Spray and Detonation Gun Processes.”
2. Characterization, Structure and Properties. The
chapters in this section are: “Structure/Property
Relationships for Hard Coatings,” “Characterization of
Hard Coatings,” and “Macro- and Micromechanical
Properties and Tribological Properties.”
Introduction 3
3. Applications. The chapters are: “Applications to Cutting

Tools,” “Wear and Corrosion Resistant Coatings for
Non-Cutting Tool Applications,” and “Cubic Boron
Nitride and Diamond Materials and Films.”
A final chapter brings together “New Developments and Out-
look.”
REFERENCES
1. Aitchison, L., A History of Metals, Interscience Publishers, 2:351

(1960)
2. Aitchison, L., A History of Metals, Interscience Publishers, 1:209
(1960)
3. Bunshah, R. F., and Deshpandey, C. V., “Hard Coatings,” Vacuum,
30(10):955 (1989)
Vapor Deposition
Technologies
Rointan F. Bunshah
1.0 SCOPE
This chapter deals with the vapor deposition technologies used

to deposit hard/wear resistant coatings onto various softer substrates.
Companion volumes in this series discuss these vapor deposition
processes in much greater detail.[1][2] Until very recently, these
substrates were metal, or metal alloys (e.g., steels, titanium, alumi-
num), or cermets (e.g., W-Co). The hard coating improved the wear
resistance of these materials by diminishing the adhesive, or abrasive
type wear. Currently, polymer materials in the form of long continu-
ous webs or bulk forms such as windows or lenses have become
candidates for wear/water vapor permeation resistant coatings (usu-
ally oxides such as silica, titania, or alumina). This has been a major
challenge to vapor deposition technology since the polymer materi-
als can withstand very low temperatures (80–300°C) during film
deposition. It is only through the detailed understanding of the
plasma assisted deposition processes that this coating triumph has
been achieved.
4
Vapor Deposition Technologies 5
2.0 CLASSIFICATION OF VAPOR DEPOSITION

PROCESS
Vapor deposition processes may principally be divided into

two types, (1) physical vapor deposition (PVD) and (2) chemical
vapor deposition (CVD).[1][2] PVD processes involve the creation of
material vapors, (by evaporation, sputtering, or laser ablation) and
their subsequent condensation onto a substrate to form the film.
CVD processes are generally defined as the deposition of a solid
material from the vapor phase onto a (usually) heated substrate as a
result of chemical reactions. PVD and CVD processes can be classi-
fied as shown in Table 1.
Table 1. Classification of PVD and CVD Processes
PVD Processes
Metals Compounds
Basic PVD Processes Basic PAPVD Processes

Evaporation Deposition Direct Activated Reactive
Evaporation (ARE)
Evaporation
Sputter Deposition or Reactive Sputtering (RS)
Sputtering
Hybrid PVD Processes Hybrid PAPVD
Processes
“Ion Plating” “Reactive Ion Plating”
CVD Processes
Basic CVD Process Basic PACVD Processes

Thermal CVD RF Excitation
Microwave Excitation
Photon Excitation
Table 2 shows the characteristics of basic PVD and CVD

processes. Generally speaking, the CVD process has the advantage
of good throwing power, while the deposition rates in PVD pro-
cesses are higher than those in CVD processes at a lower deposition
temperature.
Table 2. Characteristics of Basic PVD and CVD Processes
Evaporation Sputtering Chemical Vapor

Deposition
Mechanism of Thermal energy Momentum Chemical reaction

production of transfer
depositing species
Deposition rate Can be very high Low except for Moderate
(up to 750,000 pure metals (e.g., (200–2,500
Å/min) Cu-10,000 Å/min) Å/min)
Depositing species Atoms and ions Atoms and ions Atoms
Throwing power Poor line-of-sight Good, but nonuniform Good
coverage except thickness
by gas scattering distribution
Metal deposition Yes Yes Yes
Alloy deposition Yes Yes Yes
Refractory compound Yes Yes Yes
deposition
Energy of deposited Low Can be high Can be high with
species (0.1–0.5 eV) (1–100 eV) plasma-aided CVD
Bombardment of Not normally Yes Possible
substrate/deposit
Growth interface Not normally Yes Yes (by rubbing)
perturbation
Substrate heating Yes, normally Not generally Yes
(by external means)
As with all processes, one is concerned with the process itself,

as well as the resulting microstructures and properties of the product.
In order to understand various vapor deposition processes, one has to
model them in terms of three steps[3] illustrated in Fig. 1.
Figure 1. The three steps in film deposition.
Step 1: Creation of Vapor Phase Specie—Material can

be put into the vapor phase by evaporation, sputtering,
laser ablation, gases, vapors, etc.
Step 2: Transport from Source to Substrate—The vapor
species are transported from the source to the substrate
with, or without collisions between atoms and molecules.
During transport, some of the vapor species can be
ionized by creating a plasma in this space.
Step 3: Film Growth on the Substrate—This process
involves the condensation of the vapor species onto the
substrate and subsequent formation of the film by
nucleation and growth processes. The nucleation and
growth processes can be strongly influenced by
bombardment of the growing film by ionic species
resulting in a change in microstructure, composition,
impurities, and residual stress.
The degree of independent control of these three steps deter-

mines the versatility or flexibility of the deposition process. For
example, these three steps can be independently controlled in PVD
processes, and therefore give greater degree of flexibility in control-
ling the structure, properties, and deposition rate, whereas all of the
three steps take place simultaneously at the substrate, and cannot be
independently controlled in the CVD process. Thus, if a choice is
made for a process parameter such as substrate temperature which
governs deposition rate in CVD, one is stuck with the resultant
microstructure and properties.
3.0 PHYSICAL VAPOR DEPOSITION (PVD)

PROCESSES
3.1 Introduction
The basic PVD processes fall into two general categories: (1)
sputtering and (2) evaporation. The application of PVD techniques
ranges over a wide variety of applications from decorative, to high
temperature superconducting films. The thickness of the deposits
can vary from angstroms to millimeters. Very high deposition rates
(25 µm/sec) have been achieved with the advent of electron beam
heated sources. A very large number of inorganic materials—metals,
alloys, compounds, and mixtures—as well as some organic materials
can be deposited using PVD technologies.
The terms evaporation and sputtering refer to the mechanisms
whereby a condensed species is transferred to the vapor phase,
discussed in detail below. Of major importance is the bombardment
of the growing film by energetic species (ions and energetic neutrals)
from the adjacent plasma. Therefore, the method of plasma genera-
tion, and the ion energies are of crucial importance to the structure,
and properties of the deposited film. Figure 2 illustrates the various
effects on film formation by ions and fast neutrals impinging on the
film as a function of their energy. These phenomena include:
Thermal activation of the condensing species resulting

in an increase in their surface mobility.
Chemical activation to form a compound by reaction
between the depositing species where the dominating
role is played by excited particles.
Desorption of impurity atoms, either originally present
on the substrate or co-deposited.
Creation of activated centers (charge defects, structural
defects, bulk defects).
Sputtering of impurity atoms.
Sputtering of deposited film species followed by re-
condensation, which helps in filling up the gaps between
the film islands thus resulting in a continuous film at
lower total thickness than without particle bombardment.
Implantation, which is not of interest here.
Thus, for thin continuous films, the electron energies should be

in the range of 10 to 103 eV, and the electron density should be in the
range of 1012 to 1013 cm-3. Figure 3 shows the range of plasma
density, and electron energy for different low pressure plasma dis-
charges. Ion Plating is a hybrid PVD process since it is defined as an
atomistic film deposition process in which the substrate surface and/
or the depositing film is subjected to a flux of high energy particles
sufficient to cause changes in the interfacial region between the film
and the substrate, as well as in the properties of the deposited film as
compared to a non-bombarded film. These changes may be in the
adhesion of the film to the substrate, film morphology, density, or
stress. The source of the depositing species can be evaporation,
sputtering, gases, or vapors. A classification of PVD processes is
given in Table 1.
10
Handbook of Hard Coatings
Figure 2. Effect on layer formation by bions and fast neutrals impinging on the layer as a function of their energy.
(Courtesy Dr. S. Schiller, FEP.)
Figure 3. Range of plasma density and electron energy in different low pressure
plasma discharges. (Courtesy Dr. S. Schiller, FEP.)
3.2 Sputter Deposition
Sputtering Phenomena. Sputtering is the phenomena of mo-

mentum transfer from an incident energetic projectile to a solid or
liquid target resulting in the ejection of surface atoms or molecules.
In the sputter deposition process, the target, (a source of coating
material) and the substrate are placed in the vacuum chamber, and
evacuated to a pressure typically in the rage 10-4 to 10-7 torr. A
schematic diagram of the sputter coating process is shown in Fig. 4.
The target, (also called a cathode) is connected to a negative voltage
supply, and the substrate generally faces the target. A glow discharge
is initiated after an inert gas (usually argon gas) is introduced into the
evacuated chamber. Typical working pressure is in the range of 20 to
150 mtorr. The sputter target erosion rate is given by:
JSM A
R = 62.3 Å/min
ρ
where J is the ion current density in mA/cm2, S is the sputtering yield

in atoms/ion, MA is the atomic weight in grams, and ρ is the density
gm/cm3 of the target material. Sputtering yield assuming perpen-
dicular ion incidence onto a target consisting of a random array of
atoms can be expressed as:
M i M t Ei
S = constant ×
( M i + M t ) U Atom/ion
where: Mi: mass of the incident atom

Mt: mass of the target atom
Ei: kinetic energy of the incident ion
U: heat of sublimation of the target material
Planar Diode Glow Discharge Sputter Deposition. This is

the simplest sputtering system. This configuration consists of the
cathode (the target) and anode facing each other (Fig. 4). The
substrates are placed on the anode. The target, which is usually
water-cooled, performs two functions during the process, one as the
source of coating material, and the other as the source of second-
ary electrons sustaining the glow discharge. The distance between
the cathode and anode is usually about 5 to 10 cm.
Figure 4. Planar diode sputter deposition.

The discharge current increases with the applied voltage, thus,

increasing the sputtering rate. However, the discharge current does
not increase linearly with the applied voltage as the voltage increases
above 100 eV, since the ionization cross section decreases with
increasing electron energy. The sputtering rate can be increased if the
working gas pressure increases at a given voltage, due to an increase
in ion collection by the cathode.[4] However, the deposition rate
starts decreasing at high gas pressures due to gas scattering of the
sputtered atoms.
The deposition rate is mainly determined by the power density
at the target surface, size of the erosion area, source-substrate dis-
tance, source material, and working gas pressure. Some of these
factors are interrelated, such as pressure and power density. There-
fore, the optimum operating condition is obtained by controlling the
parameters to get the maximum power which can be applied to the
target without causing cracking, sublimation or melting. The maxi-
mum power limit can be increased if the cooling rate of the target is
increased by designing the coolant flow channels properly, and
improving the thermal conductance between the target and the target
backing plate.
Even though planar diode glow discharge sputter deposition
techniques are widely used due to their simplicity and the relative
ease of fabrication of targets for a wide range of materials, they have
several disadvantages, such as low deposition rate, substrate heating
due to the bombardment of high energy particles, and relatively
small deposition surface areas.
Magnetron Sputter Deposition. By employing magnetic fields
to the diode sputtering process, the ionization efficiency near the
target can be greatly increased. In the conventional planar diode
process, ions are generated relatively far from the target and the
probability for ions to lose their energy to the chamber walls is large.
Furthermore, the number of primary electrons hitting the anode at
high energies without experiencing collisions is increased as the
pressure decreases, thus reducing ionization efficiency. These elec-
tron losses are not offset by impact-induced secondary emission.
Therefore, ionization efficiencies are low and the discharge cannot
be sustained in planar diodes at pressures below about 20 mtorr.
In magnetron sputter deposition process, an applied magnetic

field parallel to the cathode surface forms electron traps and restricts
the primary electron motion to the vicinity of the cathode. The
magnetic field strength is in the range of few hundred gauss and
therefore, it can influence the plasma electrons but not the ions. The
electrons trapped on a given field line can advance across the mag-
netic field to an anode or walls by making collisions (mostly with gas
atoms). Therefore, their chances of being lost to the walls or anode
without collisions are very small. Because of the higher efficiency of
this ionization mechanism, the process can by operated at pressures
around 1 mtorr with high current densities at low voltages, thus
providing high sputtering rates.
There are several configurations of magnetron sputter deposi-
tion technologies. Figures 5 and 6 show the cylindrical magnetron
and the planar magnetron respectively. The cylindrical magnetron is
a very useful technique to prepare uniform coatings over large areas,
since long cathodes are employed in these techniques. Furthermore,
the cylindrical-hollow magnetron technique is effective for coating
complex-shaped objects. The cylindrical-post magnetron can be
used to substantially decrease substrate bombardment by energetic
particles, thus minimizing the heat of the substrate. Metallic films,
and dielectric films can be deposited with high deposition rates using
planar magnetron sputtering, as compared to diode sputtering.
Even though magnetron sputtering techniques have the advan-
tages of high sputtering rate and low bombardment rates of energetic
particles onto the substrate, the utilization of this technique is im-
peded by the limitations in choice of target materials, and the diffi-
culties in fabrication of the target. For example, if ferromagnetic
materials are used as the sputtering target, their thickness should be
thin enough so as to be saturated by the magnetic field. Since high
power is localized in a selected area in magnetron sputtering process,
targets should be prepared without voids or bubbles to avoid local
melting and spitting.
Figure 5. Cylindrical magnetron (post cathode) sputter deposition.
Figure 6. Planar magnetron sputter deposition.

Balanced and Unbalanced Magnetrons. The magnetic field

in a normal/balanced magnetron is designed to keep the electrons,
and hence, the plasma in the vicinity of the substrate. The balanced
magnetron was developed for microelectronic applications, where
bombardment of the growing film by energetic particles was to be
avoided. For hard coatings, it is necessary to bombard the growing
film with energetic particles. One, then, has the choice of setting up
a secondary plasma near the substrate as a source of energetic
particles, use a separate ion gun, or suitably alter the magnetic fields
between the inner, and the outer sets of magnets in the sputtering
cathode, so as to permit the plasma to extend to the region of the
substrate, resulting in the desired ion bombardment of a positively
biased substrate. This is called the unbalanced magnetron, and ex-
amples are shown in Fig. 7. It may be noted that multiple pairs of
cathodes may be used to ensure that the plasma covers the entire
deposition volume without any dark spots with the south and north
poles alternating from one set to the adjacent set. [5]–[7] This ar-
rangement is now commonly used for deposition of hard nitride/
carbide coatings onto alloy steel and carbide tools.
Figure 7. Balanced and unbalanced sputter deposition.

RF Sputter Deposition. The development of the RF sputter

deposition technique made it possible to deposit films from noncon-
ducting sputtering targets which cannot be sputtered by the DC
methods due to charge accumulation on the target surface.
Most ions are almost immobile, as compared to electrons
which can follow the temporal variations in the applied potential at
the typical RF frequencies used for sputtering (5–30 MHz). When
the electrode is coupled to an RF generator, a negative voltage is
developed on the electrode due to the difference in mobility between
electrons and ions. Since the insulating target electrode constitutes a
capacitor in the electrical circuit, there should be no DC component
to the current flow. Therefore, the voltage on the electrode surface is
required to be self-biased negatively to compensate for the differ-
ence in mobility of electrons and ions, and to satisfy the condition of
net zero current (averaged over each cycle). The magnitude of the
resulting negative bias is almost the same as the zero-to-peak voltage
of the RF signal. The period for the electrode to act as a anode is very
short and the electrode mostly acts as a cathode during the RF cycle.
Therefore, we can expect the target to be sputtered as in the DC case.
Significant numbers of ions are not accumulated on the target surface
while the electrode acts as a cathode, due to high frequency em-
ployed in RF sputtering (normally 13.56 MHz).
Deposition can be performed at considerably lower pressures,
such as 5 to 15 m torr in RF sputtering, as compared to the planar DC
discharge, since electrons oscillating at high frequency can obtain
sufficient energy to cause ionizing collisions, and the number of
electrons lost (without making collisions) can be reduced.
RF sputtering is widely used to deposit various kinds of con-
ducting, semiconducting, as well as insulating coatings despite the
complexity of the RF power source. This technique can also be
applied to magnetron sputtering sources.
Pulsed Mid-Frequency Sputter (PMS) Deposition Process
for Insulating Compounds. This is a recent development for the
high rate sputter deposition of insulating compounds (see discussion
on reactive sputtering in Sec. 4.3). Professor S. Schiller and his

colleagues at FEP (Fraunhofer Institut Elektronstrahl und
Plasmatechnik) are the pioneering group for the industrial
development.[8a][8b] The stimuli for the development of this technol-
ogy can be gauged from the deficiencies of DC sputter and RF
sputter deposition of insulating compounds, and is illustrated in Fig.
8. For the DC case, one has the undesirable phenomena of arcing
(both macro and micro arcs), see Fig. 9, shifting of electrical poten-
tials by insulating layers, and low plasma density near the substrate
surface. For the RF case, one has low deposition rates and high
power losses (reflected power), high expenditure for impedance
matching and associated high equipment cost, limited large area
application, high voltages, and high particle energies. The solution is
mid-frequency sputter deposition process in the frequency range of
10–100 kHz. The pulse width, off-time (if used), and the polarity can
be varied. The pulse can be unipolar, where the voltage is typically
negative with a no-voltage (off) time or the bipolar, where the
positive and negative voltages are equal (symmetric) or unequal
(asymmetric), perhaps with an off-time. These arrangements are
illustrated in Fig. 10. In the positive bias and off-time, electrons
move to the surface from the plasma to neutralize any charge build
up produced during the negative portion of the cycle, thus preventing
arcing over the dielectric surface and through a dielectric film being
deposited on a metallic surface. In the negative portion of the cycle,
energetic ion bombardment can sputter dielectric films that have
been formed on metallic targets during reactive sputtering, thus
keeping the target in the metallic mode. For the asymmetric case, the
negative pulse (-350 V for example) is greater than the positive pulse
(+100 V for example). The negative pulse time is typically 80 to 90%
of the voltage cycle, and the positive pulse time is 10 to 20% of the
voltage cycle, resulting in high energy bombardment of the growing
film over a much greater portion of the cycle, as compared to RF
sputtering.
Figure 8. Deficiencies of DC and RF sputter deposition processes. (Courtesy Dr.

S. Schiller, FEP.)
Symmetrical pulsed DC is generally used in a dual magnetron

sputtering configuration where each of the magnetrons are alterna-
tively biased positive and negative. This eliminates the “disappear-
ing anode” effect with DC sputtering of insulating films. However,
the dual cathode pulse sputtering arrangement does not sputter dur-
ing as large a portion of the cycle, as compared with asymmetrical
pulsed DC sputtering from a single magnetron.
Figure 9. Simple model of arcing formation during reactive DC sputter process.

(Courtesy Dr. S. Schiller, FEP.)
Figure 10. Bipolar mid-frequency pulsed magnetron sputtering. (Courtesy Dr. S.

Schiller, FEP.)
The material being sputtered influences the frequency, dura-

tion, and pulse height. The frequency is lower for an insulating oxide
material, as compared to a somewhat higher frequency for a poorly
conducting material such as TiC or TiN, or ITO due to the easier
drainage of the charge from the surface of these materials.
Ion Beam Sputter Deposition. A relatively recent variation

called ion beam sputtering (Fig. 11) provides: (i) excellent adhesion,
(ii) a resulting high purity deposit due to a result of low operational
pressures (about 0.1 mtorr), and (iii) very low substrate heating
effects since the substrate is not in contact with the plasma. In this
technique, an ion beam of high energy (hundreds to thousands of
electron volts) extracted from the ion source is directed at a sputter-
ing target of the desired material. An inert, or reactive gas is used for
the ion beam source. The substrate is suitably located to collect the
sputtered species from the sputtering target, as shown in Fig. 11.
There are two kinds of ion sources mainly used in practical thin film
deposition, the Kaufman source and the duoplasmatron. No further
discussion of these sources will be presented in this chapter; please
refer to the Handbook of Ion Beam Processing Technology. [9]
Figure 11. Ion beam sputter deposition.
Since the substrate can be isolated from the plasma generation

source, this permits independent control over the substrate tempera-
ture, gas pressure, and the type of particle bombardment of the
growing film. Also, it is possible to control the energy and the target
current density independently in this technique, whereas it can be
done only by varying the working gas pressure in glow discharge
sputtering technology. The deposition rate in ion beam sputtering is
lower than in conventional sputtering. The reason for this low rate is
mainly due to the low beam current in the conventional ion beam
sputtering (dual-grid) system. This low-beam current can be greatly

increased by using a single-grid system; Nishimura, et al.,[10] ob-
tained a very high deposition rate over 90 nm/min for various
materials, such as Al, Cu, etc., using single-grid ion beam sputtering.
A small coverage of deposition area due to the small ion beam size
(about 1 cm) is another drawback of the ion beam sputtering. The
beam size can be increased to about 10 cm by adaptation of space-
type extraction optics.
3.3 Evaporation Deposition
Introduction. In evaporation processes, vapors are produced

from a material located in a source which is heated by various
methods. A schematic of an evaporation system is shown in Fig. 12.
It consists of an evaporation source to vaporize the desired material,
and the substrates which are located at an appropriate distance facing
the evaporation source. Resistance, induction, arc, electron beam, or
lasers are the possible heat sources for evaporation. The substrate
can be heated and/or biased to the desired potential using a DC/RF
power supply. Evaporation is carried out in vacuum, in a pressure
range of 10-5 –10-10 torr. In this pressure range, the mean free path
(MFP) is very large (5 × 102–10 7 cm) as compared to the source-to
substrate distance. Hence, the evaporated atoms essentially undergo
a collisionless line of sight transport prior to condensation on the
substrate, thus, leading to thickness build up directly above the
source, and decreasing steeply away from it. Planetary substrate
holders are therefore used in some cases so as to even out the vapor
flux on multiple substrates. In some cases an appropriate gas such as
argon at pressures of 5–200 mtorr is introduced into the chamber to
reduce the MFP so that vapor species undergo multiple collisions
during transport from the source to substrate thus producing reason-
ably uniform thickness coatings on the substrate. The technique is
called gas scattering evaporation or pressure plating.[11]
Figure 12. Evaporation deposition process schematic.
The transition of solids or liquids into the vapor phase is an

atomistic phenomenon. It is based on thermodynamics and results in an
understanding of evaporation rates, source-container reactions, and the
accompanying effect of impurity introduction into the vapor state,
changes in composition during alloy evaporation and stability of
compounds. An excellent detailed treatment of the thermodynamic
and kinetic bases of evaporation processes is given by Glang.[12]
The rate of evaporation is given by the well known Hertz-
Knudsen equation,
dNe − 1
dt = α υ (2πmkT ) ( p∗ − p)
2
Ae
where αυ is the evaporation coefficient, dNe /Aedt is the number of

molecules evaporating from a surface area Ae in time dt, p* is the
equilibrium vapor pressure at the evaporate surface, p is the hydro-
static pressure acting on the surface, m is the molecular weight, k is
Boltzmann’s constant, and T is the absolute temperature. The evapo-
ration coefficient, αυ, is very dependant on the cleanliness of the
evaporate surface, and can range from very low values for dirty
surfaces, to unity for clean surfaces.
For reasonable deposition rates (100–1000 nm/min) at a source
to substrate distance of 20 cm, the vapor pressure should be about
10-2 torr. The source temperature should be adjusted to give this
value of the vapor pressure.
The directionality of evaporating molecules from an evapora-
tion source is given by the Cosine Law. Holland[13] and Graper [14]
have thoroughly discussed the theoretical distribution of vapor from
a point, a wire, a small surface, an extended tip, and from cylindrical
and the ring types, or sources. For the ideal case of deposition from a
clean, uniformly emitting point source onto a plane receiver, the rate
of deposition varies as cosθ /r2 (Knudsen’s cosine law), where r is
the radial distance of the receiver from the source, and θ is the angle
between the radial vector and the normal to the receiver direction.
Evaporation Sources, General Considerations. Evaporation
sources are classified by the mode of heating used to convert the
solid or liquid evaporant to the vapor phase. Thus, one talks of
resistance, arc, induction, electron beam, arc imaging, lasers, and
exploding wire types of sources. A very important fact to be noted is
that we cannot evaporate every material from any of the types of
sources listed above for the following reasons:
1. Chemical interaction between the source material
and the evaporant which would lead to impurities in
the deposit. For example, evaporation of titanium
from a MgO source would cause oxygen and
magnesium contamination of the deposit; the
titanium would reduce the MgO. Therefore, for the
evaporation of reactive metals like titanium,
zirconium, etc., we use water cooled copper crucibles.
2. Reaction between metallic source (such as a W or

Ta boat), and evaporant (Ti), could occur. In many
cases, at high temperatures two metals can mutually
dissolve in each other leading to a destruction of the
source.
3. The power density (i.e., watts per sq cm) varies
greatly between the various heat sources.
Table 4.2 from Ref. 12, the article by Glang, lists the tempera-
ture, and support materials to be used in the evaporation of elements.
Similar tables are found in the literature of many of the manufacturers.
Evaporation of alloys and compounds pose additional prob-
lems and they are considered later.
Resistance Heated Sources. The simplest vapor sources
are resistance heated wires and metal foils of various types
shown in Fig. 13.
Figure 13. Wire and metal foil resistance heated evaporation sources. (a) hairpin
source; (b) wire helix; (c) wire basket; (d) dimpled foil; (e) dimpled foil with
alumina coating; (f) canoe type. (From The Handbook of Thin Film Technology,
©1970, McGraw Hill.)
They are available in a variety of sizes and shapes, and at

sufficiently low prices so that they can be discarded after one experi-
ment if necessary. They are usually made from the refractory metals,
tungsten, molybdenum, and tantalum which have high melting points,
and low vapor pressure so as not to contaminate the deposit.
Platinum, iron, or nickel are sometimes used for materials
which evaporate below 1000°C. The capacity (total amount of
evaporant) of such sources is small. The hairpin and wire helix
sources are used by attaching the evaporant to the source in the form
of small wire segments. Upon melting, the evaporant must wet the
filament, and be held there by surface tension. This is desirable to
increase the evaporation surface area and thermal contact. Multistrand
filament wire is preferred because it increases the surface area.
Maximum amount held is about 1 gram. Dimpled sources and basket
boats may hold up to a few grams.
Since the electrical resistance of the source is small, low
voltage power supplies, 1 to 3 kW, are recommended. The current in
the source may range from 20 to 500 amps. In some cases, the
evaporant is electroplated onto the wire source.
The principal use of wire baskets is for the evaporation of
pellets or chips of dielectric materials, which either sublime or do not
wet the wire on melting. In such cases, if wetting occurs, the turns of
the baskets are shorted and the temperature of the source drops.
The rate of evaporation from such sources may vary consider-
ably due to localized conditions of temperature variation, wetting,
hot spots, etc. Therefore, for a given thickness of film, the procedure
is to load the source with a fixed weight of evaporant and evaporate
to completion, or use a rate monitor and/or thickness monitor to
obtain the desired evaporation rate and thickness.
Sublimation Sources. For materials evaporating above 1000°C,
the problem of non-reactive supports may be circumvented for
materials such as Cr, Mo, Pd, V, Fe, and Si which reach a vapor
pressure of 10-2 torr before melting. Hence, they can sublime and
produce a sufficiently high vapor density. The contact area between
the evaporant and the source crucible is held to a minimum. Figure
14 shows such a source designed by Roberts and Via.
Figure 14. Chromium sublimation source. (After Robert and Via.) The current
flows through the tantalum cylinder (heavy lines). (From The Handbook of Thin
Film Technology, ©1970, McGraw Hill.)
A different type of sublimation source is used for the vaporiza-

tion of thermally stable compounds such as SiO, which are com-
monly obtained as powders or loose chunks. Such source material
would release large quantities of gases upon heating, thus, causing
ejection of particles of the evaporant which may get incorporated
into the film. Figure 15 shows two sources which solve this problem
by reflection of the vaporized material.
Figure 15. Optically dense SiO sources; (a) the Drumhellar source; (b) Compart-
mentalized source. (After Vergara, Greenhouse, and Nicholas.) (From The Hand-
book of Thin Film Technology, ©1970, McGraw Hill.)
Evaporation Source Materials. We have already discussed

the potential problems concerned with the reaction between metal
sources and evaporates. Oxides and other compounds are more
stable than metals. There are many metals which can be evaporated
from refractory oxide sources. Note that there is no such thing as an
absolutely stable oxide, nitride, or other compound. Reaction is
controlled by kinetics, i.e., temperature and time.
Oxide crucibles have to be heated by radiation from metal
filaments, or their contents can be heated by induction heating. This
is illustrated in Figs. 16 and 17 for resistance heated sources.
Figure 16. Oxide crucible with wire coil heater. (From The Handbook of Thin
Film Technology, ©1970, McGraw Hill.)
Figure 17. DaSilva crucible source. (From The Handbook of Thin Film Technol-
ogy, ©1970, McGraw Hill.)
Other source materials are nitrides, such as boron nitride. A

50% BN–50% TiB2 is also well established as a crucible material.
This material (HDA composite, Union Carbide) is a fairly good
electrical conductor and hence can be directly heated to evaporate
materials. It can be readily machined to shape. Pyrolytic BN and

carbon are also used.
Induction Heated Sources. Figure 18 shows the induction
heated sources using a BN–TiB2 crucible. Figure 19 shows an
induction heated evaporation sublimation source using a water cooled
copper crucible. This is suited to the evaporation of reactive metals
such as Ti, Be, etc., which will react with all the refractory oxides,
nitrides, etc.
Figure 18. RF heated aluminum source with boron-nitride/titanium diboride

crucible. (After Ames, Kaplan, and Roland.) (From The Handbook of Thin Films
Technology, ©1970, McGraw Hill.)
Figure 19. Induction heated sublimation source using a water cooled copper
base.
Electron Beam Heated Sources. Electron beam heated sources

have two major benefits. One, is a very high power density, and
hence, a wide range of control over evaporation rates from very low,
to very high. Two, the evaporant is contained in a water-cooled
copper hearth, thus, eliminating the problem of crucible contamina-
tion.
The evaporation rate for pure metals like Al, Au, Ag, which are
good thermal conductors, from water-cooled copper crucibles de-
creases due to heat loss to the crucible walls. In such cases, crucible
liners of carbon, and other refractory materials are used.
Any gun system must consist of at least two elements, a
cathode and an anode. In addition, it is necessary to contain these in
a vacuum chamber in order to produce and control the flow of
electrons, since they are easily scattered by gas molecules. A poten-
tial difference is maintained between the cathode and the anode. This
varies from as little as a few kilovolts to hundreds of kilovolts. In
melting systems, a normal operational range is of the order of 10–40
kV. In the simple diode system, the cathode emits electrons, which
are then accelerated to the anode across the potential drop. Where the
anode is the workpiece to be heated, this is termed a work-acceler-
ated gun. It is shown schematically in Fig. 20. In a self-accelerated
gun structure, an anode is located fairly close to the cathode, elec-
trons leave the cathode surface, are accelerated by the potential
difference between the cathode and anode, pass through the hole in
the anode and continue onward to strike the workpiece. Self-acceler-
ated guns have become the more common type in use and offer more
flexibility than the work-accelerated gun.
Electron beam guns may be further subdivided into two
types depending on the source of electrons: (1) thermionic
gun and (2) plasma gun.
(a)
(b)
Figure 20. Simple electron beam guns, (a) work accelerated gun; (b) self
accelerated gun.
Thermionic Gun. In thermionic guns, the source of electrons is

a heated wire, or disc of a high temperature metal or alloy, usually
tungsten or tantalum. Such guns have the limitation of a minimum
operating gas pressure of about 1 × 10-3 torr. Higher pressures cause
scattering of the electron beam, as well as a pronounced shortening
of the cathode life (if it is a wire or filament) due to erosion by ion
bombardment. Figure 21 shows examples of thermionic electron
beam heated work-accelerated sources. The close cathode gun shown
in Fig. 21 is not a desirable configuration since molten droplet
ejection from the pool impinging on the cathode will terminate the
life of the cathode due to low melting alloy formation. Thus, cath-
odes are hidden from direct line-of-sight of the molten pool and the
electron beam is bent by electrostatic fields (Figs. 21b and 21c), or
magnetic field (Figs. 22 and 23) generated by electromagnets. The
latter is a preferred arrangement since variation of the X and Y
components of the magnetic field can be used to scan the position of
the beam on the molten pool surface.
(a) (b)
(c)
Figure 21. Work accelerated electron beam sources. (a) pendant-drop source, (b)
shielded filament (Unvala), (c) shielded filament. (Chopra and Randlett.) (From
The Handbook of Thin Film Technology, ©1970, McGraw Hill.)
Figure 22. Bent beam electron gun with water cooled evaporant support. (With
permission of Temescal Metallurgical Co., Berkeley, CA.) (From The Handbook
of Thin Film Technology, ©1970, McGraw Hill.)
Figure 23. Transverse electron beam gun.
Figures 21, 22, and 23 show linear cathodes (i.e., wires or

rods), and are referred to as transverse linear cathode guns. Figure 24
shows a disc cathode which is characteristic of a high power Pierce
type electron beam gun. Low power Pierce type guns may have a hair
pin filament, or a wire loop as the cathode. In either case, the beam
geometry of the Pierce gun is different than that of the transverse
linear cathode guns. In some instances, the electron emitter assembly
is located at a distance from the crucible in a separately pumped
chamber to keep the pressure below 1 × 10-3 torr, with a small orifice
between the emitter chamber and the crucible chamber for the pas-
sage of electrons.
Figure 24. Schematic representation of a Pierce gun.
Plasma Electron Beam Gun. A plasma is defined as a region of

high-temperature gas containing large numbers of free electrons and
ions. By a proper application of electrical potential, electrons can be
extracted from the plasma to provide a useful energy beam similar to
that obtained from thermionic guns. There are three types of plasma
e-beam guns:
1. The Cold Cathode Plasma Electron Beam Gun.
The plasma electron beam gun has a cylindrical
cathode cavity made from a metal mesh or sheet,
(Fig. 25) containing the ionized plasma from which
electrons are extracted through a small aperture in
one end. The cathode is maintained at a negative
potential, e.g., -5 to -20 kV, relative to the workpiece

and remainder of the system, which are at ground
potential. After evacuation of the system, a low
pressure of ionizable gas in the range of 10-3 to 10-
1 torr is introduced. Depending upon the high voltage
level, a long path discharge between the cathode,

and other parts of the system will occur in the gas at
a particular pressure. Ionizing collisions in the gas
then produce positive ions which are accelerated to
the cathode, causing electrons to be released from
the cathode surface. Although the cathode may heat
up slightly due to ion bombardment, no heating is
required for electron emission. Upon proper
adjustment of cathode voltage and gas pressure, a
beam mode of operation is established, since
interaction between the plasma inside and outside
of the cathode, and the electric fields between
cathode surface and plasma boundary will largely
confine electron emission to the end of the cathode
and its interior. In argon, a beam mode is supported
at about 10-2 torr with 5–0 kV. Beam currents range
up to 3 A for a 3 inch diameter cathode in argon at
20 kV. With lighter gases, e.g., helium, higher
pressure to about 10-1 torr will yield a beam mode in
this same voltage range. Beam current will vary
with voltage and pressure control, also. More specific
information is given by Cocca and Stauffer.[15] The
beam is self-collimating because of the focusing
effect of positive ions in the beam path, and the
electrostatic lensing action of the aperture since it
separates regions of different potential gradient.
The beam has a cross section equal to that of the
cathode aperture. Adjustment of focus can be
achieved to some extent by varying pressure and
voltage, but external focusing may also be used if
desired, with magnetic or electrostatic lenses, as
with conventional electron beams.
Figure 25. Cold cathode electron beam gun.
2. The Hot Hollow Cathode Discharge Beam Gun.

The hollow cathode discharge beam applied to
vacuum processing has been reported by Morley,[16]
and differs in a number of respects from the plasma
beam. A schematic of the hollow cathode discharge
beam is shown in Fig. 26. Here the cathode must be
constructed of a refractory metal since it operates at
elevated temperature. An ionizable gas, usually
argon, is introduced into the system through the
tubular-shaped cathode. A pressure drop across the
orifice in the cathode provides a sufficient amount
of gas inside the cathode to sustain the plasma,
which generates the beam. A low voltage, high
amperage DC power source is utilized. When RF
power from a commercial welding starter is coupled
to the gas, it becomes ionized, and the plasma is
formed. Continued ion bombardment of the cathode

results in heating of the cathode and increased
electron emission. Ultimately, a high current “glow
discharge” will occur, analogous to that experienced
in vacuum arc melting at higher pressures. At this
point, the discharge appears as a low power density
beam “flowing” from the cathode aperture and
fanning out in conical shape into the chamber.
However, a parallel axial magnetic field is imposed
on the beam (as seen in Fig. 26) which then forms a
high power density, well-collimated beam. The
hollow cathode discharge beam is operationally
stable, and efficient over the pressure range from
10-4 to 10-1 torr. A more detailed description of
physical aspects, operational characteristics, and
cathode design has been given by Morley.[16]
Figure 26. Schematic of the hot hollow cathode electron beam gun.
3. RF Hollow Cathode Low Voltage EB Guns.[17] The

construction, and operation of this source is very
similar to the DC version. The pressure of the gas
inside the hollow cathode is almost independent of
the reactor pressure, which allows to generate
discharge at high collision frequency and transport
it into the low pressure reactor. The discharge forced
out from the hollow cathode forms a decaying
plasma channel with extraordinary properties. Gas
metastables excited inside the cathode can act in
selected gas mixtures as a source of additional heat,
thereby enhancing thermionic electronic emission
and ionization of the gas. An arc regime can be
started from the glow discharge by simply increasing
the RF power. Hollow cathode arc in the metal
vapor can be sustained even without the working
gas. It is used for film deposition and dry etching.
Comparisons. Thermionic, as well as the plasma e-beam guns
can be used equally well for evaporation. Focusing of the beam spot
is easier for the thermionic guns. The plasma guns have the advan-
tage of being able to operate at higher pressures, which can be
important for gas scattering evaporation, reactive evaporation, and
ion plating.
Arc Evaporation. The definitions of arcs are: Karl T. Compton,
Princeton University: “A discharge in a gas or vapor.” Lafferty, “The
arc is a self-sustained discharge capable of supporting large currents
by providing its own mechanism of electron emission from the
cathode.” Berghaus[18] describes the use of arcs to form refractory
compounds by reactive evaporation. Since 1940, consumable and
nonconsumable vacuum-arc melting processes have been developed
to melt and refine various reactive metals such as Ti, Hf, Zr, etc.
More recently, arc techniques have been used to deposit metals[19][20]
and refractory compounds, and even for extraction of ions from the
vacuum arc plasma for the deposition of metals.[21]
Wroe[22] in 1958 and Gilmour, et al.,[23] suggested vacuum arcs

as a source for metallic coatings. The U.S. patents to Snaper[24][25] in
1971, and the Russian patents to Sablev[26][27] in 1974, set the stage
for the commercial production of arc coatings, which were achieved
in the USSR around 1977–78. The first commercial use of the arc
evaporation-deposition method was for TiN coatings deposited at
low temperatures, particularly for high-speed steel cutting tools by
arc evaporation of titanium in a nitrogen plasma. This follows on the
heels of the Activated Reactive Evaporation (ARE) process devel-
oped in 1971 for deposition of refractory compounds such as TiN
using electron beam evaporation techniques. There is very extensive
Russian literature on vacuum arc coating technology and the reader
can find a convenient source in recent reviews by Sanders[28] and by
Martin,[29] and in a recent companion volume in this series.[30]
There are two types of cathodic arc systems—pulsed and
continuous. In the pulsed devices, the arc is repeatedly ignited, and
extinguished using a capacitor blank to supply the arc power. Pulsed
arcs have the advantage of letting the target cool between the pulses.
The disadvantage is the decrease in steady-state coating rates.
The continuous cathodic arc can be random in nature or con-
trolled. By the use of an insulating ring, a random arc source can be
constrained at the edge of the target, but allowed random motion
within that constraint. Random arc sources have the advantage of
simplicity and excellent target utilization because the entire target
(except near the very edge) is utilized in the arc of very large parts.
The main disadvantage of random arcs is the formation of
macroparticles, which may cause the resulting coating to be unsuit-
able in some applications. Figure 27 shows that macroparticles are
ejected at small angles with respect to the target surface, and can
therefore be minimized using appropriate shielding. Such a strategy
has made possible arc-produced decorative coatings where surface
finish and optical specularity are of concern.
Figure 27. Phenomena occurring at a discrete cathodic arc spot.
Magnetic fields can be used to control the trajectories of the

arcs. These fields can be used to discourage the arc from leaving the
desired portion of the target surface or can actually be used to define a
well controlled path for the arc to follow in the so-called “steered arc”
devices. While the mechanism is still the subject of some debate, it is
clear, at least in the case of ceramic coatings based on refractory
metals, that steered arcs can produce coatings having extremely low,
or no measurable macroparticle component.
Macroparticles can also be removed by the use of suitable
filters as discussed by Sanders[28] and by Martin.[31] This is the so-
called “filtered arc evaporation process.”[30] Other strategies for
macroparticles involve the production of diffuse arcs. In one case,
the cathode is contained in a crucible which is allowed to heat up to
a temperature where the target material has a substantial vapor
pressure.[32] This causes a decrease in the arc voltage and current
density, the discharge becomes diffuse and macroparticles no longer
form. The other approach is the so-called “anodic arc,”[33]–[35] (see

Fig. 28). In this process, the cathode initially supplies electronics, as
well as ions until the anode heats up. At this point, with sufficient
electron emission, a diffuse arc forms on the hot anode target mate-
rial which supplies the ions necessary to sustain the discharge. The
cathode material is not evaporated, and the coating material now
emanates from the anode. There are no macroparticles formed. High
deposition rates (several µm per minute) are obtained for a variety of
metals including Al, Ti, V, Ca, Mn, Fe, Ni, Cu, Pd, Ag, Au, and
Pt.[33] Since the substrate is left relatively cool, the process makes it
possible to produce adherent coatings on plastics at temperatures less
than 70°C, which makes this relatively new process a competitor for
sputter deposition. Alloy coatings such as stainless steel can be
readily deposited with good stoichiometric transfer. For example,
Ni, Al, and stainless steel coatings less than 1 µm thickness impart
excellent corrosion protection to iron.[36]
Figure 28. Schematic of the anodic arc evaporation process.

One of the main advantages of arc deposition processes is the

relatively high level of ionizing atoms in the plasma. This makes it
convenient to extract ion beams from the plasma and deposit
macroparticle free coatings entirely from the ion beam.[28][31]
Laser Induced Evaporation/Laser Ablation/Pulsed Laser
Deposition (PLD). This technique with many names was first used
by Smith and Turner[37] in 1965 to deposit thin films in a vacuum
chamber used a pulsed ruby laser. Systematic studies in the 1970s were
performed to provide a better understanding of the physics of
laser-solid interactions, and the related issues of deposition mecha-
nisms and film quality. More recently, the process has been extensively
used for growing highly crystalline dielectric films,[38] compound
semiconductor epitaxial layers, layers for bandgap engineering,[39][40]
and very extensively for high Tc superconducting films.[41][42] The
reader is referred to an excellent review by Cheung and Sankur.[43]
In this technique, material is vaporized and ejected from the
surface of a target as it irradiated by a laser beam. Films are formed
by condensing the material ablated from the target onto a solid
substrate. Absorption characteristics of the material to be evaporated
determine the laser wavelength to be used. To obtain the high power
density required in many cases, pulsed laser beams are generally
employed. Pulse width, repetition rate, and pulse intensity are se-
lected for specific applications.
In some studies on YBCO film deposition, the laser version of
a plasma-assisted reactive evaporation process was used. Oxygen
was bled into the system, and a plasma was created in the target-
substrate space by the use of a positively biased electrode placed
some distance above the target. This is the ARE process geometry
developed earlier and described later.
Although laser evaporation is an attractive approach for syn-
thesis of high purity metal alloys and compound films, it suffers from
the following limitations:
• Complex transmitting and focusing systems need
to be employed to direct the beam from the laser
located outside the vacuum system onto the
evaporant placed inside the system. This involves

special optical path designs and increases the cost
of the setup. Also, a window material which
efficiently transmits the wavelength band of the
laser must be found and mounted in such a way that
it is not rapidly covered up by the evaporant flux.
• It is not always possible to find a laser with
wavelength compatible with the absorption
characteristics of the material to be evaporated.
• Energy conversion efficiency is very low-usually
around 1 to 2%.
• The size of the deposited film is small (10 to 20 mm
or 0.4 to 0.8 in., diameter), resulting from the small
size of the laser impact spot.
• The “splashing effect,” which involves the
production of microparticles between 0.1 and 10
µm in size, diminishes film quality.
The main advantages of this technique are:
a. The production of high-energy species which
enhances film quality.
b. Excellent transfer of stoichiometry between the target
and the film, e.g., the deposition of hydroxyl apatite
thin films for biomedical applications such as
implants.
The macroparticle density can be decreased by lowering the
power level at the expense of deposition rate. The latter may not be
important for many thin film applications.
The question of large area deposition has been recently ad-
dressed by Greer.[42] He has constructed a vacuum deposition system
in which the laser beam is scanned on a rotating YBCO target and the
substrate is itself rotated. This rather complex apparatus is capable of
depositing YBCO films onto two or three inch diameter substrates.
4.0 PVD TECHNIQUES FOR DEPOSITION OF METALS,

ALLOYS, AND COMPOUNDS
The great versatility of the PVD processes is their ability to

deposit a very large number of materials including metals, alloys,
semiconductors, superconductors, polymers, and fabricate compos-
ites of various types (particulate, fibrous, or laminate).
4.1 Single Element Deposition
This can be carried out by evaporation or sputter deposition

processes. The deposition rate depends on the process and process
parameters.
4.2 Alloy Deposition
Alloys consist of two or more components, which have differ-

ent vapor pressures and hence, different evaporation rates. As a
result, the composition of the vapor phase and therefore the deposits
are constantly varying. The following solutions have been used to
obtain alloy films with stoichiometry close to the source composi-
tion by evaporation based techniques.
a. Co-evaporation or co-sputtering using multiple sources.
This technique involves simultaneous co-evaporation
of the constitutive elements of the alloy. The
composition of the deposited film is controlled by
adjusting the evaporation/sputtering rate of the
respective elements. In elaborate systems, separate
deposition rate monitors are used with appropriate
feedback networks to control the deposition rate from
each individual source independently. Near
stoichiometric films of many binary alloys have been
successfully deposited using this technique. Dispersion
strengthened alloys such as Ni-ThO2 have also been
successfully deposited by this technique.
b. Evaporation from a single source. This technique

involves evaporation of an alloy using a rod-fed electron
beam source. Evaporation operates under steady-state
conditions where the composition and volume of the
liquid pool on the top of a solid rod are kept constant
(see Fig. 29). A detailed description is given in Ref. 45.
c. Flash evaporation. In this process, pellets of the alloy
are dropped onto a very hot strip, and are vaporized
completely, thus maintaining the composition of the
alloy in the deposit. It works very well for elements
with high vapor pressures.
d. Sputter deposition from an alloy target.
e. Sputter deposition from a segmented target where the
segments consist of each of the two components of the
alloy and the ratio of the target sample area of each
element is inversely proportional to the sputtering yields.
Figure 29. Schematic of direct evaporation of an alloy from a single rod-fed

source.
46
Figure 30. Magnetron sputter source with double ring plasma on electrically insulated separate targets. (Courtesy Dr. S.
Schiller, FEP.)
f. Sputter deposition from a double ring magnetron

target[46] as illustrated in Fig. 30.
g. Laser ablation from an alloy target.
4.3 Deposition of Compounds
Deposition of compounds can be performed in two ways: (1)

Direct evaporation, using conventional heating methods or laser
ablation, where the composition of the evaporant is the same as that
of the compound that is to be deposited and (2) Reactive evapora-
tion, where the elements of the compound are evaporated, and react
with the gas to form the compound. Plasma assisted reactive evapo-
ration process are often used since they activate the reactions leading
to compound formation. Reactive evaporation, and plasma assisted
reactive evaporation will be discussed later.
Direct Evaporation. When a compound is heated, evaporation
can occur with or without dissociation of the compound into frag-
ments. There are a small number of compounds which are evapo-
rated without dissociation, specific examples being SiO2, MgF 2,
B2O3, GaF2, and other Group IV divalent oxides.
In the more general case, when a compound is evaporated, the
material is not transformed to the vapor state as compound mol-
ecules but as fragments thereof. Subsequently, the fragments have to
recombine on the substrate to reconstitute the compound. Therefore,
the stoichiometry (anion: cation ratio) of the deposit depends on
several factors including the vaporization rate, the ratios of the
various molecular fragments in the vapor, the impingement of other
gases present in the environment on the film, the surface mobility of
the fragments (which in turn depends on their kinetic energy and
substrate temperature), the mean residence time of the fragments of
the substrate, the reaction rate of the fragments on the substrate to
reconstitute the compound and the impurities present on the sub-
strate. For example, it was found that direct evaporation of Al2O3
resulted in a deposit which is deficient in oxygen. In other cases, the
deposit from direct evaporation of TiB2 contains both the monoboride,
and diboride phases.[47]
Laser Ablation. Laser ablation deposition techniques have

attracted great interest in recent years for the synthesis of semicon-
ducting, and insulating thin films. Very recently these techniques
have been applied successfully for the deposition of high Tc super-
conducting films.[48][49]
In this technique, material is vaporized and ejected from the
surface of a target as it is irradiated by a laser beam. Films are formed
by condensing material ablated from the target onto a solid substrate.
Absorption characteristics of the material to be evaporated deter-
mine the laser wavelength to be used. To obtain the high power
density required in many cases, pulsed laser beams are generally
employed. Pulse width, repetition rate, and pulse intensity are se-
lected for specific applications.
Although laser ablation is an attractive approach for the syn-
thesis of high purity metal, alloy, and compound films, it suffers
from some of the same limitations as laser evaporation:
1. Complex transmitting and focusing systems need
to be employed to direct the beam from the laser
located outside the vacuum chamber onto the
evaporant placed inside the system. This involves
special designs and increases the cost of the setup.
Also, a window material which efficiently transmits
the wave length band of the laser must be found and
mounted in such a way that it is not rapidly covered
up by the evaporant flux.
2. It is not always possible to find a laser with
wavelength compatible with the absorption
characteristics of the material to be evaporated.
3. Very low energy conversion efficiency.
4. The size of the deposited film is small, 1–2 cm dia.,
resulting from the small size of the laser impact
spot.
Reactive PVD Process. Reactive PVD processes are those in
which at least one of the elements of the coating is introduced in the
gas phase. Various compounds have been synthesized from metal

targets with reactive gases: air, O2, or H2O (oxides), N2 or NH3
(nitrides), O2 + N2 (oxynitrides), H2S (sulfides), etc. There are
several advantages to these processes: (i) various kinds of com-
pounds can be prepared using relatively easy-to-fabricate metallic
targets, (ii) insulating compound films can be prepared using DC
power supplies (reactive sputtering), and (iii) graded composition
films can be formed.
Reactive Sputtering (RS). Sputter deposition is performed in
the presence of reactive species in the gas phase in reactive sputter-
ing process. Reactions can occur on the cathode surface, at the
substrate, and in the gas phase. However, reactions in the vapor
phase are precluded by considerations of momentum and energy
conservation unless the process is performed at high pressures to
allow multiple many body collisions in the gas phase.
In this process, the target is a nominally pure metal. The
compound film is synthesized by sputtering in a pure reactive gas or
an inert gas-reactive gas mixture. Usually, the inert-reactive gas
mixture is preferred to the pure reactive gas from considerations of
sputtering rate. In the RS technique, a compound target also can be
used. In this case, the target is chemically decomposed by inert gas
ion bombardment. It is usually necessary to add the reactive gas to
compensate for the loss of reactive component by dissociation.
The main problem in reactive sputtering is target poisoning as
shown in Fig. 31. As the reactive gas partial pressure increases, the
rate of compound formation exceeds the removal rate of compounds
on the target surface, resulting in a decrease in deposition rate due to
the low sputtering yield of compound formed on the cathode surface,
and the fact that compounds have higher secondary electron emis-
sion yield than pure metals. The increase in secondary electron
emission results in a reduction in both the discharge voltage, and ion
component in the cathode current at constant voltage. In other words,
more of the energy of incoming ions is consumed to produce and
accelerate secondary electrons.
Figure 31. Schematic showing target poisoning effect during reactive sputtering.
The poisoning effect has two practical problems. One is the

aforementioned loss in deposition rate. The second relates to the
transition behavior, the material being deposited often passing abruptly
from metal to a nearly stoichiometric compound. This transition
behavior has been observed in planar diodes, planar magnetrons, and
cylindrical magnetrons. The transition behavior is most abrupt for
oxide deposition and more gradual for nitrides and carbides.
Much of the analysis of the transition behavior is incomplete
since it considers only the phenomena occurring at the cathode while
ignoring the total system. It is important to realize that reactive
sputtering processes are dependent on the total system, i.e., its
geometry, the accumulation of coating on the walls and fixtures, and
the location of the gas injection. All of these have to be carefully
controlled to use reactive sputtering effectively in production. Non-
linear models show that with a sufficiently high overall pumping
speed, there is a smooth transition between metal and compound
sputtering modes. This very high pumping speed becomes a practical

concern in the design of the vacuum system and process for reactive
sputtering.
Several solutions have been found to reduce the effect of target
poisoning. They are:
1. Very high pumping speed.
2. Feedback control techniques using mass
spectrometric control and varying the reactive gas
pressure to operate in the metal mode.
3. Incorporation of getter surfaces on the surfaces
surrounding the target and substrate, and the latter
at the target.
4. Incorporation of a baffle over the target to limit the
impingement of reactive gas molecules on the target.
Such target shields may have a single opening or a
distributed opening using a mesh.
5. Creation of a plasma near the target using secondary
electronics from the target and inserting a positively
biased electrode near the substrate.
All of these are discussed in Ref. 50 by Karim, et al.
Reactive Evaporation Process. The difficulties involved in
direct evaporation processes due to fragmentation of the vaporized
compounds are overcome in reactive evaporation where a metal is
evaporated in presence of the reactive gas. The compound is formed
by reaction of the evaporating metal species with the molecules of
the reactive gas. Even though this technique has been extensively
used to deposit a variety of oxide films for optical applications, it is
generally observed that the films are deficient in oxygen. It is also
observed in some cases, especially in the synthesis of carbide films,
that the deposition rate becomes a limiting factor governing the
growth of the films. In such cases stoichiometric TiC films could
only be deposited at very low rates (about 1.5 Å/sec). This limitation
of deposition rate in the case of the reactive evaporation process is
due to the reaction kinetics of the compound formation by this
process. The presence of a plasma in the Activated Reactive Evapo-

ration (ARE) process influences the reaction kinetics by providing
activation energy to the reactive species, thereby making it possible
to synthesize compound films at considerably higher rates[51] and
lower temperatures.
Plasma Assisted Reactive Evaporation Processes. The ARE
process which was the earliest of these processes generally involves
evaporation of a metal, or an alloy in the presence of the plasma of a
reactive gas.[52][53] For example, TiC and TiN coatings are deposited
by this process by evaporating Ti in the presence of C2H2 and N2
plasmas respectively. The two basic variants of the ARE process are
shown in Figs. 32 and 33. For more information on the other modifi-
cations of the ARE process, refer to a review by Bunshah and
Desphandey.[54] The role of the plasma is twofold:
1. To enhance the reactions that are necessary for
deposition of compound films.
2. To modify the growth kinetics, and hence, the
structure/morphology of the deposits.
Figure 32. The activated reactive evaporation (ARE) process using an electron
beam evaporator.
Figure 33. The activated reactive evaporation (ARE) process using resistance
heated evaporation source.
In the last few years, new modifications of the plasma assisted

evaporation process have been developed. The key element of these
processes is the very high plasma density (much higher than the prior
reactive PVD processes)—see Fig. 3. A comparison of the estab-
lished (older) processes with the new high rate evaporation tech-
niques. It may be noted that with the newer processes (Table 3), there
is almost a factor of 100 improvement in deposition rate, vapor
particle density, plasma density, and ion current density on the
substrate. A dramatic result of this high plasma density is the ability
to deposit oxides such as alumina at high rates (good economics) at
low temperatures so that substrates such as plastic webs can be
continuously coated with alumina.
These three new processes are SAD—Spotless Arc Deposi-
tion, MAD—Magnetron Activated Deposition, and HAD—Hollow
Cathode Activated Deposition.
Table 3. Typical Parameters for the deposition of compounds by

plasma activated high rate evaporation in comparison to current
plasma-activated deposition techniques. (Courtesy Dr. S. Schiller,
FEP.)
Typical parameters Established High-rate

(near the substrate) technologies evaporation
deposition rate 1–10 nm/s 100–1000 nm/s

vapor particle density* 1011 cm-3 10 13 cm-3
plasma density 10 9–10 10 cm-3 1011–1012 cm-3
reactive gas pressure 10 -2 Pa 1 Pa
mean free path 300 mm 3 mm
between reactive
gas particles*
ion current density 0.1–5 mA/cm2 5–50 mA/cm2
*Note: High gradient between source and substrate
In the SAD process, shown in Fig. 34, a dense vapor of the

evaporant is produced by focusing a high power electron beam (300
kW) onto a liquid metal pool.[55] A positively biased electrode is
located above the pool. The crucible is connected to the negative of
the power supply, and a high-current discharge is initiated. The
substrate is biased to a negative potential. A major advantage of the
SAD process over cathodic arc evaporation is the absence of the
discrete cathode spots, and the corresponding droplets. Figure 35
shows the relationship between the ion current density, deposition
rate, and the impact ratio of ions and condensing vapor particles for
the SAD process, and other PVD process. Results to date show the
considerable improvement of the mechanical properties of refractory
metals.
Figure 34. The spotless arc deposition (SAD) process. (Courtesy Dr. S. Schiller
FEP.)
Figure 35. Relationship between ion current density jbias, deposition rate aD, and
impact ratio (alpha) of ions and condensing vapor particles for various plasma
deposition processes. (Courtesy Dr. S. Schiller, FEP.)
In the MAD process, the metal is evaporated using resistance,

or high voltage electron beam heating. A dense plasma is created
over the molten pool using a pair of opposed pulsed magnetrons.
Thus, boat evaporation has been combined with bipolar pulsed
magnetron sputter technology (PMST) to develop a new plasma
activated evaporation process.[56] A schematic of the process is
shown in Fig. 36. A potential application is the deposition of alumina
on plastic webs.
Figure 36. The magnetron activated deposition (MAD) process. (Courtesy Dr. S.
Schiller, FEP.)
In the HAD process, aluminum vapor produced by an electron

beam line evaporator passes through a very dense plasma created by
a LVEB (low voltage electron beam) discharge along with oxygen
gas to deposit an alumina coating on a wide substrate. There are two
variations of the HAD process:
1. A liquid vapor source such as aluminum for alumina

coatings (Fig. 37).
2. An electron beam heated silica cylinder for silica
coatings.
From Fig. 3, we note that the HAD process has the highest
plasma density of all the Plasma Assisted Reactive Evaporation
Process.[57]
Figure 37. The hollow cathode deposition (HAD) process. (Courtesy Dr. S.
Schiller, FEP.)
4.4 Hybrid PVD Processes
Ion Plating. Ion plating is a hybrid PVD or CVD process, [58]

since the coating material is vaporized by thermal energy (i.e.,
evaporation), or momentum transfer (i.e., sputtering), or supplied as
a vapor (very similar to CVD processes). In this technique, the
vaporized (or supplied) coating materials pass through a gaseous
glow discharge on their way to the substrate, thus ionizing some of

the vaporized atoms, generally 1% (Fig. 38).
Figure 38. An ion plating configuration using a DC diode discharge and a

thermal vaporization source. (After Mattox.)
The glow discharge is produced by biasing the substrate to a

high negative potential (-2 to 5 kV) and admitting a gas, usually
argon, at a pressure of 5 to 200 m torr into the chamber, see Fig. 38.
In this simple mode, which is known as diode ion plating, the
substrate is bombarded by high energy gas ions which sputter off the
material present on the surface. This results in a constant cleaning of
the substrate (i.e., a removal of surface impurities by sputtering)
which is desirable for producing better adhesion, and lower impurity
content. The ion bombardment also causes a modification in the
microstructure and residual stress in the deposit. On the other hand, it
produces the undesirable effects of decreasing the deposition rate
since some of the deposit is sputtered off, as well as causing a
considerable (and often undesired for microelectronic applications)
heating of the substrate by the intense ion bombardment. The latter
problem can be alleviated by using the supported discharge ion
plating process, where the substrate is no longer at the high negative
potential, the electrons necessary for supporting the discharge come

from an auxiliary heated tungsten filament. The high gas pressure
during deposition causes a reasonably uniform deposition on all
surfaces due to gas scattering (Fig. 39).
Figure 39. Triode ion plating using a DC supported discharge with an electron
beam evaporation source. (After Matthews and Teer.)
Reactive Ion Plating (RIP) Process. Reactive ion plating

(RIP) is very similar to the reactive evaporation process in the metal
atoms, and reactive gases react to form a compound aided by the
presence of a plasma. Since the partial pressure of the gases in
reactive ion plating is much higher (>10-2 torr) than in the ARE
process (>10-4 torr), the deposits can become porous or sooty. The
plasma cannot be supported at lower pressures in the simple diode
ion plating process; therefore, Kobayashi and Doi[59] introduced an
auxiliary electrode biased to a positive low voltage (as originally

conceived for the ARE process) to initiate and sustain the plasma at
lower pressure (about 10-3 torr). This is no different from the ARE
process with a negative bias on the substrate reported[60] much earlier
by Bunshah, which was designated by him as the biased ARE
process (BARE).
Another variation of reactive ion plating using a triode con-
figuration[61] involves injection of electrons into the reaction zone
between the electron-beam-heated evaporation source and the nega-
tively biased substrate from a heated tungsten filament transversely
to the metal vapor path. These low energy electrons are pulled across
the reaction zone by a positively biased anode located opposite to the
cathode. The arrangement is very similar to that shown in Fig. 33
except for the use of an electron-beam-heated evaporation source,
and is also very similar to triode sputtering. This adds versatility, as
well as complexity to the process through the addition of another
process variable.
Murayama[62] used an electron-beam-heated source with a
negatively biased substrate and RF activation of the reactants by
means of a coil electrode of aluminum wire in the reaction zone to
deposit oxide and nitride films.
5.0 CHEMICAL VAPOR DEPOSITION (CVD)
5.1 Introduction
Chemical vapor deposition (CVD) processes are widely used

in industry due to their versatility for depositing a very large variety
of elements and compounds covering a wide range from amor-
phous deposits to epitaxial layers having a high degree of perfection
and purity.
CVD can be defined as a process in which the gaseous
chemical reactants are transported to the reaction chamber, activated
thermally (conventional CVD) or by other than thermal means (plasma
assisted CVD or laser induced CVD), in the vicinity of the substrate,

and made to react to form a solid deposit on the substrate surface. It
is possible to deposit films of uniform thickness and low porosity
even on substrates of complicated shape in this process.
A major area for utility of CVD is in microelectronic applica-
tions, such as gate insulating layers, passivation layers, oxidation
barriers, doped epitaxial layers of silicon, polycrystalline silicon
layers, etc. CVD techniques are also extensively employed for pro-
tective coatings for a variety of operating environments where, for
example, protection is required against wear, erosion, and high-
temperature oxidation.
5.2 Conventional CVD (Thermal CVD)
In conventional CVD (CCVD), the gaseous reactants are acti-

vated thermally in the vicinity of the heated substrate, and react to
form a film on the substrate. This technique is referred to simply as
CVD in many cases. A simple schematic representation of CVD is
shown in Fig. 40.
Figure 40. Schematic of the chemical vapor deposition (CVD) process.

Various heating sources are used in CVD.

1. Hot plate. The substrate is in direct contact with the
hot plate which is either resistively or inductively
heated.
2. Radiant heat. The substrate is heated by a thermal
radiation technique or optical technique (tungsten
filament lamp or laser).
3. Heating of a conductive substrate. Conductive
substrates can be heated resistively or by RF
induction.
5.3 Classification of CVD Reactions
CVD reactions fall into four general categories.

1. Thermal decomposition reactions (pyrolytic reactions). This
reaction is characterized by
AX(g) → A(s) + X (g)
where AX is a gaseous compound, A a solid material, and B a gaseous

reaction product.
Some examples of these reactions are:
B2H6(g) → B(s) + 3H2(g)
SiH4(g) → Si(s) + 2H2(g)
W(CO)6(g) → W(s) + 6CO(g)
2. Reduction reactions. In this reaction, a gaseous compound is

reduced by a reducing agent (usually hydrogen).
2 AX(g) + H2(g) → 2A(s) + 2HX(g)
Examples of some reduction reactions are given below:
2BCl3(g) + 3H2(g) → 2B(s) + 6HCl (g)
SiCl4(g) + 2H2(g) → Si(s) + 4HCl (g)
3. Displacement reactions. These reactions are also known as

exchange reactions. In the molecule AX, X is replaced by another
element B.
AX(g) + B(g) → AB(s) + X(g)
Some representative reactions are:
Zn(g) + H2S(g) → ZnS(s) + H2(g)
SiCl4(g) + CH4 → SiC(s) + 4HCl (g)
CrCl2(g) + Fe(s) → Cr-Fe alloy + FeCl2(g)
4. Disproportionation reactions. In these reactions, the oxida-

tion number of an element both increases and decreases through the
formation of two new species. Some typical examples are:
2GeI2(g) → Ge(s) + GeI4(g)
TiCl2(g) → Ti(s) + TiCl4(g)
Several types of reactions can be involved simultaneously in

some CVD coating processes. An example of these reactions is:
2AlCl3(g) + 3CO2(g) + 3H2(g) → Al2O3(s) + 3CO2(g) + 6HCl(g)
In this reaction, water produced from

Co2(g) + H2(g) → CO(g) + H2O(g)
is used to form Al2O3 by the following reaction
AlCl3(g) + 3H2O(g) → Al2O3(s) + 6HCl(g)
5.4 Rate-Limiting Steps
The sequence of events in a CVD process as follows:[63]

1. Diffusion of reactants to the surface.
2. Adsorption of reactants at the surface.
3. Surface events, such as chemical reaction, surface
motion, lattice incorporation, etc.
4. Desorption of products from the surface.
5. Diffusion of products away from the surface.
Among these steps, the slowest one is the rate-determining
step. The rate-limiting step is mainly determined by the process
parameters. The most important rate-limiting steps in the CVD
process are mass transport control, and surface kinetics control. The
latter produces uniform deposits on complex shaped substrates.
5.5 Reactors
There are two kinds of reactors most frequently used in the

CVD processes, hot wall reactor (Fig. 41) and cold wall reactor
(Fig. 42). In the former reactor, the reactor tube is surrounded by a
tube furnace making the substrate, and the reactor wall to be the same
temperature. A large number of substrates can be coated in this type
of reactor. A major drawback of this type of reactor is, deposition on
the reactor wall and possible contamination in the system from
chemical reactions between the reactor wall, and the vapor due to the
high temperature of the reactor wall. Therefore, the hot wall reactor
is ideal for the case where the reaction is exothermic, since the high
wall temperature prevents undesirable deposition on the reactor

walls. III–V and II–VI type semiconductors have been successfully
prepared in the hot wall reactor.
In the cold wall reactor, only the susceptor where the substrates
are placed is intentionally heated by RF induction, or high radiation
lamps. This type of reactor is predominantly used for the deposition
reaction which is endothermic, such as Si deposition from the ha-
lides. Since the substrates have a higher temperature than the reactor
wall, the reaction will proceed most readily on the hot surface of the
substrate. In this reactor type, contamination due to the interaction
between the reactor wall and the vapor can be greatly reduced. Very
frequently, the walls are water-cooled to further prevent deposition
on the wall or reactions between walls and vapor.
Figure 41. Hot wall CVD reactor.
Figure 42. Cold wall CVD reactor.

5.6 Low Pressure CVD (LPCVD)
Low pressure CVD technology is widely used in the semicon-

ductor industry due to several advantages over conventional atmo-
spheric-pressure CVD technique. By operating at lower pressure to
increase diffusibility in the gas phase, and increasingly subject the
system to surface kinetics control, uniformity of deposition is en-
hanced. The mass transfer rate and the surface reaction rate are
generally of the same order of magnitude at normal atmosphere
pressure (CCVD), while the mass transfer rate is much higher than
the surface reaction rate at lower pressure, i.e., 0.5–1 torr (LPCVD),
and thereby the rate-determining step is the surface reaction. The
transfer rates of gaseous reactants and reaction products are in-
versely proportional to pressure. If the pressure is reduced from 760
torr to 0.5–1 torr, diffusibility increases by a factor of 1000 which is
only partially offset by the increase in thickness of the boundary
layer (by the square root of pressure).[64] At this low pressure, mass
transfer cannot be the rate limiting step, and the deposition rate is
mainly controlled by surface reactions, resulting in uniform film
thickness, and properties over extended surfaces with better step
coverage and conformity, and good structural integrity with fewer
pinholes.
Another advantage of LPCVD occurs because the mean free
path is very large at lower pressures; thus wafers can be stacked on
edge instead of lying flat, and thereby a large number of wafers can
be loaded and deposited in the same run.
Polycrystalline Si films from SiH4, Si 3N4 films from SiH2Cl2
and NH3, and SiO2 films from SiH2Cl2 and N2O have been success-
fully prepared using this technique.[65][66]
5.7 Plasma Assisted CVD (PACVD)/Plasma Enhanced CVD

(PECVD)
Plasma assisted chemical vapor deposition (PACVD) can be

defined as a process in which the constituents of the vapor phase
react to form a solid film assisted by an electric discharge. In the
PACVD technique, the gas molecules are mainly dissociated by

electron impact generating very reactive neutral, radical, and ion
species. These reactive species arrive on a surface and react with
each other via an ionic or free radical mechanism in the film forming
process. Since the gas molecules are activated by the energetic
electrons instead of thermal energy, the reaction temperature can be
easily reduced. Films can be deposited at temperatures typically less
than 300°C. Furthermore, the inherent limitations of conventional
thermodynamics and of chemical availability in CCVD are elimi-
nated in plasma activation[67] due to the non-equilibrium nature of
the glow discharge plasma. A schematic of a microwave plasma
CVD apparatus is shown in Fig. 43.
Figure 43. Schematic of a radial-flow plasma enhanced CVD reactor.
Neutral radicals are believed to be the major deposition agents

among the reactive species generated in the plasma due to the
following two effects. First, the dissociation energy is usually lower
than the ionization energy for many gas molecules and therefore the
generation rate for radicals is generally greater than that for ions.
Second, positive ions may drift toward any surface and recombine
with electrons while neutral radicals may have more chance to stay in
the plasma and have a longer lifetime.
The number of neutral radicals is determined by a set of

variables, which are the glow discharge power, electrode spacing,
gas collision mean free path, and ion diffusibility. These variables
determine the electron temperature (energy states of the electrons in
the plasma), and thereby the radical generation rate. The radical
generation rate is roughly proportional to e-∆E/kTe, where ∆E is the
dissociation energy of gas molecules, Te is the electron temperature,
and k is the Boltzmann constant. This relation indicates that the
number of gas radicals which have higher dissociation energy will
change more with the electron temperature than those of the species
with lower dissociation energy.[66] These changes of radical ratio is
the important factor determining the film stoichiometry.
Even though ions do not contribute greatly to the film-forming
process,[67][68] the ions impinging on the film during growth signifi-
cantly affects the physical properties of the film.[69] The ionic bom-
bardment energy is determined by several factors, such as discharge
power, gas type and pressure, target bias, and frequency (RF). At
lower pressure (less than 0.05 torr), and low frequency (<< 1 MHz),
the ions can bombard onto the surface with the full energy available
to them, since the ions can follow the RF field and do not experience
appreciable scattering. The ionic bombardment energy is an average
plasma potential as the pressure and the frequency increase.
The substrate temperature still plays an important role in the
PACVD process, through the activation energy for the chemical
reaction is provided mainly by the glow discharge. During film
growth, the absorbed radical has to diffuse to a stable site to become
part of the growing film. This radical mobility on the surface of the
substrate is strongly affected by substrate temperature. The radicals
on the surface obtained more energy and diffuse to the stable site
easily at high substrate temperature producing a denser film, while,
at low temperatures the diffusion of adatoms on the surface is much
retarded and thereby the film has more defects and a lower density.
Furthermore, the stoichiometry of the film can be influenced by
substrate temperature. In the case of plasma CVD silicon nitride, the
content of hydrogen in the grown film can be reduced greatly by

increasing the substrate temperature.
5.8 Advantages and Disadvantages of PACVD
There are several advantages of PACVD processes over con-

ventional CVD processes:
1. Ability to deposit films at much lower temperature.
2. Almost unique method to prepare the heavily
hydrogenated amorphous silicon films for solar cells.
3. Good adhesion of the films to their substrates and
the bond strength.
4. Higher deposition rates than by the CCVD
technique.
The present limitations of PACVD processes are:
1. More defects in the films and a lower density of the
film compared to a high temperature deposited film.
2. Difficulties of deposition of pure materials.
3. Extreme difficulty in controlling stoichiometry.
5.9 Advanced CVD Techniques
Hot-Filament CVD. Yasui, et al.,[70] prepared high-quality

silicon nitride films with a low concentration of hydrogen at high
deposition rates by the hot-filament CVD technique. In this tech-
nique, the hot filament (W filament), which is heated to very high
temperatures (such as 2400°C) is placed close to the substrate (8 cm).
They deposited films at the deposition rate which was one order of
magnitude higher with a smaller activation energy of the growth rate
than that of CCVD technique using the same gases. The same
techniques are now used to grow diamond films.
Laser Induced CVD (LCVD). Since the substrate is bom-

barded by charged and energetic particles, as well as by high energy
radiation in the PACVVD process, this technique can introduce
deleterious effects for radiation sensitive electronic devices. Re-
cently, laser induced CVD processes have been attracting much
attention. In this process the reaction energy is selectively provided
by photons.
There are two types of processes in LCVD, pyrolytic LCVD
and photolytic LCVD. In pyrolytic LCVD, polyatomic gas molecules
are dissociated near a gas-substrate interface by localized heating of
the substrate which is exposed to the laser beam. This technique is
limited by the choice of the laser/source gas/substrate. The gas vapor
sources are required to be relatively transparent at the exciting laser
wavelength and the substrates strongly absorbing. In photolytic
LCVD, a molecule near the substrate is decomposed by means of a
photochemical reaction. Specific chemical bonds in polyatomic mol-
ecules can be broken selectively through the choice of the laser
wavelength. In this technique, photodissociation of the vapor source
fixes the maximum allowable wavelength of the laser, since only
radiation which is absorbed by the reactants can lead to a photo-
chemical reaction. LCVD techniques have been successfully em-
ployed to synthesize polycrystalline Si, and Ge,[71] aluminum ox-
ide,[72] and silicon.
6.0 MATERIALS DEPOSITED BY PVD AND CVD

TECHNIQUES
In addition to the deposition of simple element species, e.g.

refractory metals by CVD for large scale integrated circuits as well
as for chemical applications, Al for metallization and decorative
coatings, etc., a number of more complex materials are deposited
using PVD and CVD techniques particularly the plasma assisted
versions as illustrated in Table 4.[74]
Table 4. Examples of some compounds synthesized and deposition

rates obtained by ARE, reactive sputtering, and PACVD processes.
ARE Reactive PACVD

Compound (Å min-1) (Å min-1) (Å min-1)
Carbides 2000–3000 400–500 150–400

TiC, HfC, ZrC, VC
Nitrides 2000–3000 300–400 60–150
TiN, HfN, ZrN
Oxides 1000–2000 200–800 200–300
TiO2, ZrO2, Al 2O3, SiO2
Sulfides 1000–2000
TiS2, MoS2, MoS3
Novel Materials
Superconducting materials 1000–1500
Nb3Ge, CuMo6S 8
Photovoltaic materials 1500–2000 50–200
a-SiH, CuInS2
Optoelectronic materials 500–1000
Indium-tin-oxide,
zinc
Cubic BN 1000–1500
Diamond 1000 Å h-1
Carbon 300 200
SUGGESTED READING
1. Hurkmans, T., Hauzer, F., Buil, B., Engel, K., Tietema, R., “A New
Large Volume PVD Coating System Using Advanced Controlled
Arc and Combined Arc/Unbalanced Magnetron ABStm Deposition
Techniques,” Surface and Coatings Technology, 92:62–68 (1997)
2. Hogmark, S., Hedenquist, P., Jacoson, S., “Tribological Properties

of Thin Hard Coatings: Demands and Evaluation,” Surface and
Coatings Technology, 90:247–257 (1997)
3. Stals, L. M. M., Nesladek, M., Quaeyhaegens, C., “Current Industrial
Practice—Critical Issues in Hard PVD and PA-CVD Coatings,”
Surface and Coatings Technology, 91:230–239 (1997)
4. Pochet, L. F., Howard, P., and Saeed, S., “CVD Coatings: From
Cutting Tools to Aerospace Applications And its Future Potential,”
Surface and Coatings Technology, 94/95:70–75 (1997)
5. Strafford, K. N., “Tribological Properties of Coatings—Expectations,
Performance and the Design Dilemma,” Surface and Coatings
Technology, 81:106–117 (1996)
6. Prengel, H. G., Pfouts, W. R., and Santhanam, A. T., “State of The
Art in Hard Coatings For Carbide Cutting Tools,” Surface and
Coatings Technology, 102:83–190 (1998)
REFERENCES
1. Handbook on Deposition Technologies for Films and Coatings, 2nd

Ed., (R. F. Bunshah, ed.), Noyes Publications, Park Ridge, NJ
(1994)
2. Pierson, H. O., Handbook of Chemical Vapor Deposition, Noyes
Publications, Park Ridge, NJ (1992)
3. Deposition Technologies for Films and Coatings, 2nd Ed., (R. F.
Bunshah, ed.), p. 463, Noyes Publications, Park Ridge, NJ
4. Schiller, S., Heisig, U., and Goedicke, K., Thin Solid Films, 40:327
(1997)
5. Teer, D. B., Surface and Coatings Technology, 36:901 (1988)
6. Musil, J., et al., Surface and Coatings Technology, 39/40:270 (1990)
7. Sproul, W. D., et al., Surface and Coatings Technology, 43/44:270
(1990)
8. (a) Kirchoff, V., et al., German Patent No. DD252, 205, (September
1986; (b) Schiller, et al., Surface and Coatings Technology, 61:331
(1993)
9. Wilson, R. G., and Brewer, G. R., Ion Beams with Applications to

Ion Implantation, John Wiley and Sons, New York (1973)
10. Nishimura, C., et al., J. Vac. Sci. Tech. A, 5(3):363 (1987)
11. Beale, H. A., and Bunshah, R. F., Proc. 4th Intl. Conf. on Vac. Met.,
p. 238, Iron and Steel Company of Japan, Tokyo, Japan (1973)
12. Glang, R., Handbook of Thin Film Technology, (L. Maissel and R.
Glang, eds.), McGraw-Hill, New York (1970)
13. Holland, L., Vacuum Deposition of Thin Films, Chapman and Hall,
London (1957)
14. Graper, E. P., J. Vac. Sci. Technol., 10:100 (1973)
15. Cocca, M. A., and Stouffer, L. H., Trans. Vac. Met. Conf., Am. Vac.
Soc., 203 (1963)
16. Morley, J. R., Trans. Vac. Met. Conf., Am. Vac. Soc., 186 (1963)
17. Bardos, L., Surface and Coatings Technology, 86/89:648 (1996)
18. Berghaus, B., German Patent No. 683,614 (1939)
19. Sabalev, L. P., et al., U.S. Patent No. 3,783,231 (January 1, 1974);
3,793,179 (February 19, 1974)
20. Dorodnov, A. M., Soviet Phys. Tech. Phys., 23:1058 (1978)
21. Osipov, V. A., et al., Soviet Rev. Sci. Inst., 21:1651 (1978)
22. Wroe, H., Br. J. Appl. Phys., 9:488–491 (1958)
23. Gilmour, A. S., Jr., Lockwood, D. L., Proc. IEEE, 60(8):977–991
(1972)
24. Snaper, A. A., Arc Deposition and Apparatus, U.S. Patent No.
3,625,848 (1971)
25. Snaper, A. A., Arc Deposition and Apparatus, U.S. Patent No.
3,836,451 (1974)
26. Sablev, L. P., Apparatus for Vacuum Evaporation of Metals Under
the Action of an Electric Arc, U.S. Patent No. 3,783,231 (1974)
27. Sablev, L. P., Apparatus for Metal Evaporation Coating, U.S.
Patent No. 3,793,179 (1974)
28. Sanders, D. M., Handbook of Plasma Processing Technology, (S.
Rosnagel, J. J. Cuomo, W. D. Westwoods, eds.), p. 419, Noyes
29. Martin, P. J., et al., Thin Solid Films, 153:91 (1987)
30. Boxman, R. L., Martin, P. J., Sanders, D. M., Handbook of Vacuum

Arc Science and Technology, Noyes Publications, Park Ridge, NJ
(1995)
31. Martin, P. J., Netterfield, R. P., Kinder, T. J., Thin Solid Films, 193/
194:77 (1990)
32. Vasin, A. I., Dorodnov, A. M., et al., Sov. Tech. Phys. Lett., (English
Translation of Pis’ma Zh. T. Fiz.), 5:23–24 (1979)
33. Ehrich, H., Hasse, B., et al., Proc. 8th Intl Conf. On Discharge
Appl., pp. 591–592, 596, Essen Univ. (1985)
34. Dorodnov, A. M., Kunetsov, A. N., et al., Sov. Tech. Phys. Lett.,
(English Translation of Pis’ma Zh. T. Fiz.), 5:418–419 (1979)
35. Ehrich, H., J. Vac. Sci. Technol. A, 6:143–138 (1988)
36. Meassick, S., Chan, C., and Allen, R., Thin Film Deposition
Techniques Using the Anodic Arc, to be published
37. Smith, H. M., and Turner, A. F., Appl. Opt., 4:147 (1965)
38. Sankur, H., DeNatale, J., Gunning, W., and Nelson, J. G., J. Vac.
Sci. Technol. A, 5:2869 (1987)
39. Cheung, J. T., and Madden, J., J. Vac. Sci. Technol. B, 5:705 (1987)
40. Cheung, J. T., Chen, J. S., and Otsuka, N., Proc. IRIS IR Detector
Specialty Meeting, Seattle, WA, (August 1987); This work was
followed by several other similar investigations presented at the
34th National Symp. of Am. Vac. Soc., Anaheim, CA (November
1987)
41. Dijkkamp, D., Venkatesan, T., Wu, S. D., Shaheen, S. A., Jisrawi,
N., Min-Lee, Y. H., McLean, W. L., and Croft, M., Appl. Phys. Lett.,
51:619 (1987)
42. Wu, X. D., Dijkkamp, D., Olgale, S. B., Ina, A., Chase, E. W.,
Miceli, P. F., Chang, C. C., Tarascon, J. M., and Venkatesan, T.,
Appl. Phys. Lett., 51:861 (1987)
43. Cheung, J. T., and Sankur, H., Solid State and Materials Sciences,
16:63 (1988)
44. Greer, J. A., J. Vac. Sci. Technol., 10(4):1821 (1992)
45. Bunshah, R. F., Handbook of Deposition Technologies for Films

and Coatings, 2nd Ed., p. 179, Noyes Publications, Park Ridge, NJ
(1994)
46. Frach, P., Goedicke, K., Gottfried, Chr., Walde, H., and Hentsch,
W., Surface and Coatings Technology, 74/75:85 (1995)
47. Bunshah, R. F., Nimmagadda, R., Dunford, W., Movchan, B. A.,
Dernchishin, A. V., and Chusanov, N. A., Thin Solid Films, 54:85
(1978)
48. Rosa, B., and Schultz, L., Appl. Phys. Lett., 53(16):1557 (1988)
49. Humphreys, R. G., Satchell, J. S., Chew, N. G., and Edwards, J. A.,
Appl. Phys. Lett., 54(1):75 (1989)
50. Karim, A. A., Deshpandey, C., Doerr, H. J., and Bunshah, R. F.,
Thin Solid Films, 172:111–112 (1989)
51. Bunshah, R. F., Thin Solid Films, 107:21 (1983)
52. Bunshah, R. F., U.S. Patent No. 37,791,852 (1972)
53. Bunshah, R. F., and Raghuram, A. C., J. Vac. Sci. Technol., 9:1385
(1972)
54. Bunshah, R. F., and Deshpandey, C. V., Physics of Thin Films, (J. L.
Vossen and H. Francombe, eds.), (1987)
55. Goedicke, K., Scheffel, B., and Schiller, S., Surface and Coatings
Technology, 68/69:799 (1994)
56. Reschke, J., Goedicke, K., Schiller, Surface and Coatings
Technology, 76/77:763 (1995)
57. Neumann, M., Morgner, H., Staach, S., SVC Conf., (1996)
58. Mattox, D. M., Deposition Technologies for Films and Coatings,
(R. F., Bunshah, ed.), p. 320, Noyes Publications, Park Ridge, NJ
(1994)
59. Kobayashi, M., and Doi, Y., Thin Solid Films, 54:57 (1978)
60. Bunshah, R. F., New Trends Mater. Process, Am. Soc. Met., p. 200,
(1978)
61. Matthews, A., and Teer, D. G., Thin Solid Films, 80:41 (1981)
62. Murayama, Y., J. Vac. Sci. Technol., 12:818 (1975)
63. Kern, W., and Ban, V. S., Thin Film Processes, (J. L. Vossen and W.
Kern, eds.), p. 267, Academic Press, New York (1978)
64. Kern, W., and Ban, V. S., Thin Film Processes, (J. L. Vossen and W.
Kern, eds.), p. 274, Academic Press, New York (1978)
65. Yee, K. K., Intl. Metals Rev., 1:19 (1978)
66. Dun, H., Pan, P., White, F. R., and Douse, R. W., J. Electrochem.
Soc., 178(7):1555 (1981)
67. Coburn, J. W., and Chen, M., J. Applied Physics, 51:3134 (1980)
68. Zesch, J. C., Lujan, R. A., and Deline, V. R., J. Non-Cryst. Solids,
35–36:273 (1980)
69. Jansen, F., Plasma Deposited Thin Films, (J. Mort and F. Jansen,
eds.), p. 4, CRC Press Inc., Boca Raton, FL (1985)
70. Yasui, K., Katoh, H., Komaki, K., and Kaneda, S., Appl. Phys. Lett.,
56(10):898 (1990)
71. Andreata, R. W., Abel, C. C., Osmundsen, J. F., Eden, J. G., Lubben,
D., and Greene, J. E., Appl. Phys. Lett., 40(2):183 (1982)
72. Minakata, M., and Furukawa, Y., J. Electron. Matter., 15(3):159
(1986)
73. Peters, J. W., Gebhart, F. L., and Hall, T. C., Solid State Technol.,
p. 121 (1980)
74. Bunshah, R. F., IEEE Transactions on Plasma Science, 18:846
(1990)
Thermal Spraying and Detonation Gun Processes 77
Thermal Spraying and

Detonation Gun Processes
Otto Knotek
1.0 HARD COATING WITH THERMAL SPRAYING

PROCESSES
The earliest records for thermal spraying are patents by the

Swiss engineer M. U. Schoop, originating in the early 1900s. At first
lead and tin wires were melted in a welding torch by the energy of an
acetylene/oxygen flame. The torches then were modified for the use
of powdered materials. The wire-arc spraying process was patented
around 1908, also by Mr. Schoop, making the deposition of more,
and various metals possible. Due to the development of thermal
plasmas, and the increasing demand of high temperature and wear
resistant materials and coating systems, the thermal spraying tech-
nologies expanded in the 70s (Fig. 1). Since the 80s the major
developments are leading towards increasing particle velocities.
Nowadays thermal spraying includes all processes where coat-
ing materials are partially, or totally molten, either inside or outside
of a spray torch, and the liquid or solid particles are deposited onto a
surface, where the arriving droplets form a coating.
77
78
Figure 1. Historical development of thermal spraying equipment, processes, and materials.

One of the advantages of thermal spraying is the fact that the

molten, or partly molten coating material droplets are deposited onto
a substrate material without melting it, since only slight heating of
the substrate occurs.
Therefore, usually no influence on heat treatments, chemical
compositions, etc., is observed due to the moderate heat input by
thermal spray coating processes, and substrate temperatures seldom
exceed 150°C.
The tendency of substrate distortion due to substrate heating is
considerably lower than for hard facing welding processes (arc, oxy/
fuel, electron beam, laser, etc.), where the substrate material is
partially molten. This may be the reason for a decrease in certain
material properties. However, these fusion processes do have a
metallurgical bond, which increases the bond strength to values that
thermal spray processes (without posttreatment) do not achieve.
The coating material may be heated and accelerated through
various types of energy release. Figure 2 illustrates the major catego-
ries of thermal spray processes.
While the particles are accelerated in a gas jet (flame, plasma),
they are heated up, and partly or totally molten, depending on their
dwell time, which is a function of the average particle size distribu-
tion, and temperature distribution within the jet as well. During the
flight the particles may interact with the surrounding medium, e.g.,
oxidation may occur due to their high temperature at their active
surface when sprayed in air.
In the electric wire-arc spray process, however, the sprayed
materials are wires, which are melted by an electric arc. Therefore,
the accelerated droplets are always in a molten state, but their
temperature starts to decrease immediately after they are formed
from the wire tips.
The liquid spray particles form a splat when impacting at the
substrate due to their temperature and high kinetic energy, i.e., they
are flattened, fractured, spread, and quenched within a very short
period of time. Due to the radial spreading, and the increase in
surface area, and the rather small mass, a rapid quenching of the
particles takes place. The necessary time for solidification is be-
tween about 10-8 and 10-6 seconds.
80
Figure 2. Thermal spray processes.

Even when the temperature of the spray particles is higher than

the solidus temperature of the substrate material, no fusion with the
substrate material occurs since the cooling rate is very high (splat
cooling).
Any thin oxide layer around a molten particle will be dispersed
and entrapped in fragments upon impact and may be seen in metal-
lurgical cross sections as oxide inclusions (stringers) between the
typical lamellae structures formed by various splats.
The spray particles impacting upon other, already solidified
ones are rapidly cooled also, since the periods between impacting
particles is about 10-2 seconds and therefore much higher/longer than
the time necessary for cooling.
Even easy inflammable materials such as wood, or plastics
may be used as substrate materials due to the low interface tempera-
tures ranging between 100–280°C.
1.1 The Bonding Mechanisms of Thermally Sprayed Coatings
During the rapid solidification of the spray particles there is a

close contact between particles and/or the substrate surface. This
leads to bonding effects due to mechanical interlocking, adhesion,
diffusion, chemical reactions, and sometimes partial fusion of the
contact surfaces. These bonding effects permit the formation of a
continuous coating layer as illustrated in Fig. 3. Mechanical inter-
locking is the main mechanism of thermal spray coating adhesion/
adherence, if the substrate temperature is kept within the above
mentioned range. The bond between impacting particles and the
substrate surface is established to a high degree through the arriving
droplets liquid flowing and solidifying around the substrate asperi-
ties/roughness. Usually these asperities are formed by mechanical
means such as abrasive blasting, grit blasting, or rough turning,
which also activate and clean the surfaces prior to coating. The
quenching stresses within the spray particles increase the interlock-
ing effects. In Fig. 4 the possible bonding interactions are illustrated.
82
Figure 3. Schematic of a typical structure during thermal spray coating.

Figure 4. Bonding interactions within a spray coating.
The adhesive bonding mechanisms are effective at microcontact

areas between particles and/or substrate. These forces correspond to
the atomic attractive forces within crystalline solid matter, since the
particles approach one another as close as atomic levels. Depending
on the type of atomic bonding, on the one hand van der Waals forces
(physisorption) and on the other hand an exchange of valence elec-
trons (chemisorption) may take place.
Between spray particles, as well as between spray particles and

the substrate, melting and diffusion within micro dimensions may
occur due to the interface temperature being somewhat higher than
the melting point of the substrate, despite the high cooling rates.
The industrial applications of thermally sprayed coatings largely
depend on the bond quality between the coating and substrate.
Therefore, a thorough substrate surface preparation is a necessity.
Contaminants such as rust, scale, grease, moisture, etc., must be
removed from the surface. After cleaning, usually a roughening of
the substrate surface follows to ensure coating adhesion. Common
methods for surface roughening, which are often combined, are:
• Dry abrasive grit blasting.
• Machining or macro roughening.
• Applying a bond coat.
The thermal spray coating process should start as soon as
possible after the surface preparation is completed, since the pre-
pared surface is very active, and oxidation, recontamination, etc.,
should be avoided. In some cases, preheating of the substrate materi-
als may be necessary just prior to the coating in order to drive
moisture from the substrate. For this purpose substrate preheating in
air up to 150–200°C for approximately 60 seconds is sufficient.
2.0 PROCESSES
In general, the thermal spraying processes used for hard coat-

ings may be divided into 2 basic processes: combustion spraying,
and plasma spraying.
In this chapter the equipment for thermal spray technologies is
described.
2.1 Combustion Spraying
Low Velocity, Flame Spraying. Being the oldest of the spray-

ing processes, flame spraying is well established. It uses the thermal
energy released through an oxygen-fuel flame to melt feedstock

materials in powder or wire form. The acceleration of the particles in
the direction of the substrate is due to the exiting combustion gases
or a dispersing gas (e.g., compressed air). Flame spraying is practi-
cally applied using:
• Wire or rod feedstock
• Powder material without thermal post treatment
• Powder material with immediate thermal treatment (one
step process)
• Powder material with a thermal post treatment (two
step process)
It has rather low costs and permits deposition rates of
5 to 8 kg/h, depending on the process, feedstock material, etc. A
schematic of a powder-flame spraying process is given in Fig. 5.
Figure 5. Schematic of a flame spraying process for powder material.

The most commonly used fuel gases yield different flame

temperatures. The triple-bond of acetylene is chemically unstable. It
is rather reactive and therefore yields the highest flame temperature,
and the highest ignition velocity. Another advantage of acetylene is
that it may be used as a fuel gas for reducing, neutral, and oxidizing
flames.
With a surplus of oxygen the deposition rate increases, but so
do the oxide inclusions in the coating. A reducing flame stoichiom-
etry may prevent the decomposition of certain alloy components. An
acetylene flame also prevents the formation of boron hydride, which
would cause a decrease of boron, and therefore a decrease of hard
phases within the coating.
An overview of the achievable flame temperatures for various
fuel-oxygen mixtures is given in Fig. 6 and some chemical and
physical properties for the main types of fuel gases for flame spray-
ing processes is listed in Table 1.
Figure 6. Flame temperatures for various fuel-oxygen mixtures for (a) acety-
lene, (b) methane, (c) methylacetylene, (d) propylene, (e) propane.
Table 1. Chemical and Physical Properties of the Main Fuel Gases

used for Flame Spraying Processes
The spray particles gain their kinetic energy through the pres-
sure of the process gases. An overview of flame spray processes,
feedstock materials, and applications is given in Fig. 7.
The selection of the substrate material is of great importance
for all flame spraying processes and especially for the application of
self fluxing alloys requiring a thermal post treatment, i.e., fusion.
Here the substrate materials have to withstand the thermal treatment,
and may not undergo great changes in chemical composition and
properties. Especially steels, e.g., the martensitic hardening, nickelless
13% chromium steels, and high carbon heat-treated steels are caus-
ing problems, since the post heat-treatment creates temperatures that
enforce sudden changes of structures with changes of volume. Sub-
strates made of various types of steels, as well as copper and copper
alloys have to be preheated just prior to the coating process, in order
to compensate the differing thermal expansion behaviors/coefficients.
88
Figure 7. Processes, feedstock materials, and applications for flame spraying.

The application of these flame spray coatings are mainly wear

protection in various industrial fields, for example:
• Steel industry
• Mining industry
• Cement and concrete industry
• Chemical industry
• Glass producing industry etc.
Typical examples are chutes, milling and grinding devices,
seals, valve seats, glass molds, etc.
2.2 Combustion Flame Spraying
High Velocity, Discontinuous, Detonation Gun Processes.

The starting point for the development of the high velocity flame
spraying processes was the detonation gun coating process by Union
Carbide, known since the mid 50s for the best coating qualities/
properties. With the detonation gun process (Fig. 8), also known as
D-gun, the reaction gases (acetylene and oxygen), and a carrier gas
for the powdered coating material are introduced into the combus-
tion chamber, mixed and ignited by a spark from a spark plug. The
powder material is heated by the detonation energy release, and
accelerated up to 900 m/s when it hits the surface of the component
to be coated. After each combustion process (detonation) a flushing
gas (argon, nitrogen) cleans the combustion chamber before a new
cycle starts. There are about 4 to 8 detonations per second. The very
high kinetic energy of the particles upon impact is responsible for
very good contact between spray particles, and partial welding/
fusing of particles and substrate, being the reason for the very dense
coatings and the high bond strengths achieved by D-gun processes.
Figure 8. Thermal spray coating with the detonation gun process.
High Velocity, Continuous, HVOF Processes. The High

Velocity Oxy-Fuel (HVOF) thermal spray processes are rather young
technologies, and industrially applied just within the last years. The
first of these processes was the Jet-KoteTM process, introduced by
Browning Engineering. The major difference in comparison to the
principle of conventional flame spraying, i.e., the combustion of a
oxygen/fuel gas-mixture, is the confined combustion process under
higher pressure in a combustion chamber within a spraying torch,
Fig. 9. The combustion gases are expanding through a nozzle sys-
tem, and the injected powder particles gain significantly higher
velocities due to the high gas velocities.
Figure 9. HVOF thermal spray system.

Compared with the D-gun, the combustion process in HVOF

thermal spraying is a continuous one. The combustion in the cham-
ber takes place under higher pressure (up to 8 bar), and therefore the
fuel gases used have to be stable under pressure, e.g., propane, and
propylene. Also hydrogen, heptane, and kerosene may realize the
necessary large gas flows through higher gas pressures, and there-
fore, in HVOF the flame temperatures, and energies of fuel gases
used is not as high as that of acetylene/oxygen flames, since only a
few types of HVOF systems work with acetylene, requiring that very
high security standards have to be met.
The various HVOF systems differ in the type of powder injec-
tion, the design of the combustion chamber, and the nozzle geom-
etry. The powder feedstock is usually axially injected into the hot
exhaust gases, or into the combustion chamber, depending on the
type of HVOF-system and nozzle system. Heating, and acceleration
of the particles takes place within the barrel of the torch, and in the
free jet outside. The flame temperature is about 3000°C, and the
velocity of the exiting and further expanding gases may be as high as
1200 m/s. The powder particles can reach velocities of up to 600 m/
s, depending on the density, shape, and size of the feedstock mate-
rial. Typical for these HVOF technologies are the visible shock
diamonds in the free jet, showing that the Mach number of this
compressible fluid flow has been equal or greater than 1 in this
process.
Due to the high particle velocities, and the rather moderate
temperatures, these processes are preferably used for coating materi-
als that tend to decompose at higher temperatures. This is the reason
for their main field of application, the deposition of hard metals/
cemented carbides such as WC/Co and Cr3C2/NiCr, since the density
and the wear resistance of these coatings is extremely good/high.
Also the FeNiCo-alloys, i.e., HastelloysTM , StellitesTM, Tribaloys TM,
and self-fluxing Ni-based alloys can be efficiently deposited.
The main field for HVOF are applications for abrasion and
sliding wear. Components like nozzles of water jet cutting tools,
rolls for paper and foil producing industries, sliding areas of pressing
irons, valves and pumps in petrochemical applications, and mechani-
cal seals are only some examples.
2.3 Plasma Spraying
Plasma spraying is the most flexible/versatile thermal spray

process with respect to the sprayed materials. The high temperatures
of plasma spray processes permit the deposition of coatings for
applications in areas of liquid and high temperature corrosion and
wear protection, and also special applications for thermal, electrical,
and biomedical purposes.
An electric arc, started with a high frequency spark, is the
energy source for plasma spraying. The arc in DC plasma gun burns
between a cylindrical tungsten cathode and radial, concentric copper
anode which is also the nozzle. Into this nozzle the arc gases, usually
argon Ar, helium He, or nitrogen N2 or mixtures of these, are
introduced/fed. When these gases pass the arc, there are reactions
between the electrons from the cathodic source, and the gas mol-
ecules and/or atoms, causing a dissociation or ionization of the gas,
and thus, creating the plasma state.
The gases are heated as they pass over and through the arc, they
expand radially and axially, and accelerate through the nozzle exit.
Then the exothermic recombination of atoms and molecules of the
plasma gases follows, and the temperature in the core region can be
higher than 10000 K due to the released heat and radiation. Thermo-
electric and electromagnetic effects increase the energy density
through confining the core region of the arc. The temperature distri-
bution of a plasma jet is shown in Fig. 10.
The energy content, and the temperature of the plasma depends
mainly upon the physical properties of the plasma gas or the mixture
of plasma gases used. The enthalpy of the gases indicate their heating
potential, and the enthalpy of monatomic gases (Ar, He) is signifi-
cantly lower than that of diatomic gases (N2, H2), and the influence
on the achievable temperatures is reverse.
The feedstock material (mainly powder) is fluidized with a
carrier gas and injected into the high energy plasma jet. The point of
injection, depending on the design of the plasma burner, may take
place externally or internally through powder ports at various angles
to the jet axis.
Thermal Spraying and Detonation Gun Processes
Figure 10. Temperature distribution and geometry of a plasma jet.

93
The particle velocities may reach 300 to 700 m/s, depending

on the process design, the plasma gases, and the feedstock material.
The superposition of radial and axial velocities combined with prop-
erties of the spray powder (size distribution, density, etc.) leads to
various trajectories of the powder particles through the plasma jet
with various effects/results concerning their melting behavior.
The differing symmetry axes (angle β in Fig. 10) of plasma and
powder cone is a function of powder properties and injection condi-
tions. The determination of an angle θ > 0 increases the dwell time of
the particles in the hot core region of the plasma jet, thus improving
the melting behavior of materials with high melting points. Negative
angles θ (θ < 0) will inject low melting point metals, or plastics into
the outer and therefore cooler regions of the plasma jet. The interde-
pendencies of the most important plasma spray parameters are shown
in Fig. 11.
Figure 11. Interdependencies of the major plasma spray parameters.

The powder mass flow rate, the velocity of the plasma torch,
and the number of passes determine the final coating thickness
(ca. 50–250 µm).
Geometric conditions of thermal spraying (“line of sight pro-
cess”) only permit the coating of easy accessible areas or rotating
surfaces without undercuts. Special types/designs of torches, how-
ever, permit internal coating processes of tubes with a diameter of
about 40 mm.
Process Variants of Plasma Spraying. The plasma spray
process can take place in different atmospheres at different pressure
levels and the resulting various process variants of plasma spraying
are listed in Fig. 12.
Atmospheric Plasma Spraying (APS), Low Velocity. Eco-
nomically, most important is atmospherical plasma spraying (APS)
in air. The powder particles can interact with the air atmosphere,
which may limit the choice of the spray materials, since the originat-
ing oxides are built into the coating. The major field for APS-
applications are coatings for wear- and corrosion (liquid and gas-
eous) protection, often based on oxide ceramic materials. Other
typical coating materials are metals, and some alloys especially
insensitive to oxidation. The porosity of APS-coatings is generally
between 1 and 5%. The spraying distance between substrate and
plasma torch is about 100 to 150 mm, depending on the materials
and spraying parameters.
Atmospheric Plasma Spraying: High Velocity, High Power
Plasma Spraying (HPPS, High Velocity Plasma Spraying). A
further development of APS is the HPPS process. This development
was necessary to meet the requirements/demands for better coating
densities and bond strengths of high melting ceramic materials. This
process variant leads to advantages when spraying with high power
levels (ca. 200 kW), higher gas flows, deposition rates, and efficien-
cies. A typical example for a HPPS application is the advanced
coating of chromium oxide Cr2O3 on printing rolls.
96
Figure 12. Process variants of plasma spraying.

Vacuum Plasma Spraying (VPS, Low Pressure Plasma

Spraying, LPPS). The coating process of VPS takes place in a
closed chamber with reduced pressure. The coating process, started
after the chamber is evacuated to pressures < 10-1 mbar and refilled
with an inert gas atmosphere, takes place at about 50 to 400 mbar.
To be able to constantly follow the working chamber pressure,
efficient pump systems have to be employed in order to remove the
steadily injected plasma gases.
Another advantage of VPS technology is the option to clean the
substrate surface especially from oxide layers and preheat the sub-
strate, both giving better adhesion.
Large turbulences of a VPS plasma jet take place at greater
distances from the nozzle exit only, which permits an elongation of
the hot core region. The decrease of the energy density can be partly
compensated through an increase of chamber pressure above
200 mbar, with consequences for particle velocity and dwell time.
The substrate material heats to higher temperatures since the
cooling possibilities and convective heat transfer to the surrounding
atmosphere is reduced. These higher component temperatures, how-
ever, lead to better coating bond strengths due to improved diffusion,
and may reduce the tensions caused by cooling off, when substrate
temperatures are kept below critical values.
The stand off distance for VPS-processes is usually between
250 to 300 mm, the porosity of VPS-coatings are below 1%. The
industrial application of VPS is done in fully automated systems.
The advantages of VPS processes compared to APS are:
• Through reducing the chamber pressureto about 10-1 mbar
the largest part of the oxygen is removed. Therefore, it
is possible to process materials (feedstock and substrate)
that are highly sensitive to oxygen, e.g., refractory
metals.
• Due to the lower pressure inside the chamber, the less
oxidized particles and substrates result in better bond
strengths, dense, and low oxide coatings.
There are, however, also some disadvantages of VPS

technologies:
• The rather high technical effort/expense for achieving
the vacuum conditions.
• The time necessary for one VPS process/coating is
longer than for APS and there is a deposition of
overspray materials inside the chamber that has to be
removed thoroughly.
• For some applications using low melting point materials
as substrates, there is insufficient convective heat transfer
within the chamber.
The VPS-coating of jet engine parts, or gas turbines, e.g.
turbine blades and turbine hot sections components with MCrAlY
materials has gained importance over the last years. The rather high
price for a single turbine blade justifies the expensive VPS-coating
process, typical coating thickness is about 100 to 150 µm. This en-
sures a sufficient high temperature corrosion resistance without
changing the geometry of these blades. In order to reach a high
efficiency/productivity of the coating system, the blades are mounted
on a type of conveyor belt, and coated one after another.
When using Argon in VPS/LPPS processes, it is possible to
clean electrically conductive substrate materials by applying the
reversed-transferred arc, i.e. when the substrate is cathodic/negative
relative to the plasma gun, most thin insulating oxide layers are
locally and rapidly (> 10-6 sec) vaporized due to cathodic arc dis-
charges, as shown in Fig. 13. This cleaning phenomena is one of the
major reasons for the commercial success of turbine airfoil coatings.
Inert-Gas Plasma Spraying (IPS). The main difference to
VPS is the fact that the spraying chamber after evacuation is flushed
with an inert gas (Ar, nitrogen) at 1 bar pressure. Oxidation reactions
of both substrate and coating materials are therefore suppressed.
Figure 13. Cleaning phenomena with VPS due to reversed transferred arc
with Ar-ions.
Shrouded Plasma Spraying (SPS). This plasma spray variant

has been developed to reduce the effort for chamber and pumping
system for less expensive applications. Similar to shrouded welding
processes an envelope of an inert gas, which is not ionized, protects
the plasma jet from the surrounding oxygen containing atmosphere
and also improves substrate cooling.
Controlled Atmosphere Plasma Spraying (CAPS). Com-
bining vacuum plasma spraying, and inert-gas plasma spraying in
one system leads to controlled atmosphere plasma spraying (CAPS).
For a pressure range from 50 to 400 mbar VPS-coatings are pro-
duced, at pressures around 1 bar IPS-coatings, and with pressures up
to 4 bar high pressure, plasma spray coatings are deposited.
The advantage of a CAPS system is the possibility to coat with
APS- and VPS-coating systems in one step, and additionally to use
materials with high vapor pressures. Furthermore, reactive plasma

spraying is possible.
2.4 Special Processes
Reactive plasma spraying is a further development of the VPS

process. One has to distinguish between spraying processes and
CVD (Chemical Vapor Deposition) processes.
Reactive Plasma Spraying. During the coating process the
feedstock material is injected into a normal plasma torch for VPS.
However, at the end of the nozzle a reactive gas is injected into the
plasma jet, chemically reacting with the melted powder particles.
This way, for example, titanium carbide TiC is obtained by process-
ing titanium powder and injecting methane (CH4), and due to the
very short reaction periods (dwell time of the particles) TiC is
dispersed within the coating. This fine dispersion of this hard phase
is advantageous for the wear behavior of the coated component. This
processing technology is in the state of research and development,
and the optimal process parameters are obtained in an low pressure,
inert-gas atmosphere.
Diamond Coatings by Plasma Spraying Processes. This
process is used for depositing diamond coatings on various substrate
materials. The thermal plasma is the tool for creating/obtaining the
necessary thermal conditions for the diamond synthesis.
The reasons for the great interest in depositing diamond coat-
ings are the extraordinary properties that this material combines:
high hardness, very good heat transfer coefficient, and high electric
resistivity. The fields for applications of diamond coatings are wide-
spread, e.g., from electronics (heat sinks), to tools.
The difference to the above mentioned reactive spray coating
process is that there is no powder material injected into the plasma
jet, just a hydrocarbon gas, e.g., methane. Through the high enthalpy
of the plasma flame the hydrocarbon is decomposed, and deposits in
diamond structures and also graphite structures onto the substrate
surface. Due to a high hydrogen content in the plasma flame, and the
high enthalpy atomic hydrogen is formed, acting as a selective
etching for graphite. This way it is possible to deposit pure diamond

coatings with coating rates of 80 µm/h onto high melting substrate
materials, since the substrate temperatures are about 1000°C. With
the substrate being cooled intensively, the best results for diamond
synthesis are obtained with substrates made of carbide forming
materials, e.g., Mo, Si, W, and Ti. It is also possible to use
cemented carbides as substrate materials, indexed tools and drills
coated with diamond are already used in practical tests. The typical
setup for diamond deposition from a thermal DC plasma jet is shown
in Fig. 14.
Figure 14. Schematic set up for diamond deposition with a thermal DC plasma
torch.
3.0 COATING MATERIALS/CONSUMABLES/

FEEDSTOCK MATERIALS
Nowadays there are a large number of modern coating materi-

als available for thermal spray surfacing technologies, Fig. 15. Basi-
cally one can say that any materials may be deposited as a coating by
thermal spray processes that are available in a “sprayable” form and
have a melting point, i.e., that do not decompose or evaporate.
Materials that do tend to chemically decompose may be treated in
order to use them as feedstock for spraying, e.g. coating them with
other materials.
Figure 15. Coating materials used in thermal spraying.
The biggest part of feedstock materials are in powder form.

Powder materials have to meet further requirements such as chemical
homogeneity, density, flow behavior, size and shape distribution, low
cost, etc. Due to the advances of the past years almost every material
is available for spraying applications. The standard techniques for
producing powder materials are water and gas atomization, crushing
and milling, and production through chemical techniques, e.g.,
solgel, agglomeration, or spray drying, etc.
3.1 Metals and Alloys
The processing of metals by thermal spraying is one of the

main fields. They are primarily used for corrosion protection of
wet corrosion in aqueous solution, gaseous corrosion at higher
temperatures, and as bond coat. Aluminum, nickel, copper, chro-

mium, zinc, and molybdenum have gained the biggest significance.
Of great significance are the refractory metals, which are processed
with VPS due to their high sensitivity to oxygen.
A large number of metallic alloys are also in use for thermal
spray applications. NiAl-, NiCr-alloys are preferably used as bond
coat. Due to an exothermal reaction in the Nickel-Aluminum alloy
partial fusion/welding between coating and substrate takes place,
improving the bond. The main reason for applying these materials as
bond coatings, however, is their ductility, which allows the reduc-
tion/mitigation of stresses between substrate and coating material.
Of great technical and economical significance are the MCrAlY-
alloys. These materials on nickel, cobalt, or iron basis prove to be
very resistant against high temperature corrosion. Applications for
these coatings are therefore preferably found in stationary gas tur-
bines, jet engines, and as intermediate layers.
By adding carbon, silicon, or boron to alloys based on Fe, Co,
and Ni, the resulting hard alloys play an important role when wear
resistance is required. This high wear resistance, often combined
with good corrosion resistance, is due to the formation of hard phases
(carbides, borides), which precipitate as primary/secondary carbides
or as binary/ternary eutectics. Table 2 lists some hard alloys and their
components.
The metallic hard alloys are realized mainly with chromium,
tungsten, molybdenum, and vanadium. Chromium is also applied for
reasons of corrosion protection. The metalloids carbon, boron, and
silicon form together with the hard compounds homogeneously
dispersed, hard phases in a ductile, eutecticly solidified matrix
(binder). Of great significance are hard alloys on Co-basis. These
stellite materials are, as well as Fe-hard alloys, mainly applied
through welding, or thermal spray processes.
Economical reasons, i.e., the substitution of expensive cobalt,
forced developments of Ni-hard alloys. Especially for flame spray-
ing with post-heat-treatment these materials succeeded, since they
have self-fluxing properties due to boron and silicon contents.
Table 2. Hard Alloys for Wear Protection
Further systems are CoMoSi (Tribaloy TM) and NiMo

(HastelloyTM). Tribaloys are mainly used for friction and wear
applications. They consist of hard, intermetallic laves-phases
embedded in a ductile matrix. Hastelloys are advantageous for
corrosion protection applications and the performance of nickel
in reducing corrosion media is improved through adding molyb-
denum. In case of oxidizing corrosion conditions NiMo-alloys
with added chromium have proven to be successful.
3.2 Hard Compounds, Hard Metals
The combination of a metallic hard phase with a metallic

binder is a hard metal. For thermal spraying these composite materi-
als/hard metals are mainly consisting of tungsten carbide (WC,
W2C), and chromium carbide (Cr3C2) as hard phases and cobalt and/
or nickel as a ductile binder phase metal, embedding the hard phases.
These materials are produced through powder metallurgy only, since
the carbides would decompose or dissolve when smelted/fused/
liquefied. The tungsten carbides are used with cobalt, or nickel as

matrices with contents between 8 and 20%. The binder phase used
for thermally sprayed coatings of chromium carbide is NiCr with
contents of 20–35%. Their advantage in comparison to WC/Co
coatings is their better stability under high temperature condi-
tions: WC/Co-coatings may be used in temperature ranges up to
550–600°C only, Cr 3C 2/NiCr-coatings are withstanding tempera-
tures up to 900°C.
3.3 Ceramics
Ceramic coatings are gaining importance in thermal spraying

due to their corrosion behavior, hardness, and temperature stability.
Ceramic hard materials are used as thermal barrier coatings (TBC’s),
as well as for wear and corrosion resistant coatings. Widely used are
aluminum oxide Al2O3, aluminum oxide plus titanium oxide
Al2O3xTiO2, stabilized and partially stabilized zirconium oxide ZrO2,
and chromium oxide Cr2O3, due to their properties like hardness,
dielectrics, and resistance against chemical attacks etc. Examples for
ceramic coating systems are thermal barrier coatings in jet engines,
wear protection in textile industry, and coatings on various large
rolls for purposes in the paper and printing industry.
Ceramic coatings have generally a high degree of porosity,
which may be improved by alloying/mixing of various oxides. In
order to obtain a good bond, the substrates are roughened and coated
with a bond coat, which is usually NiCr, but for ZrO 2 it is a type of
MCrAlY. These porous coatings are sealed with various plastic
materials especially to improve their corrosion resistance. All coat-
ing materials are summarized in Fig. 16.
Figure 16. General view of coating materials.

4.0 CONCLUSIONS
Thermal spraying is one of the advanced hard facing technolo-

gies for surface preparation and protection. The diversity of thermal
spraying processes used for hard coating is due to the variety of
applications and the required properties, as well as economic as-
pects. A systems analysis of thermal spraying for hard facing pur-
poses shows common advantages, e.g., little restrictions on the
shape, and the materials of the component to be coated. The limita-
tions are the relatively low bond strengths, and the degree of poros-
ity. The large potential of this universal technology has yet only been
used to a small degree so that new applications will be discovered
and developed.
REFERENCES
1. Smith, R. W., Novak, R., “Advances and Applications in U.S.

Thermal Spray Technology, I., Technology and Materials,” Powder
Metallurgy International, 23(3) (1991)
2. Smith, R. W., Novak, R., “Advances and Applications in U.S.
Thermal Spray Technology, II., The Market and R&D,” Powder
Metallurgy International, 23(4) (1991)
3. Budinski, K. G., Surface Engineering for Wear Resistance, Prentice
Hall, Englewood Cliffs, NJ, USA (1988)
4. “Thermal Spray Research and Applications,” Proceedings of
NTSC’90; (T. F. Bernecki, ed.), ASM International (1991); “Thermal
Spray Coatings: Properties Processes and Applications,” Proceedings
of NTSC’91, (T. F. Bernecki, ed.), ASM International (1992)
5. “Thermal Spray: International Advances in Coatings Technology,”
Proceedings of ITSC ’92, (C. C. Berndt, ed.), ASM International
(1992)
6. Knotek, O., Lugscheider, E., Eschnauer, H., Hartlegierungen zum
Verschleiss-Schutz, Verlag Stahleisen GmbH., Düsseldorf, Germany
(1975)
Structure/Property
Relationships For Hard
Coatings
Lars Hultman and Jan E. Sundgren
1.0 INTRODUCTION
Since the macroscopic properties of a material are strongly

dependent on its structure, studies of structure/property relationships
are among the most important issues in material science. Due to the
fact that there exists several structural levels ranging from the
elementary particle level up to engineering structures and covering
a size scale of more than 10 orders of magnitude, these studies can be
quite different from each other. However, most commonly the struc-
tural studies that are correlated to the physical properties of the
material involve studies of the atomic structure, and of the micro-
structure (phase, grain, and defect structure).
For thin films, the structure/property relationships are particu-
larly important to study since thin films usually are deposited far
from equilibrium conditions. This implies that thin films in most
cases contain a much higher content of lattice defects such as vacan-
cies, dislocations, grain boundaries, and second phase precipitates
than conventionally prepared bulk materials. Furthermore, since the
material flux during growth of a thin film often is directional, the
108
Structure/Property Relationships for Hard Coatings 109
microstructure, and thus also most of the physical properties of thin

films are often highly anisotropic. Many attempts have also been
made to make generalized models describing the microstructure as a
function of process parameters of thin films deposited from the vapor
phase. However, due to the complex nature of the deposition pro-
cess, and the large number of parameters imposing kinetic limita-
tions to the growth process, it is difficult to predict the microstructure
and the physical properties of the deposited films from a knowledge
of the used deposition conditions. Furthermore, the microstructure is
not determined by the deposition process alone, but is a function also
of the substrate constitution. Finally, the microstructure is seldom
homogeneous throughout the coating, but evolves with thickness.
Even though the term “hard coatings,” in principle, encom-
passes a broad spectrum of materials, especially materials such as
transition metal nitrides and carbides, and diamond or amorphous
diamond-like C:H materials, have achieved the largest attention.
This is, of course, due to their extremely high hardness, and high
chemical inertness. The high hardness in combination with a low
coefficient of friction decreases the rate of abrasive wear of the
coated component, while the chemical inertness decreases the chemi-
cal interaction between, for example, a coated tool and the work
piece. Recently, so called superlattice coatings, or multilayered
coatings with layer periodicity in the range of 5–50 nm, have at-
tracted interest due to their extremely high hardness values. Despite
the increasing commercial importance of hard coatings, and the large
volume of scientific literature available, there is still a relatively poor
understanding of the relation between film deposition parameters,
and the microchemistry and microstructure of the films. This is
partly due to the complexity of the deposition process, as mentioned
above, but also to the fact that the coatings are, in many cases, poorly
or not adequately characterized.
Quantitative chemical and structural analysis of thin films,
particularly of compounds, are often quite difficult and troublesome
to carry out. For example, many techniques used for “bulk” materials
cannot be applied to chemical thin film analysis due to the limited
amount of material available in a film only a few µm thick. Also,

microstructural analysis of thin films can be problematic. The small
grain sizes, and large defect contents complicate phase identifica-
tion, accurate determination of lattice parameters, and quantitative
determination of the defect type and character. Furthermore, sample
preparation for transmission electron microscopy (one of the most
important tools for microstructural investigations), especially for
cross-sectional imaging, is itself a nontrivial problem due to the large
differences between film and substrate etch rates.
The aim of this chapter is to give some typical examples of hard
coating microstructures, and to discuss some of the variations in
microstructures that occur as a function of process parameters. Par-
ticular emphasis will be given to nitride films deposited by ion
assisted deposition techniques due to the high commercial success of
such coatings and the large body of literature available. It should,
however, be mentioned that the chapter does not give a complete
review of all the literature in the field, but rather gives some typical
examples of what has been published. In order to more clearly
understand the requirements that are placed on hard coating materi-
als, the chapter begins with a brief discussion about the fundamental
properties of hard materials in general. The third and fourth sections
of the chapter discuss the correlations between process parameters,
microstructure, and stresses and hardness of the coatings. Finally,
the last part gives typical examples of the structure of the coating/
substrate interface region, and how its constitution relates to the
coating adhesion.
2.0 HARD MATERIALS
The hardness of a material is determined by its intrinsic hard-

ness, (the hardness characteristic of the material in the single-crystal
form) and by microstructural features that affect the deformation
mechanisms. Thus, the understanding of a hard material requires
detailed knowledge both about the electronic and atomic structure,
and about the microstructure of the sample of interest. It is also
essential to have a hardness measurement technique that provides

reproducible results in order to at least qualitatively compare differ-
ent samples. In this section we will briefly discuss some parameters
that influences the intrinsic hardness, and some general features that
affect the measured hardness values. Table 1 shows structural, me-
chanical, and thermal properties of some hard coating compounds.
Table 1. Structure, Mechanical, and Thermal Properties of Some

Hard Coating Materials (from Refs. 1–8).
Thermal Melting or
Hardness Young’s expansion decomposition
[kg mm-2 ] modulus Poissons coefficient temperature
Material Bulk Film [kN mm-2] ratio [10-6 K-1] [°C]
TiN 2000 ≤3680 440 0.25 9.35 2949

TiC 2900 ≤4000 450 0.19 7.4 3067
TiB 2 3370 480 8.0
VN 1500 460 8.1 2177
VC 2900 430 2648
ZrM 1500 ≤4500 510 0.25 7.2 2982
NbN 1400 480 10.1 2204
NbC 2400 580 0.22 6.6 3600
TaN 1000 3.6 3093
CrN 1100 400 2.3 1500
Cr3C2 1300 400 10.3 1810
HfN 1600 ≤4700 6.9 3387
HfC 2700 ≤3900 460 0.18 6.6 3928
WC 2100 695 0.185 4.3 2776
Al2O3 ≤2100 400 0.23 9.0 2300
SiC 2600 ≤4000 480 5.3
BN 3000 –5000 660
B4C 3000 –4000 440 4.5–5.6
C, diam. 9000 440–590 0.8
2.1 Intrinsic Hardness
The intrinsic hardness of a material is determined both by the

strength of the interatomic forces and the crystal structure. Materials
exhibiting a high intrinsic hardness can be characterized as having

high cohesive energy, short bond length, and a high degree of
covalent bonding.[2][9][10] The strength of the interatomic forces, and
the bonding length plays an important role in determining the elastic
properties of a material, and thus, also the hardness. For example, a high
elastic modulus implies that the atomic potential wells have a large
curvature at the atom position corresponding to the minimum (the
equilibrium atom position) which in turn implies both a deep well and
a small bond length. An intrinsically hard material also requires that the
resistance to dislocation generation and propagation is high, and that
the number of operative slip systems is low.
A high resistance to dislocation propagation and multiplication
can, for example, be achieved in materials with a high bond strength,
and thus, a high Peierls stress (intrinsic lattice resistance to disloca-
tion motion). Also, highly directional bonds such as covalent bonds
will restrict dislocation propagation. Covalent bonds are character-
ized by comparable orbital energies for neighboring atoms, a strong
overlap between orbitals and similar orbital sizes (for example, see
Ref. 11). Covalent bonds are also highly directional. Diamond, the
hardest material known, is a completely covalent bonded material.
However, many compounds do not have pure covalent bonds, but
often contain bonds of mixed character. In general, the hardness
decreases as the proportion of covalent bonding decreases. This can be
demonstrated with the three compounds TiC, TiN, and TiO, which
have the same crystal structure (B1-NaCl), and similar lattice param-
eters. For these compounds the degree of covalent bonding de-
creases, and the amount of ionic and metallic bonds increases when
the nonmetallic element is changed from C to N to O. The hardness
for these three compounds are approximately 3000 kg mm-2, 2000 kg
mm-2, and 1000 kg mm-2, respectively (see Ref. 3 for a more detailed
discussion). Notably, it was predicted that the covalent carbon-
nitrogen solid, β-C3N4 (β-Si3N4 structure) would have a bulk modu-
lus comparable to or greater than diamond.[10]
In spite of the complicated dependence of the hardness on
material parameters such as type of bonds, bond strength, slip sys-
tems, etc., many attempts to relate hardness to the atomic structure
through a single parameter have been done. For example, the hard-
ness has been related to the cohesive energy,[12] heat of formation, [13]
surface energy,[14] and vibrational entropy.[15] However, such simple
approaches can only give rough predictions. As already mentioned, a
hard material requires both a high cohesive energy (deep potential
well), and a short bond length, the latter implying a large curvature in
the bottom of the well. Thus, if the cohesive energy is divided by a
parameter related to the bond length, such as the molar volume, a
relatively good correlation with measured hardness values can be
obtained.[2] This is demonstrated in Fig. 1 for several materials.
Figure 1. Relationship between hardness and cohesive energy per molar volume.
(Reprinted with permission Almond, E., Vacuum, 34:835, 1984)
Since hardness values are measured by indentation methods,

knowledge of the plastic deformation processes is essential in order
to fully understand the obtained results. The processes occurring
during a hardness indentation not only involve large localized stresses
beneath the indenter, but also a very complex stress distribution[16]
making the obtained results difficult to interpret in terms of physical
models. This also implies that the obtained results strongly depend
on the hardness indentation method. However, in spite of the com-
plex nature of a hardness indentation measurement, the plastic defor-
mation occurring beneath the hardness indenter predominantly oc-
curs via dislocation slip even though processes such as structural
densification, and micro-crack propagation also might contribute.
Thus, defects that hinder dislocation propagation will effectively
increase the hardness of the material. Some typical microstructural
features that influence the hardness values are grain boundaries,
precipitates, dislocations, impurity atoms, and vacancies. Thin films
often have completely different microstructures compared to bulk
materials, e.g., a much smaller grain size, and a higher concentration
of other defects. Thus, it is understandable that thin films often show
much higher hardness values compared to bulk materials.
The most generic and striking feature observed in thin film
microstructures is the small grain size d (usually < 1 µm). In fact d is
often less than 100 nm and in some cases as small as 5–10 nm if
the films are deposited at low temperatures with respect to their
melting temperature Tm. In metal and alloy films the small grain size
cause a hardening of the material in agreement with the Hall-Petch
relation (for example, see Ref. 17).
Eq. (1) H = Ho + kd-1/2
where H is the film hardness, Ho the intrinsic hardness, and k a

material dependent constant. This relation which originally was
developed for yield strength is based on dislocation pileups at grain
boundaries. For hardness, the Hall-Petch relation has been found to be
valid for d values as low as 20 nm for single-phase metal films.[18]–[20]
Also, Eq. (1) has been used to fit hardness increases in Cu/Ni
multilayered thin films if d is taken as half the bilayer repeat length
(characteristic “grain size”).[21] However, for refractory compound
films several reports have shown that the hardness not always de-
pend on the grain size as will be discussed further in Sec. 4 of this
chapter.
3.0 MICROSTRUCTURE AND MORPHOLOGY OF

HARD COATINGS
Due to the non-equilibrium nature of most vapor-deposition

methods, the resulting film microstructure and morphology are gen-
erally highly dependent on the deposition technique and the used
growth parameters. Despite the importance of the physical structure
on the hard coating properties, there is presently only a qualitative
understanding of the origin of the various microstructures observed.
This section begins with an overview of the most important analyti-
cal tools for structural determination. We then review the current
understanding of the mechanisms by which the various structures
develop, and the classification into structure zone models. Computer
simulations, presently being employed to model film growth, will
also briefly be described. The possibilities of tailoring film micro-
structure is discussed and exemplified with studies of TiN, diamond,
and multilayered coatings. Finally, the large influence of the substrate
constitution on film nucleation and resulting structure is discussed.
3.1 Characterization Methods
In this section we give a brief overview of some of the most

useful methods for structural characterization. A review of some
methods used for characterization of hard coatings has been given
by Sundgren, et al.[22]
Electron optical methods in general are the most important
tools for microscopical studies, because of their superior resolving
power to optical microscopy. Scanning electron microscopy (SEM)
is used successfully to study the surface morphology of coatings, as
well as the grain structure on fractured surfaces. However, SEM-
cross sections are very sensitive to the fracturing conditions making
clear and representative areas difficult to find. Also, since hard
coatings often have small grains, and a high defect concentration,
transmission electron microscopy (TEM) is the best tool available to
achieve a detailed knowledge of the microstructure. By combining
direct imaging, and electron diffraction it is possible to obtain infor-

mation about phase composition, grain size and shape, degree of
preferred orientation, and density of defects such as dislocations,
cracks, voids, and gas bubbles. Voids and gas bubbles are easily
viewed in either an under- or overfocused condition through which
the phase contrast is enhanced. However, care should be taken when
correlating the amount of defects observed with growth parameters,
because similar artifacts can also be introduced during the sample
preparation. TEM also has the limitations of small sample thickness
(typically < 150 nm), and that three-dimensional features are pro-
jected into two-dimensional images. Both TEM and SEM can also be
used to study the nucleation stage by depositing very thin (< 100 nm)
films. However, such studies are scarce in the field of hard coat-
ings,[23] in spite of their importance for understanding the effects of
the initial nucleation on the subsequent film growth. Cross-sectional
TEM (XTEM) nucleation studies of the nucleation stage in as-
deposited thicker films is complicated by superposition of contrast
from secondary nucleated grains, and subsequently deposited material.
The use of electron diffraction analysis as a complement to x-
ray diffraction (XRD) also has several advantages. Because of the
much shorter wavelength of fast electrons compared to x-rays, the
line-broadening of electron diffraction patterns are considerably
reduced with respect to that obtained by XRD, which facilitates the
detection of crystallites much smaller than 10 nm. However, the
accuracy in the determination of the interplanar spacings is generally
not as good as in XRD due to smaller diffraction angles. On the other
hand, the lateral area over which the diffraction analysis is carried
out can be several orders of magnitude smaller in TEM compared to
in XRD. In principle it is possible to probe individual grains or
precipitates that have sizes of the order of the smallest electron beam
available (~2 nm).
TEM investigations are most commonly performed by viewing
the structure in the plane of the coating. However, since the micro-
structure often changes with distance from the substrate surface,
cross-sectional studies give a more complete picture of the structure
evolution. However, great care should be taken when estimating
renucleation rates from XTEM images since not all columns are in
the plane of the section observed. Instead, many columns may be
intertwined, and go in and out of the section studied giving rise to an
apparent increased renucleation rate. Through XTEM it is also pos-
sible to investigate the coating-substrate interface region,[24][25] and
thus, to find correlations between the observed structure, and proper-
ties like adhesion. However, due to difficulties in the sample prepa-
ration, XTEM is not near as straight forward as plan-view studies.
Helmersson and Sundgren[26] have developed an ion milling prepa-
ration technique for cross-sectional samples that include also coat-
ing-substrate combinations with large differences in etching rates.
The technique, in which the ion beam is allowed to etch the sample
only in the substrate-to-coating direction, has proved successful in
the preparation of XTEM samples of TiN[25][27] and TiAlN[28] coat-
ings on steel with coating thicknesses up to 4 µm.
XRD is perhaps the most commonly used tool for structural
analysis of as-deposited coatings. Important to note is that XRD
gives integrated information from the full coating thickness, and also
often from the substrate for coatings < 5–10 µm thick. However, as
already been pointed out, the observed diffraction peaks are often
quite broad since the films usually have small grain sizes, and high
defect concentrations. However, the accuracy in XRD analysis is
normally sufficient for general phase identification, texture (pre-
ferred grain orientation), and lattice parameter determination. The
peak broadening itself can be employed to deduct grain size and
defects that are causing non-uniformly distributed strains[29] often
observed in PVD deposited coatings.
In order to separate the size and strain contributions to the peak
broadening, a Fourier analysis of the peak shapes have to be done,
and multiple reflections have to be used (e.g., the Warren-Averbach
method).[30] Obtaining reasonable intensities for high-order reflec-
tions is, however, difficult in coatings with small grain sizes. If the
broadening of only first order peaks are used to calculate the grain
size (Scherrer’s equation)[30] too low values may be obtained since
the defect contribution has not been separated.
Finally, it is emphasized that in order to achieve as much

knowledge as possible of the coating microstructure, TEM, XRD,
SEM, and other analyzing techniques should be used in combination.
3.2 Influence of Deposition Parameters on Microstructure
Structure and Morphology Models. Extensive studies of the

correlation between structure and deposition parameters have been
performed over the past 20 years. This has lead to the development,
and refinement of structure zone models (SZM), which predict the
morphology, and microstructure of a film, independently of its
material, as a function of the adatom mobility. The recent under-
standing of film microstructure has been reviewed by Thornton[31],
for example. Movchan, and Demchishin[32] were the first to observe
that the morphological structure of evaporated coatings of Ti, Ni, W,
ZrO2, and Al2O3 could be related to a normalized, or reduced tem-
perature Ts/Tm (where Ts is the actual film temperature during depo-
sition, and Tm is its melting temperature, both in K). Figure 2a shows
the three different structure zones that were defined.
(a)
Figure 2. Structure zone models for film growth. (a) Model proposed by Movchan
and Demchishin, and (b) Model proposed by Thornton for sputtered metal
coatings. Ts is the substrate temperature and T m is the coating material melting
point. (Reproduced, with permission, from the Annual Review of Materials
Science, Vol. 7, ©1977 by Annual Reviews Inc.)
(b)
Figure 2. (Cont’d.)
In Zone 1 (Ts/Tm < 0.3) a fine-grained structure consisting of

textured and fibrous grains with domed tops and a high density
of lattice imperfections and porous grain boundaries is formed. The
structure is produced by the process of continued nucleation of
grains during deposition due to the low adatom mobility (incident
atoms stick at the point of impact) in combination with shadowing
effects.[31][33] Zone 1 grains tend to point in the direction of the
coating flux vector. The Zone 2 structure (0.3 ≤ Ts/Tm < 0.5) consists
of uniform columnar grains defined by metallurgical grain bound-
aries. Grain sizes increase with Ts/Tm, and the grains may extend
through the film thickness at high Ts/Tm. The structure is believed to
be due to the onset of surface diffusion. Also, Zone 2 grains tend to
have highly faceted top surfaces giving rise to a smooth, but mat
surface appearance. In Zone 3 (Ts/Tm ≤ 0.5) the columnar grains have
increased in width through bulk diffusion or by recrystallization.
These grains are sometimes misleadingly[34] referred to as equiaxed.
Thornton[35][36] extended this zone classification to include the
bombardment-induced mobility process in sputtering through the
use of the sputtering gas pressure parameter (see Fig. 2b). In this
model a Zone T was added as a transition between Zones 1 and 2,
consisting of a dense array of poorly defined fibrous grains without
voided boundaries. Zone T films are denser and have a much smoother
surface morphology than the two surrounding zones. The reduction

in film morphology size and resulting film densification with de-
creasing gas pressure has been related to ion-bombardment phenom-
ena from elastically backscattered sputtering gas atoms.[37] For high
pressures, the collisional scattering by the inert gas atoms enhances
the oblique component in the deposition flux which results in an
increased Ts/Tm needed to obtain a Zone T or Zone 2 structure.
Messier, et al.,[38] found that the Zone 1–T boundary in, for
example, TiB2, BN, and SiC is non linear and varies as the applied
substrate bias potential varies as a function of gas pressure. Figure 3
shows the SZM by Messier, et al.,[38] including the effects of both
thermal, and bombardment-induced mobility. As the bombardment
energy increases, the width of the Zone T increases on expense of
Zone 1. It was also found that inside the columnar structure of Zones
1 and T, the intrinsic structure can be either polycrystalline or
amorphous.
Figure 3. Structure zone model for film growth showing the effect of both
bombardment and thermal-induced mobility. (Reprinted with permission, Messier,
R., Giri, A. P., and Roy, R. A., J. Vac. Sci. Technol., A2(2), pp. 500–503, 1984.)
The present SZM are derived mainly from plan-view and

cross-sectional SEM investigations. However, XTEM observations
should indicate if there is additional fine-structure in Zones 1 and T
than previously observed.
For purposes of further generalizing the SZM, a bombardment-

induced mobility parameter have been suggested which represents
the average energy deposited per incoming vapor atom, <Ed>. Such
a parameter would be obtained by integrating the full spectra of
energetic particles impinging, and by normalizing the obtained en-
ergy to the growth rate,[39][40] which requires detailed knowledge of
both the flux and energy of the bombarding species which are not
easy to record. However, it was recently demonstrated in the case of
Ti0.5 Al0.5N films that the ion-induced changes in film microstruc-
ture, texture, phase composition, and nitrogen-to-metal ratio follow
distinctly different mechanistic pathways depending upon whether
the ion energy or flux was varied, resulting in different properties for
the same value of <Ed>.[41] Thus, <Ed> is clearly not a universal
parameter.
The physical structure of a thin film can be related to the
mobility of the adatoms during growth. The different sources of
adatom energy are: (i) thermal, (ii) energetic particle bombardment,
the flux and energy distribution of which are pressure dependent, and
(iii) chemically induced mobility (compound growth). The third
contribution comes from the nature of the vapor species; their ad-
sorption-desorption characteristics, and thus the adatom bond
strengths. This factor has this far not been considered in terms of
SZM.
It has been realized that especially the low-mobility structures,
(Zones 1 and T) undergo an evolutionary growth development, i.e.,
steady state is never reached. The Zone T structures are generally
considered to be the internal structures of the Zone 1 structure,[36][38]
and that the Zone 1 structure forms because of preferential nucle-
ation at inevitable substrate inhomogeneities. Messier, et al.,[38][42]
proposed an extended SZM that includes also the evolution of
morphology with thickness. With both experiments and models, it is
shown how a self-organized structure results through random aggre-
gation of atoms and competition for cone-shaped cluster growth.[42][43]
Figure 4 is a simplified representation of the competition for cone
growth and is a fractal construction in the sense that it shows scale
invariance.[42] The cone-shaped morphology observed in SiC,[44] for

example has been taken as evidence that there is a growth competi-
tion between adjacent grains, and that this general process occurs at
all levels of morphology development from initial aggregate clusters
(1–3 nm) all the way to surface structural features with centimeter-
size-dimensions for high enough film thicknesses. For conditions of
low Ts/Tm and particle bombardment, the dominant morphology size
in sputtered SiC films has been reported to evolve nearly linear with
the 3/4 power of the film thickness.[44] The exponent decreased with
increasing Ts/Tm, and decreasing discharge pressure. Messier, et
al.,[38] also have constructed an evolutionary SZM in which the
different structural sizes grow larger with increasing thickness.
Figure 4. Schematic drawing of the geometry of thin film morphology. (Re-

printed with permission, Messier, R., J. Vac. Sci. Technol., A4(3), pp. 490–495
1986.)
However, if a power law exists for microstructure evolution,

XTEM investigations of TiN[28][45] shows that the exponent can be
close to zero, indicating little evolution. Figure 5a is an XTEM
micrograph of a TiN film sputter-deposited at Ts/Tm = 0.17, P = 1.1
Pa, and Vs = 0 V in which the column width is nearly constant with
film thickness after an initial strong competition for growth in the
first 0.5 µm next to the substrate. [45] A similar microstructure
evolution is observed in Ti 0.5Al0.5N films deposited at Ts = 400°C
(see Fig. 5b).[28]
Figure 5. Cross-sectional TEM micrographs showing the evolution of micro-

structure in (a) a TiN film sputter-deposited at Ts = 300°C and P = 1.1 Pa and V s
= 0 V and (b) a Ti 0.5N film deposited at T s = 400°C. Both films have columnar
microstructures with a rapid increase in column diameter in the region closet to
the substrate. (Reprinted with permission, Hakansson, G., et al., Thin Solid
Films, 153:55 1987.)
Computer Simulation of Film Growth. Monte Carlo

(MC),[33][46]–[50] and Molecular Dynamics (MD) [51] computer simu-
lations have been used to graphically demonstrate how porous mi-
crostructure forms because of shadowing on the atomic size scale
due to low adatom mobilities. Figure 6 is a two-dimensional (2-D)
growth model, using the MC technique, of Ni that illustrates a
sudden change from a porous to a densely packed film when Ts/Tm is
increased from 0.20 (350 K), to 0.26 (450 K). The three bottom
layers represent the substrate. At Ts/Tm = 0.26 the migration rate of
adatoms to shadowed regions obviously was large enough to surpass
the rate of void incorporation during growth. MC and MD simula-
tions by Müller[46][51][52] of films grown under ion irradiation also
show an increase in film density towards bulk values. Ion incorpora-
tion, sputtering, and recoil implantation results in a film density that
increases almost linearly with the ion-to-vapor flux ratio. The calcu-
lations also predict that there should be an optimum ion energy E*i
for densification. The optimization results from the fact that at Ei <
E*i, the number of recoil events is small, while at Ei > E*i an
increasing fraction of the ion energy is lost deeper in the lattice
leaving vacancies which cannot be filled by arriving vapor species.
2-D MD calculations have been used to describe Ar+- assisted Ni
growth at Ts = 0 K.[51] Figure 7 shows a collision sequence induced
by a 100-eV Ar+ ion impinging on a porous Ni film. The incident ion
transfers kinetic energy to other atoms as the collision sequence
develops in time. It is thus demonstrated how low-energy ion irradia-
tion during growth, even in the absence of thermal diffusion, can
result in film densification.
Srolovitz[53] also used MC simulations to model the evolution
of microstructure in the low-mobility case, and observed growth
competition between adjacent grains resulting in a columnar-like
microstructure. The mean grain size increased during growth with
time as t0.4, independent of temperature.
Finally, the model by Lichter, et al.,[54] also predicts that
columnar microstructure should result due to low adatom mobilities,
and that the column width and mean square diffusion length of
adatoms are related.
Figure 6. Monte Carlo simulations of the microstructure of a two-dimensional Ni

film during deposition at different times t for substrate temperatures T s = 350,
420, and 450 K. The deposition rate R is 1 nm s -1 for a packing density of 0.87.
The angle of vapor inclination is a = 45°. (Reprinted with permission, Muller, K.
H., J. Appl. Phys., 58:2573, 1985.)
Figure 7. Molecular dynamic simulation of the structure of a two-dimensional

Ni film deposited at Ts = 0 K at various times t after bombardment by a 100 -eV
Ar+ ion. Atomic displacements (not trajectories) are indicated by straight-line
segments with origins at the zero-time positions of the relocated atoms. (Re-
printed with permission, Muller, K. H., Phys. Rev. B, 35:7906, 1987.)
Effects on Film Microstructure and Preferred Orientation

by Energetic Bombardment. It is well established that low-energy
ion irradiation during film growth from the vapor phase has a strong
influence on microstructure.[55] Energetic neutral species are typi-
cally present in PVD processes, and their effects on film growth are
often inseparable from that of the ions. XTEM studies of, for ex-
ample, reactively sputter-deposited TiN films,[27] and TiAlN films[28]
have shown that increasing ion energy leads to (i) densification of
porous column boundaries, (ii) defect incorporation, and (iii) en-
hanced renucleation rate from ion bombardment-induced surface
defects leading to a suppression of the columnar-type growth. The
rate-determining ion energy values for these stages are functions of
the ion species, ion current density, the pressure, and the deposition
temperature.[27][39][56] These ion bombardment-induced effects will
be exemplified below from XTEM studies of reactively sputter-
deposited TiN with predominantly Ar+ ion bombardment. Figure 8
shows the defect incorporation (dislocation networks and point de-
fect aggregates) as a function of substrate bias Vs in a multilayered
film bombarded by Ar ions accelerated by Vs = 120, 80, and 40 V,
and Ar ion-to-Ti arrival rate ratio Jion/JTi ~0.3. The figure shows a
decreasing defect density with decreasing Vs from 120 to 80 V. As
the ion energy is increased, the projected range of the ions becomes
larger giving rise to a defect generation at increasingly larger dis-
tances below the growth surface, thus reducing the probability of the
point defects becoming annihilated. Figure 9 shows a closing of
intercolumnar porosity when increasing Vs from 80 V to 120 V at Ts
= 300°C, as a result of forward sputtering and ion irradiation-
enhanced adatom mobilities (Jion/J Ti ≈ 0.3). Figure 10 shows the
disruption of columnar growth for Vs > 200–240 V with Jion/JTi = 0.4.
The renucleation is caused by surface defects created by the imping-
ing ions.
Figure 8. Cross-sectional TEM micrograph showing the evolution of defect

incorporation in a TiN film sputter-deposited at T s = 700°C. The negative
substrate bias V s was changed stepwise without interrupting the growth from 120
to 80 and to 40 V.
Figure 9. Cross-sectional TEM micrograph showing the evolution of microstruc-

ture in a TiN film sputter-deposited at T s = 300°C. The negative substrate bias Vs
was changed stepwise without interrupting the growth. The voided regions along
column boundaries (indicated by arrows) become dense when increasing Vs from
80 to 120 V.
Figure 10. Cross-sectional TEM micrograph showing the evolution of the

microstructure of a TiN film sputter-deposited at Ts = 900°C. The negative
substrate bias Vs was changed stepwise without interrupting the growth. The
columnar growth was suppressed for Vs above 200–240 V (Ar ion-to-Ti atom
arrival rate ratio Jion/JTi = 0.4).
The effects of increasing the ion flux at constant ion energy

during unbalanced magnetron sputtering of TiN films in mixed
Ar:N2 (4:1) discharges was investigated by Hultman, et al., [56]
and are shown in Fig. 11 a–d. Increasing the value of Jion/J Ti from 0
to 7 at Vs ≈ 100 V, Ts = 350°C, and P = 5.0 Pa, resulted in: (i) film
densification through forward sputtering, and enhanced surface mo-
bility of adatoms (Fig. 11a–c), (ii) increasing renucleation through
ion-bombardment-generated surface defects (Fig.11c) leading to
smaller grain sizes, and (iii) grain enlargement because of enhanced
adatom mobilities dominating over increased renucleation rates (Fig.
11d). Figure 12 shows a similar grain enlargement at the substrate
for an ion-plated TiN film deposited on stainless steel substrates
at Ts = 450°C, Vs = 50 V, and Jion/J Ti ≈ 2.5.[25]
Another effect of ion bombardment during deposition can

be to alter the film texture.[55]–[59] For example, the preferred
crystallographic orientation of the films in Fig. 11 changed from 111
at Jion/JTi = 0.3 to 200 at Jion/JTi = 2.1, and back to 111 for Jion/J Ti ≥
4.5.[56] A systematic study on the development of preferred orienta-
tion in sputter-deposited Ti1-xAlxN films as a function of both ion
energy, and ion flux was recently presented by Adibi, et al.[59] The
physical explanation for the texture formation is complex and
involves effects from the anisotropy of incident ion penetration and
sputtering in different crystallographic directions. In the case of
reactive sputtering of compound materials, also a variation in surface
chemistry, and thus surface energy as a function of deposition
parameters may affect the type of preferred orientation. As a result of
this interplay, only certain favorable oriented grains will survive and
grow. In a practical sense, texture effects should be considered
since most hard coating materials have directional dependent
properties like hardness.
(a)
Figure 11. Cross-sectional TEM micrographs and corresponding electron

diffraction pattern showing the effect of ion flux on the microstructure
evolution of sputter-deposited TiN films grown at Ts = 350°C, Vs = 100 V, and
P = 5 Pa. (a) 0.3 ions per Ti, (b) 2.1 ions per Ti, (c) 4.5 ions per Ti, and (d) 7.1
ions per Ti.
(b)
(c)

(d)
Figure 12. Cross-sectional TEM micrograph showing the microstructure evolu-

tion of an ion-plated TiN film grown on stainless steel at Ts = 450°C, Vs = 50 V,
and ion-to-Ti arrival rate ratio Jion/JTi ≈ 2.5.
Morphology of Diamond Coatings. Diamond, and so called

diamond-like carbon coatings are today produced by various pyrolytic
and plasma-activated chemical vapor deposition, (CVD) as well as
PVD methods from hydrocarbon-hydrogen mixtures.[19][60]–[63] Al-
though different CVD techniques produce very similar crystalline
diamond coating morphologies, the mechanism of diamond growth
is controversial. However, it is established that for certain mixtures
of hydrocarbon gas and hydrogen, for example, CH4 diluted to 1%
with H2, there are conditions of pressure and temperature where
growth of the metastable diamond phase is significant, and while at
the same time growth of the graphite phase is suppressed. A neces-
sary gas activation to generate atomic hydrogen and chemically
active fragmentary hydrocarbons is obtained through the use of, for
example, a plasma created by microwave, or radio frequency sources,
or by a hot filament. Two fundamentally different hypothesis have
emerged in the effort to explain the stabilization of crystalline dia-
mond, and not graphite or vitreous carbon.[63] According to the first
proposition, graphite is the thermodynamically stable phase and the
role of the atomic hydrogen is to etch any graphite that may form and
hence to make diamond kinetically stable with respect to graphite.
Alternatively, it is argued that at the growth interface, diamond
surfaces are stabilized by termination of hydrogen to make crystal-
line diamond thermodynamically stable. The growth rate is strongly
temperature dependent with a maximum at about 900–1000°C, where
the surface mobility of absorbed species is high.
Because of growth competition, the coating may consist of
many phases with variable amounts of sp1, sp2, and sp3 bonding,
different degrees of crystallinity and various fractions of incorpo-
rated hydrogen. Diamond is composed of sp3 bonds exclusively,
whereas diamond-like carbon is composed of a mixture of sp2 and
sp3 bonds. The degree of sp3 coordination can be determined using a
variety of methods, including photoelectron spectroscopy, electron-
energy loss spectroscopy, infrared spectroscopy, Raman spectros-
copy, cathodoluminescence, or nuclear magnetic resonance spec-
troscopy. Useful working definitions of what is meant by “diamond”
and “diamond-like” coatings are given in Refs. 19 and 64. In the case
of diamond, the definition demands:[64]
1) A crystalline morphology of the deposit discernibly

by electron microscopy.
2) A single-phase crystalline structure identified
by diffraction technique such as x-ray or electron
diffraction.
3) A Raman spectrum typical for diamond, i.e., a
single narrow line at ≈1332 cm-1.
Although methane is the most commonly investigated source
gas for the deposition of diamond, other hydrocarbon species mixed
with hydrogen have been shown to be effective in CVD. Bachmann
constructed a ternary phase diagram for carbon-hydrogen-oxygen.[65]
The nucleation of diamond is quite complex as briefly de-
scribed further in Sec. 3.3. Initially, diamond particles are formed on the
substrate, and increase their sizes almost uniformly until the sub-
strate is entirely covered. However, renucleation and twinning is
commonly observed.[64][66] A characteristic of the polycrystalline
diamond coatings grown by CVD is that they have well-defined crystal
habits which strongly depend on deposition conditions. The diamond
coatings are commonly bound by octahedral {111} or cubic {100}
planes or combinations of the two and with the <110> crystallographic
direction normal to the substrate (the growth direction).[19][60][66]
Diamond-like carbon films are generally nanocrystalline-to-
amorphous, and can contain considerable amounts of hydrogen (usually
≤50 at%). This makes them significantly smoother than diamond
coatings with hardness exceeding 2000 kg mm-2. The high hardness,
and the possibility to deposit at temperatures as low as room temperature
make diamond-like carbons useful for protective coatings. However,
their thermal conductivities are lower than for pure diamond coat-
ings, and they decompose to graphite at temperatures above ~400°C.
A necessary condition for growth of diamond-like films with
high hardness, and a large fraction of sp3 bonds, is that the films are
subjected to an intense ion bombardment during growth. Different
preparation techniques for diamond-like carbons are reviewed by
Catherine[67] and Bachmann, et al.[68] Generally, film deposition
processes are divided into ion beam, reactive plasma, and laser
techniques. The characteristic feature of most processes for growing

dense carbon films is that energetic species, usually in the range
50–500 eV hit the surface of the growing film. In CVD, the most
commonly employed method for deposition of diamond-like amor-
phous hydrocarbon (a-C:H) is by impact of hydrocarbon ions in self-
bias plasma-assisted CVD.[69] Also, ion-assisted PVD methods have
been successful in producing diamond-like carbon (a-C) films
containing little hydrogen at relatively low temperatures (a few
hundred °C).
The use of wear-protective a-C:H films is today focused on
optics, read and write heads for magnetic recording media, and other
low-load technologies. The reason for this is the severe problems
with film adhesion that occur when the coated component surfaces
are exposed to high loads. The adhesion problem of a-C:H films can
be related to high compressive stress levels in the films,[69] low
interaction between the film and substrate, and microstructural de-
fects in the film/substrate interface. Especially for low-temperature
depositions, defects, and weak boundary layers at the interface can
occur.[70] This is mainly due to the difficulties of keeping the sub-
strate surface free of contamination in the initial stages of film
growth at low temperatures. There is a great need for characteriza-
tion of microstructure and phase composition of the interfaces be-
tween a-C:H films and steel substrates. Two types of a-C:H films,
representative of different commercially available low-temperature
ion-assisted deposition techniques for a-C:H, were recently investi-
gated: high-energy nitrogen ion-beam decomposition of large hydro-
carbon molecules and magnetron plasma decomposition of C2H2 in
mixed Ar-C2H2 discharges using W or Mo targets for metal
codeposition.[70] Characterization by high-resolution TEM and AES
showed that substrate surface oxides and contaminants could not be
removed during sputter-etching in high-vacuum background pres-
sure condition, but that interfacial layers of reactive metals Mo or W
between steel and a-C:H could significantly reduce residual con-
taminants. Structural characterization shows that the a-C:H films
were amorphous on the atomic scale, and that nm-sized metal-rich
clusters could form in the a-C:H matrix during codeposition of metal
and hydrocarbon.
Microstructure and Stability of Metastable TiAlN Coat-

ings. TiAlN has proven to be a useful and illustrative model system
for investigating growth of metastable ternary compounds. In this
section the preparation and characterization of TiAlN will be de-
scribed in some detail. TiAlN coatings have a much improved
oxidation resistance compared to TiN. The alloys have both similar
microstructure and mechanical properties. However, electrical prop-
erties, and color vary with AlN content. TiAlN is metastable in the
NaCl-structure, and decomposes to form TiN, and Wurtzite-struc-
ture AlN at elevated temperatures. As a matter of fact, the solid
solubility of TiN and AlN is extremely limited.[71][72] Nevertheless,
metastable TiAlN alloys can be stabilized in NaCl-structure during
thin-film deposition by two mechanisms. The first is by operating
under kinetically limited conditions, i.e., low substrate temperature
Ts, and high deposition rate. In combination, low-energy ion bom-
bardment during growth may be applied to enhance adatom mobil-
ity, and provide collisional mixing of the topmost atomic layers
without significantly increasing Ts. Second, for pseudomorphic growth
(single-crystals), the metastable phase can be obtained by a combi-
nation of (i) close lattice match, (ii) high supersaturation, and (iii)
large volume misfit between pseudomorphic, and stable phases.
Ti1-xAlxN alloy coatings in the NaCl structure can be prepared
by most PVD-processes, although the stability range is process-
dependent. For example, alloys with Al-contents up to x = 0.4 was
obtained by sputter-deposition at Ts = 500°C [73], while x < 0.6–0.7
was reported for arc-evaporated coatings at Ts = 400–450°C.[74][75]
The phase-decomposition reactions have been established for
sputter-deposited TiAlN films.[73][76] Generally, phase-separation takes
place by surface nucleation of wurtzite-structure AlN during deposi-
tion. Figure 13 shows XRD patterns from polycrystalline Ti1-xAlxN
films with 0 ≤ x ≤ 1 together with the expected peak positions for
TiN(002), and AlN(0002), and (101̄1) deposited at Ts = 500°C. [73]
Films with 0 ≤ x ≤ 0.52 show only a NaCl-structure (002) peak with
a position that continuously shifts towards higher 2Θ values with
increasing x. For the film with x = 0.59, peaks indexible to the
Wurtzite structure appear. Films with x ~ 0.5 showed a very weak

diffraction intensity. Figure 14a-c show plan-view TEM micro-
graphs of films with x = 0.26, 0.40, and 0.42 with corresponding
selected-area electron diffraction (SAED) patterns. The SAED pat-
terns in Fig. 14a and b show that the films with x ≤ 0.4 were single-
phase with a NaCl structure, and a strongly (002)-preferred orienta-
tion in agreement with the XRD results (c.f. Fig. 13). The average
grain size was relatively constant between 60–70 nm for x ≤ 0.40,
and the fraction of intergranular porosity was decreased with in-
creasing x. The films had a columnar microstructure. When the AlN
content was increased above 0.40, a two-phase structure consisting
of the NaCl- and wurtzite-structures appeared as can be seen in the
SAED pattern of Fig. 14c Associated with the phase separation was
a rapid reduction in grains size to ~30 nm, and an increase in the
intergranular porosity (c.f. Figs. 14b and c).
Figure 13. X-ray diffraction patterns from Ti1-xAlx N films with 0 ≤ x ≤ 1 at T s

= 500°C. Included are the expected peak positions of NaCl-structure TiN (002)
and wurtzite-structure (0002) and (1011). (Reprinted with permission, Kobashi,
K., et al., Phys. Rev. B, 38:4067, 1988.)
Figure 14. Plan-view transmission electron micrographs and corresponding selected-are electron dif-
fraction (SAED) patterns from Ti1-xAl xN films with (a) x = 0.26, (b) x = 0.40, and (c) x = 0.42. The SAED
patterns in (a) and (b) can be indexed with the sodium chloride structure while (c) is a mixture of sodium
chloride and wurtzite structures.
In order to more accurately determine the distribution of the

precipitating AlN phase, and its structural relationship to the remain-
ing Al-deficient Ti1-xAl xN, high-resolution TEM (HREM) was em-
ployed.[73] Figure 15 shows a HREM image from a sample with x =
0.42 where a wurtzite-structure AlN grain with the [0001] zone axis
parallel to the electron beam is seen located between two NaCl-
structure (TiN-rich) grains out of which one is aligned with the
electron beam parallel to the [001] zone axis. Typically, the AlN
grains were found as small separated grains located in grain-bound-
ary areas between cubic Ti1-xAlxN grains. The small grain size
together with a high defect density can explain the loss of intensity in
XRD for films with a critical AlN content for phase separation (c.f.
Figs. 13 and 15).
Figure 15. Plan-view high-resolution transmission electron microscopy image

from a Ti0.58Al0.42N film showing a NaCl-structure grain with the (001) zone axis
and a wurtzite-structure grain with the (0001) zone axis, parallel to the electron
beam. Common crystallographic directions are indicated by arrows.
The composition of the cubic phase changed during phase

separation as inferred from variations in the lattice parameter as a
function of AlN content.[73][75][77] Figure 16 shows the interplanar
distance d 002, obtained by XRD, for the NaCl-structure phase as a
function of AlN fraction, for films with 0 ≤ x ≤ 0.52.[73] The lattice
spacing decreased linearly with increasing AlN content. The line in
the figure is drawn with a least square fit to the data points of the
single-phase alloy. A linear extrapolation to x = 1 results in d002 =
0.2065 nm for NaCl-structure AlN which is close to the 0.206 value
reported in the literature.[78] The explanation for the decrease in TiN
unit cell dimensions with Al-alloying is based on a substitutional
replacement of Ti atoms with Al.[73] Al has a smaller atomic radius
than Ti, but there is also a reduced charge transfer from metal atoms
to N which gives a reduced effective size of N in TiN. The relative
influence of the two effects on the lattice-parameter reduction are not
known. Also, the microstructural changes occurring during AlN
alloying to TiN may lead to an altered intrinsic stress that could
affect the lattice parameter. There is no evidence for Al and Ti
ordering in TiN.[73]
Figure 16. Interplanar distances d002 for NaCl-structure Ti1-xAlxN films grown at
substrate temperature T s = 500°C as a function of AlN content. The line is
determined by a least squares fit using only data from single-phase films with x ≤
0.40. (Reprinted with permission, Muller, K. H., J. Appl. Phys., 62:1796, 1987.)
The phase stability of single-crystal Ti0.5Al0.5 N films was

investigated as a function of Ts.[76] In the case of single-crystals
deposited on nearly lattice-matched MgO(001), the NaCl-structure
phase could be stabilized with AlN contents up to x = 0.5 up to Ts =
540°C.[76] The higher AlN content in this case compared to the
polycrystalline Ti1-xAlxN films described above (maximum x ~0.4)
was due to the effect of pseudomorphic growth. Phase separation
occurred at Ts ≥ 560°C in epitaxial Ti0.5Al0.5 N. However, over a
narrow temperature range (540 ≤ Ts < 560°C), surface-initiated
spinodal decomposition was observed.[76] Figure 17 shows (a) plan-
view and, (b) cross-sectional TEM micrographs with SAED patterns
of a Ti0.5Al0.5N film grown at Ts = 540°C. Although the SAED
patterns indicate a single-phase epitaxial structure with the same
lattice parameter as that obtained from films grown at lower tem-
peratures, the XTEM micrographs shows a regular laminar fine
structure with a period of ~2.3 nm. The thin platelets are rectangular
with edges primarily along [100] and [010]. The modulation is due to
atomic number, or strain contrast resulting from a coherent composi-
tional modulation of TiN-rich and AlN-rich components which
presages bulk phase separation at slightly higher growth tempera-
tures.[76] The fact that the platelets extend only in the [001] growth
direction strongly indicates that the compositional modulation is
initiated at the surface during deposition rather than in the bulk of the
film, i.e., a surface-initiated spinodal decomposition. For 560 ≤ Ts <
750°C, the thickness of the single-crystal NaCl-structure alloy layer
decreased due to bulk phase separation into cubic TiN, and wurtzite
AlN, and further film growth proceeded by parallel growth of Al-
deficient NaCl-structure Ti1-xAlxN, and wurtzite-structure AlN.
The oxidation mechanism for NaCl-structure Ti0.5Al0.5N films
has also been established. While the oxidation of TiN to rutile TiO2
is pseudolinear with spallation of oxide layer,[79][80] Ti1-xAlxN has a
parabolic oxidation rate constant that gives a passive double layer.
Oxidation of Ti0.5Al0.5N proceeds by outward diffusion of Al to form
Al-rich oxide at the topmost surface and inward diffusion of O to
form Ti-rich oxide at the interface to TiAlN.[80]
Figure 17. Bright-field (a) plan-view and (b) cross-sectional transmission elec-
tron microscopy images with selected-area electron diffraction patterns from a
Ti0.5Al0.5N alloy grown on MgO (001) at T s = 540°C. Higher magnifications of
the (002) reflection reveals satellite peaks along the (010) direction in the cross-
sectional view and along both (010) and (100) directions in the plan-view.
Hard Multilayered Coatings. The deposition of multilayered

structures represents a means of tailoring the composition, and
microstructure of a coating in order to take advantage of the special
characteristics of the different hard materials and their synergetic
behavior. The motivations for using multilayers in the field of hard
wear-protective films are predominantly (i) to facilitate a strong
adhesion between film and the substrate, (ii) to obtain wear-protec-
tive films with a low chemical reactivity and low friction, and (iii) to
increase the hardness and toughness of the coating. Even though
layered materials grown by CVD have been used as hard wear
protective coatings for a long time, for example, TiN/Al2O3/TiC
structures on cemented carbide tools, it is more recently layered
structures grown by PVD have attracted attention in this field.[81]–[83]
Multilayers implies systems with the number of layers ranging from

only a few layers up to several thousand of layers in monolayer
superlattices. The term superlattice applies to films where the con-
stituent layers are repeated with such a short periodicity so as to form
a one-dimensional lattice that diffracts x-ray or electron waves. The
constituents need not be crystalline. The accuracy in controlling the
individual layer thicknesses and the interface sharpness depends on
the deposition rates, the rate fluctuations, and the speed by which the
sources can be modulated (by use of e.g., shutters). However, also
the thermodynamics of the material system will determine the sharp-
ness of each individual layer interface. For example, in cases where
the two materials are soluble in each other, one can expect some
degree of interdiffusion whereas in cases of completely immiscible
elements, sharp interfaces or an interfacial compound can be ex-
pected. However, kinetic limitations during growth, for example, a
low growth temperature, can cause the formation of very sharp
interfaces even if the materials are completely soluble.[84] In cases
where the lattice parameters of the constituents are different, also the
strain present can affect the sharpness of the interfaces by changing
the growth from a 2-D layer-by-layer growth mode to a 3-D island
growth mode.[85]
A review of the deposition and mechanical priorities of
superlattice thin films was given recently by Barnett.[86] The hard
coating superlattice system mostly investigated is that of transition
metal nitrides. The studies include the microstructure and hardness
of single-crystal superlattice films of TiN/VN,[87] TiN/NbN,[88] and
TiN/VNbN,[84] but also polycrystalline superlattice films of TiN/VN,[89]
TiN/NbN,[90] TiC/TiB2, TiN/TiB2, TiN/TiC,[82][83] and NbN/AlN.[91]
Figure 18 shows an XTEM micrograph from a lattice-matched
TiN/(V0.6Nb0.4)N superlattice with a periodicity Λ of 8.2 nm. From
the bright-field image, it can be seen that the layers are flat, and that
the layer thicknesses do not vary measurably either along the layers
or through the film thickness. Threading dislocations originating at
the MgO substrate surface can also be observed. In addition to
showing that the film was single-crystal, the electron diffraction
pattern shows satellite peaks around the Bragg spots, which can be
used to calculate Λ. The fact that satellite spots of up to 7th order
were present is an indication that the composition amplitude was

large and the interfaces fairly abrupt. However, the high-resolution
image shows that the layer interfaces are not atomically sharp. The
high-resolution image also gives the individual layer thickness.
Finally, direct observation of misfit dislocations and variations in the
lattice spacings are possible, although these were not present in this
nearly lattice matched sample.
Figure 18. Bright-field cross-sectional transmission electron micrograph, high-

resolution (200) lattice fringe image, and selected-area electron diffraction pat-
terns of a TiN/V0.6Nb0.4N film with Λ = 8.2 nm.
TEM and XRD studies provide complementary information

about the composition modulation and can be used to differentiate
between the effects of interfacial roughness and layer interdiffusion.[85]
Lattice mismatch between the constituents of a superlattice or

growth on a rough substrate surface typically lead to undulated
layers. The importance of the lattice mismatch is illustrated by the
observation that TiN/NbN (mismatch 3.6%) have less planar inter-
faces than TiN/(V0.6Nb0.4 )N (mismatch <0.2%). This can be seen by
comparing Fig. 19 which shows an XTEM micrograph of a TiN/
NbN superlattice with Λ = 4.6 nm[88] and Fig. 18. In the case of
polycrystalline superlattice coatings the layer roughness (inter-
face width) increases with coating thickness.[91] Also, the layer
modulation will follow that of the facets forming at the growth front
and not be parallel to the substrate normal.[89][91][92] These observa-
tions imply that superlattice coatings applied to objects with a
complex geometry will not necessarily be well-defined with respect
to layer uniformity and periodicity.
Figure 19. Cross-sectional transmission electron micrograph of a TiN/NbN

(100) superlattice film with Λ = 9.4 nm. High-resolution (200) lattice fringe
image of a void situated at a low-angle column boundary (upper-left) is also shown.
3.3 Substrate Effects on Microstructure
Although the structure zone models seem to be adequate in

giving a qualitative classification of film microstructure formation,
they do not, however, include effects from the substrate material
itself on the film growth mode. The observed microstructure of hard
coatings have been found to be not only dependent on process
parameters such as temperature and energetic particle bombardment
during growth, but also critically dependent on the substrate crystal
structure, morphology, composition and/or surface contamination,
all of which has strong influences on the nucleation. For example,
preferred grain orientation in coatings have been associated to ef-
fects such as minimization in surface energies[93] and a variation in
condensation coefficients.[94] It is thus essential to choose the sub-
strate material, and its pretreatment very careful. The understanding
of how changes in the substrate constitution influences the resulting
film structure is today only rudimentary from scattered observations,
although of paramount importance when optimizing film/substrate
combinations.
The substrate surface roughness often exists simultaneously on
several size scales. In Zone 1 structures, this roughness is reproduced
in exaggerated form on the coating surfaces, because of superim-
posed shadow growth.[36] The most open boundaries are associated
with the most coarse surface irregularities. Severe surface irregulari-
ties can cause Zone 1 structures to persist into the Zone 2 and 3
regions as Ts/Tm is increased. As an example, Fig. 20 shows extreme
cone-shaped growth of porous TiN originating at a substrate surface
artefact in a sputter-deposited film at Ts/Tm = 0.25.
In the case of TiN deposited on high-speed steel substrates, a
local epitaxial growth was observed on the vanadium carbide phase
due to a close structural matching between TiN and VC.[23][95][96]
Both of these phases have a NaCl structure and the lattice mismatch
is only 1.6%. On the steel matrix, and on the (Mo,W)6C phase
epitaxial growth is inhibited by the large structural mismatch, and
thus, TiN grows in a fine-grained polycrystalline mode as expected
from the SZM described above. Figure 21 shows XTEM micro-

graphs of TiN films demonstrating the effect from different phases
within the high-speed steel substrate on the grain structure of the TiN
film. Removal of substrate surface contaminants by sputter etching
prior to deposition greatly enhanced the tendency for epitaxial growth
of TiN.
In the case of diamond growth from the vapor phase, the
knowledge about the mechanism of nucleation of new diamond
crystals on the substrate and growing film is mostly empirical. The
nucleation rate is strongly dependent on the type of substrate.
For example, the nucleation density of diamonds on carbide-
forming substrates, such as Si, Mo, and W is one to two orders of
magnitude higher than the densities on non-carbide-forming sub-
strates, for example, Cu and Au. Surface defects, for example,
scratches, grain boundaries, and dislocations increase the nucleation
rate dramatically.[60][64] Regardless of the nucleation density, dia-
mond crystals initially grow unaffected by surrounding crystallite
until they come in contact. Thus, by controlling the nucleation
density the range of crystallite-size morphology can be controlled.[64]
Figure 20. Cross-sectional TEM micrograph showing a deviating microstructure

evolution of a TiN film.
Figure 21. Cross-sectional TEM micrographs showing the effect from different
phases within the high-speed steel substrate on the microstructure evolution of
sputter-deposited TiN films. (a) Fine grained polycrystalline growth on martensi-
tic steel and (Mo,W)6 C carbides and (b) epitaxial growth on VC carbides.
The nucleation density of diamond can also be increased signifi-

cantly when the substrate is roughened by abrasive powders (diamond,
silicon carbide, cubic boron nitride, and alumina).[62][97] One reason
for this effect is that diamond, diamond-like carbon or other hydro-
carbon residues from polishing or abrading with diamond powder are
left adherent to the polished surface, and that these supply nucleation
sites for subsequent diamond growth.[98] Alternatively, it can be
argued that highly disordered surface sites or microscopic crater
edges in the polished surface create high-energy sites which can be
preferred nucleation sites for diamond.[99]
4.0 INFLUENCE OF MICROSTRUCTURE ON

COATING PROPERTIES
Since the microstructure strongly varies with the growth condi-

tions, as been demonstrated in the previous section, it is natural that
also the properties of the coatings show large variations. Unfortu-
nately, the knowledge about the relationships between properties of
hard coatings and their microstructure is only rudimentary. This is
partly due to the fact that the films in most cases are insufficiently
characterized. In this section we discuss correlations between the
microstructure, intrinsic stress, and hardness. Tribological properties
and other functional properties of the coatings are not treated. This is
both due to the extreme complexity of the relations between such
properties and the microstructure and due to the absence, in the
literature, of systematic investigations on the subject.
4.1 Stress
The stresses in hard coatings are most commonly measured

through x-ray diffraction techniques such as the sin2ψ method (for
example, see Ref. 100). A generic conclusion that can be drawn from
such studies is that films grown by CVD techniques show low stress
values, whereas films grown by PVD techniques quite often are very
highly stressed. The stresses can be either tensile or compressive, but
tensile stresses are in general most damaging.[2]
The stresses in CVD films can in most cases be attributed to
thermal stresses generated by the difference in thermal expansion
coefficients between the coating and the substrate. The thermal
stresses in the coating after cooling from the deposition temperature
are tensile when the thermal expansion coefficient of the coating
material is larger than the one of the substrate and compressive if the
reverse is true. For example, in the case of TiN grown on high-speed
steel (HSS) substrates at 1000°C a compressive stress of approxi-
mately -1 GPa is obtained, whereas for TiN grown on cemented
carbide (CC) substrates a tensile stress of 1.6 GPa is obtained
(assuming thermal expansion coefficients of 9.4 x 10-6, 12 x 10-6,
and 5 x 10-6 K-1 for TiN, HSS, and CC, respectively; a Poisson
constant of 0.3, and an elastic modulus of 250 GPa for TiN; (see
Table 1 and Ref. 3).
The intrinsic or growth stresses in thin films are caused by the
non-equilibrium microstructures formed and several models de-
scribing the origin of both tensile and compressive stresses have
been postulated (see e.g., Refs. 52,101–103). Tensile stresses can be
generated by grain boundaries, small voids, and pores in the films.
These voids are most commonly located in grain or column bound-
aries as exemplified in Figs. 6 and 9. In such microstructures, which
are characteristic of low-temperature growth or Zone 1 structures in
the SZM’s, there will be attractive atomic forces acting across the
voids and thus a net tensile stress state in the films. Since the
attractive forces decreases with distance, the tensile stress compo-
nent will be reduced as the voids become large. An example of
molecular dynamics simulations of the variations in tensile stress
that occur in metal films when changing the void structure, through
ion bombardment,[104] is shown in Fig. 22.
(a)
(b)
Figure 22. (a) Molecular dynamic calculation of the intrinsic stress s in a typical
metal film as a function of the kinetic energy of the arriving adatoms. The kinetic energy
is given in units e, the depth of the atom-atom potential well. (b) Typical evolving
microstructure for three different energies of the adatoms. The chosen energies
corresponds to positions below, at, and above the maximum in s shown in (a).
The origin of the compressive stresses commonly found for

films grown by plasma-assisted PVD techniques are clearly associ-
ated with energetic particle bombardment. It has been demonstrated
that in sputtering environments the flux of backscattered (from the

target) energetic neutrals impinging on the growing film surface will
cause compressive stresses in the films as the sputtering pressure is
below a critical value.[37][105]–[109] This critical pressure value de-
pends on the geometry in the chamber, sputtering gas, target mate-
rial, target voltage, and discharge current. Some typical examples of
the variation in stresses with changing sputtering pressure are given
in Fig. 23. The value of the critical pressure reflects the energy of the
bombarding energetic neutral particles. For high pressures, scatter-
ing of the backscattered neutral particles in the gas will reduce their
energy before reaching the substrate and open voided structures will
form as discussed in Sec. 3.2.
Figure 23. Internal stress in TiN coatings versus Ar pressure for Cr, Mo, Ta, and
Pt films deposited by dc magnetron sputtering onto glass substrates at ambient
temperature. (Reprinted with permission, Hoffman, D. W., Thornton, J. A., J. Vac.
Sci. Technol., 20(3):355, 1982.)
Ion bombardment, by applying a negative substrate bias dur-

ing, for example, sputtering, ion plating, or plasma-assisted CVD,
will also result in an increase in the compressive stress levels in the
films. The bombarding particles will be implanted in the growing
films resulting in an increased defect incorporation, provided their
energy is high enough. Both of these effects will result in distortions
of the lattice as has been exemplified in Fig. 8 for the case of TiN.
The obtained stress levels first increase monotonically with increas-
ing ion energy and/or ion flux. However, for sufficiently high ener-
gies a saturation of the stress level is usually observed. The satura-
tion stress can be very high, for example, in the case of TiN films
values as high as -10 GPa, is commonly reported.[29][110] A typical
variation in stress level with increasing substrate bias is shown in
Fig. 24 for sputtered TiN films. In this case a rapid increase occurs
for bias voltages Vs > 50 V, and for Vs > 100 V a saturation is reached.
The occurrence of a saturation level has been suggested to be due to
an onset of plastic flow in the film material.[29][111][112] TEM studies
have also revealed very fine cracks in grain boundaries and along
200 and 220 planes in TiN films with a high compressive stress state
indicative of plastic deformation and fracture initiation.[113][114]
Figure 24. Internal stress in TiN coatings versus substrate bias voltage during
sputter-deposition. The films were grown on austenitic stainless steel substrate.
However, also other phenomena might contribute to a saturation

in the stress level at sufficiently high energy and flux of the impinging
ions. Petrov, et al.,[115] have shown in the case of TiN films grown at
Ts = 300°C, that the stress level in the films increases to the point
where trapped Ar in the films precipitates first at grain boundaries,
and at a later stage in gas bubbles within the grains. This precipitation
first occurred at an Ar concentration CAr of approximately 2 at%. For
lower values of CAr the Ar was suggested to be trapped in the TiN
lattice giving rise to a deformation of the lattice, and contributing to
a compressive stress state in the films. The variation in the lattice
parameter ao, and the microstructure of these films as a function of Vs
are shown in Figs. 25 and 26, respectively. As can be seen from Fig. 25,
ao increases with Vs to a maximum value at Vs = 800 V. For Vs = floating
potential (Fig. 26a), the grain boundaries are porous due to self-shadow-
ing on the atomic size level because of low adatom mobility. For Vs =
200 V, which is below the value for maximum ao, a dense structure
with a distorted grain structure characteristic of a high defect concentra-
tion is observed (Fig. 26b). At Vs = 800 V (Fig. 26c) low density
regions occurs at the grain boundaries and at Vs = 1800 V (Fig. 26d)
low density regions within the grains due to Ar gas bubble precipitation
are also seen.
Figure 25. Lattice parameter ao of reactively sputtered TiN films (determined

from the (200) Bragg reflection in x-ray diffraction voltage V s. The ion-to-Ti
arrival rate ratio increased from 0.3 at V s =100 V to 0.6 at V s = 1800 V.
(a)
(b)
Figure 26. Plan-view TEM micrograph from TiN films grown by reactive
magnetron sputtering at three different substrate bias voltages Vs. (a) V s = floating
potential (12 V), (b) V s = 200 V, (c) V s = 800 V, and (d) V s = 1800 V. The ion-to-
Ti arrival rate ratios were 0.30, 0.35, 0.50, and 0.65 respectively. The higher-
magnification inset in (d) shows Ar gas precipitates in bright contrast.
(c)
(d)

As the growth temperature is increased, the structure of the

films will become denser and the net tensile growth stress will
decease to zero. Also, in the cases when energetic particle bombard-
ment is used during growth, the defect annealing rate will eventually
become higher than the defect generation rate as the substrate tem-
perature is increased and films free of growth stresses will result. The
temperature at which this occurs depends on several factors such as;
film material, energy of the impinging particles, contamination level,
growth rate, etc. For example, a high growth rate requires a higher
temperature before obtaining films free of growth stresses. However,
as the temperature is increased the thermal stress will increase. A
schematic representation of the total stress in the films as a function
of growth temperature is given in Fig. 27. In the schematic case
shown in this figure, both the growth stress, and the thermal stresses
are assumed to be of the same sign, however, situations might of
course also occur when these stresses are of opposite sign.
Figure 27. Schematic representation of the variation in growth, thermal, and total
film stresses as a function of the growth temperature Ts.
For a coating which exhibits a high internal stress the distribution

of the stress across the coating and the substrate is ideally given as in
Fig. 28. The stress in the coating must be balanced by a stress in the
substrate, but of opposite sign. The substrate must also bend to
balance the bending moment produced by the biaxial stress in the
coating. Thus, the stress must change sign in the substrate as shown
in the figure. However, since the thicknesses of the substrate in most
cases is much larger than the film thickness, the stresses in the
substrate are usually quite low. Nevertheless, there is a stress discon-
tinuity at the coating/substrate interface. This stress discontinuity will
cause a delamination of the coating from the substrate when the
forces caused by the stress becomes higher than the adhesion forces.
Thus, for highly stressed films, the adhesion has to be very good in
order for the films to be useful.
Figure 28. Schematic figure showing the variation in stress distribution in the
coating and the substrate.
The stress in the films will also be of importance for the

coatings to withstand externally imposed stresses when the coated
objects are used in a particular application. The coating has to remain
intact, and adherent to the substrate. For example, in applications with a
high loading intensity, the imposed strain results in an additional
stress discontinuity at the coating/substrate interface if the elastic
modules of the coating and substrate are different.[116] This is shown
schematically in Fig. 29. If this stress discontinuity is of the same
sign as the stress discontinuity discussed in the previous paragraph, a
loss of adhesion might occur as the coated parts are used.
Figure 29. Schematic representation of a typical strain-stress distribution occur-

ring when an external force acts upon the coating-substrate composite, e.g.,
during a bend test. The stress discontinuity at the coating-substrate interface is
caused by differences in the Young’s module of the coating and the substrate
material.
Also, thermal stresses during usage of the coated objects can

affect the coating integrity. For example, McIntyre, et al.,[80] have
shown that the failure during high-temperature oxidation of Ti0.5Al0.5N
coatings deposited on stainless steel substrates occurs through crack
formation in the films. The tensile stress caused by the larger thermal
expansion coefficient of the substrate compared to the film was
found to be high enough to initiate crack formation in the films when
heated to temperatures above 800°C after deposition. However, it
was also demonstrated that in this case a built-in compressive stress
level in the as-deposited films, obtained by applying a negative
substrate bias during growth, the cracking resistance of the films and
thus their oxidation protective properties were considerably increased.
4.2 Hardness
Among the most measured and studied properties of hard

coatings, is the hardness. However, it is not straight forward to
measure hardness of films only a few µm thick deposited on substrates
that usually are much softer than the coating itself. Most commonly
low loads (≤10 g) are required such that the depth of the indentation
is only a small fraction of the total film thickness.[3] Procedures have

also been developed based on various “laws of mixture” schemes to
calculate the true hardness values of hard coatings with soft sub-
strates with known properties.[117]–[119] However, it should also be
pointed out that low loads, and thus small indentations can give rise
to increased hardness values even for thick bulk materials. This
effect, which is known as the indentation size effect (ISE), has been
shown to be important, particularly for ceramic materials. Several
explanations to this effect have also been given and Bull, et al.,[120]
have demonstrated that the effect can be due to a periodic buildup, and
relief of strain fields below the indenter. Independently of the origin
of the ISE, failure to recognize its importance in measuring hardness
of thin films might be responsible for many of the extreme high
hardness values for thin films reported in the literature.
However, in spite of the difficulties in measuring hardness
values of thin films correctly, the specific microstructures often
obtained in thin films will of course affect the plastic deformation
processes in the films, and thus, the film hardness. Among the most
striking features of thin film microstructures are the small grain size, and
the anisotropic columnar structure obtained in many cases. For dense
metal films,[18][20][120] the small grain size has also been shown to give
rise to increased hardness values. In these cases, the hardness in-
creased with the grain size d, down to d-values as low as 20 nm,
according to the simple Hall-Petch relationship (see Eq. 1 in Sec. 2.1).
Even for hard coatings, Rickerby and Burnett [122] have argued that the
grain size is of a major importance in explaining the higher hardness
of thin films compared to bulk materials. However, for ceramic
materials several workers have reported increased hardness values for
increasing growth temperatures Ts, as exemplified in Fig. 30 for TiN,
TiC, and TiB2 films grown by PVD and CVD techniques.[3] As the
grain size increases with Ts other structural effects than changes in
the grain size have to be used to explain this increase. Since the
results shown in Fig. 30 come from films grown without any extensive
ion bombardment during growth, it is reasonable to assume that a
reduction in the grain or column boundary void density is responsible
for the increase in hardness with increasing Ts. A small reduction in

the hardness is also observed at the highest temperature for both the TiC
and TiB2 films, where dense films with an increasing grain size can
be expected. Thus, in order for grain boundary strengthening to be
effective, the films, first of all, need to have dense column boundaries.
Figure 30. Hardness plotted as a function of growth temperature for some

ceramic coatings ( ) TiN [123], (O) TiN [124], (∆) TiC [125], and (on) TiB2 [126].
Also, other structural features, such as defect densities, texture,

intentionally layered materials, and possible precipitates have to be
considered when determining hardness/structure relationships of
thin films. Since the coatings often have pronounced preferred growth
orientations and since many ceramic materials have hardness values
that varies with orientation, it is essential to compare the coating
hardness with the correct bulk values. This effect is particularly
important for non-cubic materials, for example, the hardness of WC

varies by a factor of two depending on the orientation of the sample.[1]
The density of defects such as dislocations, point defects,
trapped gases, etc., will as discussed in the previous section, strongly
affect the stress state of the films. Correspondingly, under certain
growth conditions, especially in cases with a high flux of energetic
ions bombarding the growing film surface, compressive stress levels
as high as -10 GPa can be obtained as already been mentioned. Since
stresses will change the average distance between atoms in the lattice
away from their equilibrium positions, the elastic response of the
material will change.[127] Due to the asymmetry of the potential wells
around each atom, compressive stresses will affect the elastic re-
sponse to a larger extent than tensile stresses.[128] Since not only the
plastic properties of a material are important for the materials hard-
ness, but also its elastic properties, it is understandable that highly
compressive stressed materials will show large hardness values.
Also, the high defect concentration in a compressively stressed
material will restrict the plastic flow, and thus, be a contributing
factor in enhancing the hardness. A restriction of the plastic flow is
also naturally obtained in dense thin films only because of substrate
constraints. In films with a large grain size, a single dislocation
moving within the film as a response to an external force must leave
a dislocation line segment at the substrate/film interface. This in turn
results in an increased yield strength, compared to in a bulk material.
Freund[129] has also shown that for single-crystal films, the creation
of dislocation segments at the film/substrate interface leads to an
increased yield strength that depends inversely on the film thickness.
Compositionally, modulated structures can have quite differ-
ent mechanical properties from homogeneous materials. [81][88]
Because of the prospects for increasing the strength of materials
through compositional modulations, the efforts in the hard coating
field in these directions have started to flourish. Holleck, et al.,[82][83]
have studied growth, and properties of polycrystalline TiC/TiB2,
TiN/TiB2, and TiC/TiN multilayers with modulation wavelengths Λ
ranging from 8 to 1000 nm. The results showed that the wear
properties, as well as the strength (crack propagation measured by

indentation) of the multilayers were better than those measured for
single-component films. In the case of transition metal nitride
superlattice films, significant hardness increases were observed for
Λ around 5 to 50 nm.[88]
Figure 31 shows the Vickers microhardness H versus wave-
length Λ for nitride superlattices with different lattice mismatch
values.[88] H increased rapidly with Λ from the alloy values at Λ = 0
and reached a maximum value at 5–15 nm. Further increases in Λ
resulted in a gradual decrease in H. Similar findings were reported for
polycrystalline TiN/NbN superlattice coatings on M2 tool steel sub-
strates where 6-µm-thick coatings with Λ = 6 nm showed H = 5200 kg
mm-2.[90] Attempts have been made to explain the hardness results.
Three mechanisms for superlattice hardening have been proposed
and were discussed in Ref. 88 for the case of nitrides.
Figure 31. A plot of the Vickers microhardness H vs wavelength Λ for nitride

superlattices with different lattice mismatch values including TiN/NbN (3.6%),
TiN/V0.6Nb0.4N (<0.2%), and TiN/VN (2.4%).
First, increases in elastic moduli, for example, the supermodulus

effect, may increase H.[88] This mechanism seems unlikely, how-
ever, since Brillouin scattering results[130] show no evidence of an
elastic anomaly in TiN/NbN superlattices. Furthermore, a
supermodulus effect would yield increased H only at a specific Λ
value.
Second, Cahn[131] predicted that the alternating stress fields in
compositionally modulated materials inhibit dislocation motion. For
Λ < 9 nm in Fig. 31, where H was larger for lattice mismatched
compared to lattice-matched superlattices, XRD and XTEM showed
that substantial coherency strains were present,[88] suggesting that
coherency strains played a role in this additional hardness en-
hancement. As a matter of fact, the Λ range over which the enhance-
ment was observed was consistent with composition amplitude mea-
surements, and strain relaxation theory.[132] Also, superlattices of
lattice-matched TiN/V0.6Nb0.4N showed a relatively smaller hard-
ness increase compared to coherently strained TiN/V0.7Nb0.3N
superlattices with a 1.7% lattice mismatch.[132] The hardness peak
around Λ = 5 nm in the example of Fig. 31 would then correspond
physically to a wavelength large enough that interdiffusion is mini-
mal, yet small enough that significant strains are present. Interdiffu-
sion on the order of nm during growth of nitride superlattice films
occurs at temperatures above 600–700°C.[88]
Third, Koehler,[133] Lehoczky,[134] and Pacheco and Mura[135]
developed theories for the strengthening of multilayers based on
resistance to dislocation flow across layers with different dislocation
line energies, i.e., different shear moduli. The strength enhancement
is generally predicted to be proportional to the quotient Q written as
GA − GB
Eq. (2) Q
GA + GB
where GA and GB are the shear moduli of the superlattice layers A and
B, respectively.[136]
Recent results from nitride superlattices with different Q val-

ues, but a similar lattice mismatch, verify that the difference in
elastic moduli is the primary mechanism of nitride superlattice
hardening.[136] For example, for NbN/V0.6Nb0.4N with Q ≈ 6,
superlattice hardness varied little with Λ, reaching a maximum value
only ~400 kg mm-2 above the alloy and rule-of-mixtures values at
Λ = 7,[136] while for TiN/NbN with Q = 11–15, hardness enhance-
ments of up to 3000 kg mm-2 were observed.[88] In Fig. 31, the
decrease in H as Λ is decreased below 5 nm is believed to arise from
two sources. First, the shear stress in the softer nitride layer should
decrease as the layer thickness decreases. This results from the fact
that the dislocation strain field is mainly outside this layer for a layer
thickness of the same magnitude as the dislocation displacement
vector (Burgers’ vector) b, and hence, the associated energy barrier
to move dislocations into the neighboring layer is small. Second, the
composition modulation amplitude decreases with decreasing Λ as
discussed above. This decreases the difference in shear moduli
between the layers, and hence decreases the critical resolved shear
stress. The combination of these two effects yields the sharp drop in
H below Λ = 5 nm. The decrease in H with increasing Λ for Λ above
5 nm as shown in Fig. 31 is explained by the operation of dislocation
sources within individual layers, with stress concentrations at dislo-
cation pileups leading to slip across the layers.
Finally, the role of grain size on the hardness enhancements is
not excluded. The Hall-Petch dependence H = Ho + kd-½ (Eq. 1), has
also been used to fit data for superlattices with large Λ val-
ues.[21][88][137] Also, Lehoczky obtained a dependence of the critical
resolved shear stress given by σcr = σo + cΛ-1.[134] A best fit to the
results for TiN/NbN superlattices is given by H - HNbN = Λ-0.3±0.1 ,[88]
where HNbN = 1900 kg mm-2.
The relative importance of the above, given different factors
that influence the coating hardness is today not completely known,
and more work is needed in this area. Particularly promising should
be if measurements of the hardness, as well as other mechanical
properties, using the nano-indentation method,[138] are combined
with detailed structural studies.
5.0 ADHESION AND COATING/SUBSTRATE

INTERFACE STUDIES
Normally, failure of a hard coating is not by progressive wear,

but rather by either cohesive failure or failure of the adhesive bond
between coating and substrate. The criteria for selecting a particular
coating material should then depend on the substrate to be used, as
well as on the actual application. Also, there is a need for a fundamental
understanding of both coating/substrate interactions and interactions
between the coating, and its environment during operation.
Coating adhesion depends ultimately on the atomic bond
strength to the substrate, and on the interface microstructure. Chemi-
cal bonds are the strongest (several eV) followed by van der Waals
bonds (0.1–0.4 eV), and electrostatic bonds (0.1–0.4 eV). Because of
the complicated interactions between coating, and substrate de-
scribed earlier in this chapter, there is, as yet, neither a clear under-
standing of the various factors which control the coating integrity,
nor any accepted experimental method of measuring it. Rickerby[139]
suggests that the adhesion can be measured either as (i) the force
needed to separate two materials or (ii) the work done in separating
unit area of the materials. In the latter case the reversible thermody-
namical work WAB can be written as
Eq. (3) WAB = γA + γB - γAB
where γA and γB are the specific surface free energies of materials A

and B, and γAB is the specific free energy of the interface. However,
Tabor[140] has shown that this condition underestimates the work of
adhesion, when one or both of the materials can undergo plastic
deformation. Also Eq. (3) takes no account of mechanical interlock-
ing or interdiffusion of the two surfaces, both of which will increase
the bonding between them. However, even if this model is simple, it
indicates the importance of the energy of the interface. As can be
seen from Eq. (3) a decrease in γAB will give an increase of the
adhesion work.
Rickerby[139] and Valli[141] have reviewed methods of adhe-

sion testing including scratch test, pressure-sensitive tape test, accel-
eration (body force) test, electromagnetic stressing, shock wave test,
tensile and shear test, rolling with slip, laser techniques, acoustic
imaging, and indentation tests. The most common current test is the
scratch test, in which a hard indenter is drawn across the coating
surface under an increasing normal load L, until a change in the
adherence, or coherence of the coating is indicated by either acoustic
emission, a sudden change in friction coefficient, or observed breakup
or delamination of the coating. The force of the critical normal load
LC at which this change occurs is then taken as a comparative
measure of adhesion.
The results, however, depend on factors such as coefficient of
friction between the tip and the film, loading rate, mechanical prop-
erties of the film and substrate, and the film thickness. Typically, LC
increases with increasing substrate hardness and coating thickness.
The standard deviation of the scratch test results is usually about
10–20%. This must be kept in mind when comparing the adhesion of
different coatings. However, in spite of the difficulties in calculating
the adhesion from the LC values, they can be used to give a qualita-
tive information that is adequate for comparison between different
samples.[95][96]
Although the empirical results so far does not dictate a unique
mechanism for adhesion-enhancing processes, some general require-
ments for adhesion enhancements can be noted. The adhesion of
the coating can be enhanced in a number of different ways. Most
important is the elimination of all contamination from the substrate
prior to coating. In high-temperature coating processes, such as CVD,
organic contaminants are readily desorbed from the surface. In low-
temperature processes, such as PVD, contamination by organic films
becomes increasingly more troublesome. In-situ sputter etching has
been employed to remove contamination layers and surface oxides in
PVD processes.[24][25][142] The coating/substrate interface microstruc-
ture and microchemistry can be investigated by using cross-sectional
scanning TEM in combination with energy dispersive x-ray analysis.
In addition to removing surface contaminants, it is also essen-

tial to avoid a low nucleation density, since lateral growth results in
interfacial voids with poor adhesion due to the reduced contact area,
and facilitated crack propagation. Sputter-etching, prior to deposi-
tion can form additional nucleation sites, and promote surface diffu-
sion. However, sputter-etching also may cause changes in surface
composition, and morphology through preferential sputtering.[143] A
roughened substrate surface may assist in developing a mechanical
interlocking between the coating and substrate phases which in turn
would increase the interfacial shear strength. However, care should
be taken not to introduce additional crack generators at the surface.
Elevated substrate temperature can cause surface segrega-
tion [145]–[147] , and increased interdiffusion during growth.[148] For
films grown by CVD, a mixed interfacial zone generally develops
due to interdiffusion since high temperatures, and low deposition
rates are used. This gives rise to a strong metallurgical bond resulting
in high adhesion provided that no brittle phases are formed. Com-
pounds with high chemical affinity interact strongly, and has a high
mutual solubility indicating high adhesion. Thus, it should be desir-
able to maximize the chemical affinity of the coating for the sub-
strate, and to keep any produced brittle phases thin. Examples of
where interdiffusion at the interface contributes to strong bonding is
TiC to WC in cemented carbides,[144] or TiC to Al2O3.[148] Interfaces
between covalent hard materials seems to have little interaction with
the consequence of poor adhesion of the phases, e.g., B4C/Al2O3.[149]
During PVD, a graded interfacial zone can be formed either by
depositing an interfacial layer[95] or by using ion bombardment to
promote collisional intermixing at the interface.[150]
For PVD coatings grown at low temperatures, the adhesion can
also be improved by using substrates that are structurally similar to
the coating materials.[95][96] A structural, and chemical matching
between coating, and substrate results in a low interface energy, and
thus, promoting high adhesion, e.g., TiN/VC,[95] and TiC/TiB2.[149]
Helmersson, et al.,[95] have shown that the adhesion, measured
as the critical load in scratch test experiments of TiN films deposited
onto high speed steel substrates increased as the number density of
VC particles was increased. Since both TiN and VC has a B1-NaCl

structure with only a small difference in lattice parameters, a high
degree of structural matching is achieved (see Fig. 21b).
The in-plane stress discontinuity at the interface described in
Sec. 4.1 could be converted to a stress gradient by creating composi-
tion gradients between the substrate and coating. This can be
obtained during deposition, by modulating the fluxes from different
sources, or by ion-induced intermixing.
In the case of ion-induced collisional intermixing, the ion
irradiation would assist migration of interface atoms until a preferred
configuration is obtained. This could be obtained either during
growth by attracting low-energy ions (typically ~100 eV) from a
glow discharge, or after growth by implantation of high-energy ions
(> 10 keV) from an external source. The good adhesion observed for
ion plated coatings has been ascribed to ion-induced interdiffusion
where a high flux of low-energy ions (~100 eV) gets implanted only
a few atomic layers deep. The mechanism is, however, not clear.
Provided that a low-energy interface configuration is created, exces-
sive irradiation would have no effect, or could even be detrimental to
the adhesion by entrapment of the incident species causing interface
embrittlement. For example, a continuous decrease in the critical
load from 4.0 kg at Vs = 0 V to 0.1 kg at Vs = 1700 V (Jion/JTi ≈ 0.55)
was observed for sputter-deposited TiN on high-speed steel when the
negative bias voltage was applied during the initial stages of
growth.[150] Figure 32 shows XTEM micrographs of the microstruc-
ture of bilayered TiN films for (a) Vs = 0 V for the first 125 nm and
(b) Vs = 1700 V for the first 125 nm. For both cases, the remainder of
film growth was at Vs = floating potential. The film with Vs = 0 V/
floating potential in Fig. 32a exhibits a columnar microstructure with
voids along the grain boundaries. The film/substrate interface ap-
pears well defined and free of contaminants. The Vs = 1700 V layer
in Fig. 32b shows low-density regions which are elongated in the
substrate plane, while the Vs = floating potential layer rapidly devel-
oped a columnar microstructure identical to that in Fig. 32a for the
corresponding layer. The rough film/substrate interface for the
Vs = 1700 V layer is indicative of intermixing by the Ar/N2 ion

bombardment during growth. For Vs = 1700 V, the Ar content was
approximately 5.5 at%, indicating that the low-density regions ob-
served were due to Ar gas bubble precipitation close to the interface
which in turn resulted in a very poor adhesion.
All of the above mentioned process parameters, and phenom-
ena not only has a direct influence on film adhesion, but since they
induce chemical and structural changes at the substrate surface, they
also affect the nucleation and growth of the films, and n.b., thereby
also the structure of the films.
(a)
Figure 32. Cross-sectional TEM micrographs showing the effects of substrate

bias Vs on the film/substrate interface region for sputter-deposited TiN on high-
speed steel. (a) Bilayer TiN film with Vs = 0 V and floating potential (-12 V),
respectively. (b) Bilayer TiN film with V s = 1700 V and floating potential (-12 V),
respectively. The adhesion measured as critical load in the scratch test was 4.0 kg
in (a) and 0.1 kg in (b).
(b)
6.0 SUMMARY
Most hard coatings represent structures that are far from ther-
modynamical equilibrium with grain boundaries, texture, and crys-
tallographic defects as an integral part of the material. The relation-
ships between deposition processes, microstructure, and properties
in hard coatings are exclusively based on empirical knowledge. Only
a very limited number of materials, for example, TiN and diamond,
have been studied in some depth, and it has been shown how the thin
film microstructure can be related to individual deposition param-
eters on a physical or phenomenological basis. However, the param-
eters are often hard to vary independently, and their combined
relationship is often complex. This makes the prediction of coating

microstructure as a function of deposition conditions quite delicate.
Also, the rapid development of the deposition processes, and the
insufficient recording of important deposition parameters, like plasma
potential, substrate ion current, substrate temperature etc., makes
comparisons of coatings of even identical materials difficult.
The thorough investigations needed for better understanding
film growth employ plasma-probe measurements, sophisticated ma-
terials analysis like cross-sectional transmission electron micros-
copy and high-resolution x-ray diffraction in combination with
theoretical modeling. For the characterization of coating proper-
ties, standard methods for measuring hardness and adhesion exist
although the results are often confused by substrate-dependence
or by the inability to predict the behavior of the coating in an
actual application.
As a conclusion it can be stated that each coating/substrate
combination should be optimized for any specific application. This
is difficult to do in many cases, and the development of both new
types of substrates, and coating materials is desirable.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the financial assistance of

the Swedish National Science Research Council (NFR), Prof. U.
Helmersson, Prof. G. Håkansson, and Mr. U. Wahlström at Linköping
University, and Prof. J. E. Greene, and Prof. I. Petrov at University of
Illinois, are acknowledged for useful discussions.
REFERENCES
1. Toth, L. E., in: Transition Metal Carbides and Nitrides, Academic,

New York (1971)
2. Almond, E. A., Vacuum, 34:835 (1984)
3. Sundgren, J. E., and Hentzell, H. T. G., J. Vac. Sci. Technol. A,
4:2259 (1986)
4. Sproul, W. D., J. Vac. Sci. Technol. A, 4:2874 (1986)
5. Bunshah, R. F., Vacuum, 27:353 (1978)
6. Oakes, J. J., Thin Solid Films, 107:159 (1983)
7. Török, E., Perry, A. J., Chollet, L., and Sproul, W. D., Thin Solid
Films, 153:37 (1987)
8. Angus, J. C., Hayman, C. C., and Hoffman, R. W., in: Diamond
Optics, (A. Feldman and S. Holly, eds.), Proc. SPIE 969, p. 2 (1989)
9. Cohen, M. L., Phys. Rev. B, 32:7988 (1985)
10. Liu, A. Y., and Cohen, M. L., Science, 245:841 (1989); Phys. Rev.
B, 42:10727 (1990)
11. Ascroft, N. W., and Mermin, N. D., in: Solid State Physics, p. 387,
Holt, Rinehart, and Winston, NY (1976)
12. Plendle, J. N., and Gielisse, P. J., Phys. Rev. B, 125:828 (1962)
13. Gilman, J. J., J. Appl. Phys., 41:1664 (1970)
14. Murr, L. E., in: Interfacial Phenomena in Metals and Alloys, p. 285,
Addison Wesley, Reading (1975)
15. Grimvall, G., and Thiessen, in: Science of Hard Materials, (E. A.
Almond, Brockes, and R. Warren, eds.), p. 61, Institute of Physics
Conference Series No. 75, IOP, Bristol (1986)
16. Lawn, B., and Wilshaw, R., J. Mater. Sci., 10:1049 (1975)
17. Haasen, P., in: Physical Metallurgy, Cambridge University Press,
Cambridge (1978)
18. Komiya, S., Ono, S., and Umezu, N., Thin Solid Films, 45:473
(1977)
19. Angus, J. C., and Hayman, C. C., Science, 241:913 (1988)
20. Wierenga, P. E., Dirks, A. G., and Van den Brock, J. J., Thin Solid
Films, 119:375 (1984)
21. Cammarata, R. C., Schlesinger, T. E., Kim, C., Qadri, S. B., and
Edelstein, A. S., Appl. Phys. Lett., 56:1862 (1990)
22. Sundgren, J. E., Rockett, A., Greene, J. E., and Helmersson, U., J.
Vac. Sci. Technol. A, 4:2770 (1986)
23. Hultman, L., Hentzell, H. T. G., Sundgren, J. E., Johansson, B. O.,
and Helmersson, U., Thin Solid Films, 124:170 (1985)
24. Münz, W. D., Schroeder, J., Petersein, H., Håkansson, G., Hultman,
L., and, Sundgren, J. E., in: Proc. of SURTEC, Berlin (October 11–
13, 1989)
25. Håkansson, G., Hultman, L., Sundgren, J. E., Greene, J. E., and
Münz, W. D., Surf. Coat. Technol., 48:51 (1991)
26. Helmersson, U., and Sundgren, J. E., J. Electron Microscopy
Techniques, 4:361 (1986)
27. Petrov, I., Hultman, L., Helmersson, U., Sundgren, J. E., and Greene,
J. E., Thin Solid Films, 169:299 (1989)
28. Håkansson, G., Sundgren, J. E., McIntyre, D., Greene, J. E., and
Münz, W. D., Thin Solid Films, 153:55 (1987)
29. Perry, A. J., and Jagner, M., Thin Solid Films, 171:197 (1989)
30. Klug, H. P., and Alexander, L. E., in: X-ray Diffraction Procedures,
Wiley, NewYork (1974)
31. Thornton, J. A., J. Vac. Sci. Technol. A, 4:3059 (1986)
32. Movchan, B. A., and Demchishin, A. V., Fiz. Met. Metalloved,
28:83 (1969)
33. Dirks, A. G., and Leamy, H. J., Thin Solid Films, 47:219 (1977)
34. Grovenor, C. R. M., Hentzell, H. T. G., and Smith, D. A., Acta
Metall., 32:773 (1984)
35. Thornton, J. A., J. Vac. Sci. Technol., 11:666 (1974)
36. Thornton, J. A., Ann. Rev. Mater. Sci., 7:239 (1977)
37. Hoffman, D. W., and Thornton, J. A., Thin Solid Films, 45:387
(1977)
38. Messier, R., Giri, A. P., and Roy, R. A., J. Vac. Sci. Technol. A,
2:500 (1984)
39. Musil, J., Kadlec, S., Valvoda, V., Kuzel, R., and Cerny, R., Surf.
Coat. Technol., 43/44:259 (1990)
40. Kay, E., Parmigiani, F., and Parrish, W., J. Vac. Sci. Technol A,
6:3074 (1988)
41. Petrov, I., Adibi, F., Greene, J. E., Hultman, L., and Sundgren, J. E.,
Appl. Phys. Letters, 63:366 (1993)
42. Messier, R., J. Vac. Sci. Technol. A, 4:490 (1986)
43. Messier, R., and Yehoda, J. E., J. Appl. Phys., 58:3739 (1985)
44. Roy, R. A., and Messier, R., Mat. Res. Soc. Symp. Proc., 38:363
(1985)
45. Hultman, L., and Sproul, W. D., Unpublished
46. Müller, K. H., J. Appl. Phys., 58:2573 (1985)
48. Müller, K. H., Appl. Phys. A, 40:209 (1986)
49. Hendersson, D. J., Brodsky, M. H., and Chaudhari, P., Appl. Phys.
Lett., 25:641 (1974)
50. Leamy, H. J., Gilmer, G. H., and Dirks, A. G., in: Current Topics in
Materials Science, (E. Kaldis, ed.), Vol. 6, North-Holland,
Amsterdam (1980)
51. Müller, K. H., Phys. Rev. B, 35:7906 (1987)
52. Müller, K. H., Surf. Sci., 184:L375 (1987)
53. Srolovitz, D. J., J. Vac. Sci. Technol. A, 4:2925 (1986)
54. Lichter, S., and Chen, J., Phys. Rev. Lett., 56:1396 (1986)
55. Greene, J. E., Barnett, S. A., Sundgren, J. E., and Rockett, A., in: Ion
Beam Assisted Film Growth, (T. Itoh, ed.), Ch. 5, Elsevier,
Amsterdam (1988)
56. Hultman, L., Münz, W. D., Musil, J., Kadlec, S., Petrov, I., and
Greene, J. E., J. Vac. Sci. Technol. A, 9:434 (1991)
57. Dobrev, D., Thin Solid Films, 92:41 (1982)
58. Harper, J. M. E., Cuomo, J. J., and Hentzell, H. T. G., J. Appl. Phys.,
58:550 (1985)
59. Adibi, F., Petrov, I., Greene, J. E., Hultman, L., and Sundgren, J. E.,
J. Appl. Phys., 73:8580 (1993)
60. Kobashi, K., Nishimura, K., Kawate, Y., and Horiuchi, T., Phys.
Rev. B, 38:4067 (1988)
61. Badzian, A. R., and De Vries, R. C., Mat. Res. Bull., 23:385 (1988)
62. Yarbrough, W. A., and Messier, R., Science, 247:688 (1990)

63. Yarbrough, W. A., Mat. Res. Soc. Symp. Proc., 162:75 (1990)
64. Messier, R., Badzian, A. R., Badzian, T., Spear, K. E., Bachmann,
P., and Roy, R., Thin Solid Films, 153:1 (1987)
65. Bachmann, P., Leers, D., and Lydtin, H., Diamond and Related
Materials, 1:1 (1991)
66. Clausing, R. E., Heatherly, L., and Specht, E. D., in: Diamond and
Diamond-like Films and Coatings, (R. E. Clausing, L. L. Horton, J.
C. Angus, and P. Koidl, eds.), p. 611, Plenum Press, New York
(1991)
67. Catherine, Y., in: Diamond and Diamond-like Films and Coatings,
(R. E. Clausing, L. L. Horton, J. C. Angus, and P. Koidl, eds.), p.
193, Plenum Press, New York (1991)
68. Bachmann, P. K., and Lydtin, H., in: Diamond and Diamond-like
Films and Coatings, (R. E. Clausing, L. L. Horton, J. C. Angus, and
P. Koidl, eds.), p. 829, Plenum Press, New York (1991)
69. Angus, J. C., and Wang, Y., in: Diamond and Diamond-like Films
and Coatings, (R. E. Clausing, L. L. Horton, J. C. Angus, and P.
Koidl, eds.), p. 173, Plenum Press, New York (1991)
70. Sjöström, H., Hultman, L., Sundgren, J. E., and Wallenberg, L. R.,
Thin Solid Films, 232:169 (1993)
71. Schuster, J. S., Bauer, J., J. Solid State Chem., 53:260 (1984)
72. Holleck, H., Surf. Coat. Technol., 36:151 (1988)
73. Wahlström, U., Hultman, L., Sundgren, J. E., Adibi, F., Petrov, I.,
and Greene, J. E., Thin Solid Films, 253:62 (1993)
74. Tanaka, Y., et al., J. Vac. Sci. Technol. A, 10:1749 (1992)
75. Ikeda, T., Satoh, H., Thin Solid Films, 195:99 (1991)
76. Adibi, F., Petrov, I., Hultman, L., Wahlström, U., Shimizu, T.,
McIntyre, D., Greene, J. E., and Sundgren, J. E., J. Appl. Phys.,
69:6437 (1991)
77. Adibi, F., Petrov, I., Greene, J. E., Wahlström, U., and Sundgren, J.
E., J. Vac. Sci. Technol. A, 11:136 (1993)
78. Powder Diffraction File, JCPDC International Center for Powder
Diffraction Data, Card number 25–1495, Swarthmore, PA (1989)
79. Desmaison, J., Lefort, P., and Billy, M., Oxid. Met., 13:203 (1979)
80. McIntyre, D., Greene, J. E., Håkansson, G., Sundgren, J. E., and
Münz, W. D., J. Appl. Phys., 67:1542 (1990)
81. Sundgren, J. E., Birch, J., Håkansson, G., Hultman, L., and
Helmersson, U., Thin Solid Films., 193:819 (1990)
82. Holleck, H., Kühl, C., and Schultz, H., J. Vac. Sci. Technol. A,
3:2345 (1985)
83. Holleck, H., and Schultz, H., Thin Solid Films, 153:11 (1987)
84. Mirkarimi, P. B., Hultman, L., and Barnett, S. A., Appl. Phys. Lett.,
57:2654 (1990)
85. Birch, J., Yamamoto, Y., Hultman, L., Radnoczi, G., Sundgren, J.
E., and Wallenberg, L. R., Vacuum, 41:1231 (1990)
86. Barnett, S. A., in Physics of Thin Films, (J. L. Vossen and M.
Francombe, eds.), Academic Press, New York, 17:1 (1993)
87. Helmersson, U., Todorova, S., Barnett, S. A., Sundgren, J. E.,
Markert, L. C., and Greene, J. E., J. Appl. Phys., 62:481 (1987)
88. Shinn, M., Hultman, L., and Barnett, S. A., J. Mater. Res., 7:901
(1992)
89. Ljungcrantz, H., and Svensson, M., Master Thesis, Linköping
University, Sweden, LiTH-IFM-EX-423 (1989)
90. Chu, X., Barnett, S. A., Wong, M. S., and Sproul, W. D., Surf. Coat.
Technol., 57:3 (1993)
91. Miller, D. J., Gray, K. E., Kampwirth, R. T., and Murduck, J. M.,
Europhysics Letters, 19:27 (1992)
92. Hultman, L., Wallenberg, L. R., Shinn, M., and Barnett, S. A., J.
Vac. Sci. Technol. A, 10:1618 (1992)
93. Lewis, B., and Andersson, J. C., in: Nucleation and Growth of Thin
Films, Academic Press, London (1978)
94. Van der Drift, A., Philips Res. Rep., 22:267 (1967)
95. Helmersson, U., Johansson, B. O., Sundgren, J. E., Hentzell, H. T.
G., and Billgren, P., J. Vac. Sci. Technol. A, 3:308 (1985)
96. Helmersson, U., Hentzell, H. T. G., Hultman, L., Hibbs, M. K., and
Sundgren, J. E., in: Physics and Chemistry of Protective Coatings,
(W. D. Sproul, J. E. Greene, and J. A. Thornton, eds.), p. 79,
American Institute of Physics, New York (1985)
97. Dubray, J. J., Yarbrough, W. A., and Pantano, C. G., in: Diamond
and Diamond-like Films and Coatings, (R. E. Clausing, L. L. Horton,
J. C. Angus, and P. Koidl, eds.), p. 619, Plenum Press, New York
(1991)
98. Yarbrough, W. A., Kumar, A., and Roy, R., Mat. Res. Soc., Fall
1987 Meeting, Boston, MA
99. Yugo, S., and Kimura, T., in: First Int. Conf. on the New Diamond
Sci. and Technol., Program and Abstracts, p. 130, JNDF, Tokyo
(Sept. 24–26, 1988)
100. Noyan, I. C., and Cohen, J. B., in: Residual Stresses, Springer
Verlag, New York (1987)
101. Finnegan, J. D., and Hoffman, R. W., AEC Tech. Rep. 8, Case
Institute of Technology, Cleveland, OH (1961)
102. Klokholm, E., and Berry, B. S., J. Electrochem. Soc., 115:823
(1968)
103. Windischmann, H., J. Appl. Phys., 62:1800 (1987)
105. Hoffman, D. W., and Thornton, J. A., J. Vac. Sci. Technol., 16:134
(1979)
(1980)
(1982)
108. Thornton, J. A., and Hoffman, D. W., J. Vac. Sci. Technol., 18:203
(1980)
109. Thornton, J. A., and Hoffman, D. W., Thin Solid Films, 171:5
(1989)
110. Rickerby, D. S., and Bull, S., J. Surf. Coat. Technol., 39/40:315
(1989)
111. Window, B., Sharples, F., and Parrish, W., J. Vac. Sci. Technol. A,
6:2333 (1988)
112. Window, B., and Müller, K. H., Thin Solid Films, 171:183 (1989)
113. Ho, H. L., Kampwirth, R. T., Gray, K. E., Capone, D. W., II,
Chumbley, L. S., and Meshii, M., Ultramicroscopy, 22:297 (1987)
114. Ernsberger, C., Perry, A. J., Lehman, L. P., Miller, A. E., Pelton, A.
R., and Dabrowski, B. W., Surf. Coat. Technol., 36:605 (1988)
115. Petrov, I., Hultman, L., Sundgren, J. E., and Greene, J. E., J. Vac.
Sci. Technol. A, 10:265 (1992)
116. Arnell R. D., in: Proc. Seventh International Conference on Ion &
Plasma Assisted Techniques, p. 226, CEP Consultants Ltd,
Edinburgh, UK (1989)
117. Burnett, P. J., and Rickerby, D. S., Thin Solid Films, 148:51 (1987)
118. Burnett, P. J., and Page, T. F., J. Mater. Sci., 19:845 (1984)
119. Jonsson, B., and Hogmark, S., Thin Solid Films, 114:257 (1984)
120. Bull, S. J., Page, T. F., and Yoffe, E. H., Phil. Mag. Lett., 59:281
(1989)
121. Augis, J. A., Lo, C. C., and Pinnel, M. R., Thin Solid Films, 58:357
(1979)
122. Rickerby, D. S., and Burnett, P. J., Thin Solid Films, 157:195 (1988)
123. Münz, W. D., and Hessberger, G., Vak. Tech., 30:78 (1981)
124. Hibbs, M. K., Johansson, B. O., Sundgren, J. E., and Helmersson,
U., Thin Solid Films, 122:115 (1984)
125. Raghuram, A. C., and Bunshah, R. F., J. Vac. Sci. Technol., 9:1289
(1972)
126. Pierson, H. O., and Mullendore, A. W., Thin Solid Films, 72:511
(1980)
127. Marschall, J., and Milstein, F., J. Phys. F, 18:1913 (1988)
128. Jankowski, A. F., and Tsakalakos, T., J. Phys. F, 15:1279 (1985)
129. Freund, L. B., J. Appl. Mechanics, 54:553 (1987)
130. Mirkarimi, P. B., Shinn, M., Kumar, S., Grimsditch, M., and Barnett,
S. A., J. Appl. Phys., 71:4955 (1992)
131. Cahn, J. W., Acta Metall., 11:1274 (1963)
132. Mirkarimi, P. B., Barnett, S. A., Hubbard, K. M., Jervis, T. R, and
Hultman, L., J. Mater. Res., 9:1456 (1994)
133. Koehler, J. S., Phys. Rev. B, 2:547 (1970)
134. Lehoczky, S. L., J. Appl. Phys., 49:5479 (1978); Lehoczky, S. L.,
Phys. Rev Lett., 41:1814 (1978)
135. Pacheco, E. S., and Mura, T., J. Mech. Phys. Solids, 17:163 (1969)
136. Shinn, M., and Barnett, S. A., Appl. Phys. Lett., 64:61 (1994)
137. Menezes, S., and Anderson, D. P., J. Electrochem. Soc., 137: 440
(1990)
138. Pethica, J. B., Hutchings, R., and Oliver, W. C., Phil. Mag. A,
48:593 (1983)
139. Rickerby, D. S., Surf. Coat. Technol., 36:541 (1988)
140. Tabor, D., in: New Directions in Lubrication etc., (Loomis, ed.), p.
116, Noyes Publications, Park Ridge, NJ (1985)
141. Valli, J., J. Vac. Sci. Technol. A, 4:3007 (1986)
142. Holland, L., Chapman and Hall (1970)
143. Coburn, J. W., Thin Solid Films, 64:371 (1979)
144. Kramer, B. M., Thin Solid Films, 108:117 (1983)
145. Kumar, D., Mookerjee, A., and Kumar, V., J. Phys. F, 6:725 (1976)
146. Al-Jaroudi, M. Y., Hentzell, H. T. G., and Bengtsson, A., Thin Solid
Films, 170:293 (1989)
147. Al-Jaroudi, M. Y., Hentzell, Hörnström, S. E., and Bengtsson, A.,
Thin Solid Films, 182:153(1989)
148. Sundgren, J. E., Hibbs, M. K., Helmersson, U., Jacobson, B. E., and
Hentzell, H. T. G., J. Vac. Sci. Technol. A, 1:301 (1983)
149. Pan, A., and Greene, J. E., Thin Solid Films, 97:79 (1982)
150. Sundgren, J. E., Hultman, L., Håkansson, G., Birch, J., and Petrov,
I., Mat. Res. Soc. Symp. Proc., 268:71 (1992)
Characterization of Hard Coatings 181
Characterization of Hard
Coatings
Steve J. Bull and David S. Rickerby
1.0 INTRODUCTION
Surface coating techniques offer the possibility of modifying

the surface properties of a component, and thereby achieving im-
provements in both performance, and reliability. For instance, the
use of hard ceramic coatings deposited by physical vapor deposition
(PVD)[1] or chemical vapor deposition (CVD)[2] onto cutting tools
can result in significant improvements in tool lifetime over the
uncoated material. In order that such benefits be achieved, it is
important to understand the interrelationship between the properties
of the coating and its performance, and how these properties may be
controlled during the coating process. Figure 1 illustrates those
properties of the coating/substrate system, in terms of surface (coat-
ing dependent), bulk coating (coating and substrate dependent), and
interfacial properties. If the performance of a coated component is to
be fully understood, it is important to optimize each of these regions
and to understand the interrelationships which exist between each of
them—a very exacting goal.
181
Figure 1. Properties of a coating/substrate system which are important in

determining its performance in technological applications.
Clearly, many of these properties are only useful in certain

applications. However, there is a need to characterize a given coating
by a few fundamental properties. Of these, hardness, residual stress,
and adhesion are most relevant to a wide range of applications. These
properties are, in principle, easy to measure but, in practice, obtain-
ing useful experimental values is extremely difficult. Considerable
benefits would be gained if these basic parameters could be de-
scribed theoretically, and some progress has been made in the areas
of coating/substrate adhesion and hardness measurement. Until such
time as adequate theoretical models exist, it is important to make the
best use of current ideas to extend our understanding of the coating/
substrate system, and ultimately to relate this to the performance of
the coating in engineering applications.
2.0 EVALUATION OF THE HARDNESS OF THIN FILMS
To obtain a bulk hardness value for a film (i.e., one that is not
influenced by the presence of the underlying substrate), it is neces-
sary to satisfy the requirement that the film thickness should be
several times (typically ten) greater than the penetration depth of the
indenter in order that the measured hardness is not influenced by the
substrate deformation beneath the coating. This criterion is often
difficult to fulfill using standard microhardness equipment since
wear-resistant coatings are often deposited as layers only 1–5 µm
thick. Even when the indentation is contained within the coating,
many workers ignore the effects of indentation size on measured
hardness, even though it is well known that ceramic materials (and
many metals) show increased hardness at low loads (Ù 0.1 kgf).[3][4]
High hardness quoted in the literature may be a manifestation of this
indentation size effect (ISE).
There are two approaches to measuring the hardness of a
coated surface:
(1) Thick coatings. Measuring hardness at a range of
loads allows both the hardness at a fixed indentation
size (usually 10 µm), and the ISE index to be
determined directly.
(2) Thin coatings. A modeling approach must be used to
determine the coating hardness, and ISE index since
the substrate will now contribute to the measured
hardness.
In the field of hardness testing, models now exist which allow
the hardness of the coating to be determined independent of the
underlying substrate.[5][6] The advantages that hardness modeling
(an indirect technique), offers over direct hardness measurement will
be highlighted by reference to physical vapor deposited (PVD)
coatings.
2.1 Direct Coating Hardness Measurement
Although indentation hardness testing has been used routinely

for many years, it is not generally possible to measure the properties
of a thin film independent of the substrate, unless the coating is very
thick, using conventional testing equipment. A number of specialized
instruments have been built which can now get around this prob-
lem,[7]–[9] and these have recently become commercially available,
often called ultra-low load microhardness testers or nanoindenters. A
common feature of these instruments in continuous monitoring of
the load and displacement as the indentation is produced. A sche-
matic of a depth-sensing indentation tester is shown in Fig. 2a, and a
typical load-displacement curve recorded on this type of instrument
is shown in Fig. 2b. The feature of continuous depth and load
recording allows thin film properties to be obtained directly from the
data without the need to measure indentation diagonals. Since it is
often very difficult to find and image indentations made at these low
loads in the scanning electron microscope, the measurement errors
associated with hardness testing at low loads are greatly reduced.
(a)
Figure 2. (a) Schematic of an ultra low load hardness tester, nanoindenter, and
(b) example of the load-displacement curve it produces.
(b)
In order to obtain the hardness from such depth sensing inden-

tation tests, the projected area of the indentation needs to be deter-
mined from the load/displacement data. Since the measured dis-
placements include both elastic and plastic contributions it is
necessary to remove the elastic contributions to calculate the plastic
depth from which the area can be determined using the geometry of
the indenter. It is these elastic effects which give rise to the changes
in indentation diagonal on unloading that are responsible for some of
the indentation size effect behavior. [10] For this reason,
nanoindentation hardness measurements are not complicated by
large ISE effects, though these do still occur at very low indenter
penetrations. Methods for removing the elastic displacement, and
determining the projected area from calibrated indenter shapes
have been developed which allow accurate hardness values to be
determined.[11]
Both diamond and titanium nitride films have been investi-

gated by this technique.[12][13] Values of coating hardness, and
Young’s modulus can thus be obtained from depth-sensing inden-
tation experiments, and this has led to one such instrument being
referred to as a mechanical-properties microprobe (MPM). [13]
With a slight modification, both properties can be determined as a
function of indenter penetration, and this allows the study of ion
implanted layers, as well as, the variation in properties with coating
thickness.[14]
Given the small size of the impressions, and the size of the
loads needed to produce them, the nanohardness testers can be very
susceptible to vibration, and the roughness of the sample to be tested.
For this reason, together with their large cost, compared to conven-
tional microhardness testing apparatus, it is unlikely that these ma-
chines will be routinely used for hardness evaluation in the near
future, and other methods of thin films hardness measurement still
have a role to play.
2.2 Modeling the Hardness Behavior of Thin Films
A number of models exist for determining the contributions to

the measured hardness from the substrate and the coating.[5][6][15]–[17]
Perhaps the best approach for coatings in the thickness range 1–10
µm is the volume law-of-mixtures model which was originally
suggested by Sargent,[15] and was later extended by Burnett, and
Rickerby[5][6] to take account of the effect of the presence of the
coating on the shape of the deforming volumes. The hardness of the
coating/substrate system, Hc, is given by
Vf Vs 3
Eq. (1a) Hc = Hf + χ Hs for H f > H s
V V
Vf Vs
Eq. (1b) Hc = χ3H f + H s for H f < H s
V V
Here Hf and Hs are the hardness of the coating and substrate material
respectively. Both the substrate and coating hardness values will be a
function of the contact size—the variation is usually given by the
Meyer Index or indentation size effect (ISE) index approach
Eq. (2) H = qd m −2
where q is a constant hardness at unit diagonal size and m is the ISE

index (generally m < 2). For the substrate both q and m can be
determined from hardness data measured at a range of loads. For the
coating q and m must be chosen to give the best fit to the measured
coating/substrate hardness data. Vs and Vf are the deforming volumes
in the substrate and coating and V is the total deforming volume
given by
V = V f + χ Vs for H f > H s
3
Eq. (3a)
V = χ V f + Vs for H f < H s
3
Eq. (3b)
Chi (χ ) is an empirical factor which takes into account the modifica-

tion to the deforming volume of the softer constrained component of
the system (see Fig. 3). Chi ( χ ) is a strong function of the ratio of
plastic zone radii of the coating and substrate. The plastic zone
associated with a Vickers hardness impression in a bulk material
which behaves in an elastic/plastic manner is generally hemispheri-
cal with the deformed material being accommodated by radial dis-
placements. Using an expanding spherical cavity model to describe
the deforming volumes in Eqs. (1a) and (1b), the radius of the
plastically deformed volume, Rp, can be determined, [18]–[20] and
from this an expression for χ based on the ratio of deforming values
between the coating and substrate values derived
n
E
Eq. (4) Rp ∝  
H 
n
 Ef Hs 
Eq. (5) χ = 
 Es H f 
 
where Ef and Es are the elastic moduli of the film and substrate
material respectively. The exponent n is theoretically in the range
1/2 to 1/3 but in practice is best determined by fitting to experimental
data.[20][21]
Figure 4a shows a typical fit for Eq. 5 to experimental data for
a 14-µm TiN coating on a stainless steel substrate. The fitting
parameters are quoted in Table 1. The exponent n is close to the
theoretical value 0.5 and a good fit to the experimental data is
achieved. The variation of hardness with load for the composite
system (i.e., coating plus substrate), together with the best fit coating
hardness behavior is shown in Fig. 4b. The composite hardness has
been calculated using portions of a hemisphere for the deforming
volumes in the coating and substrate. However, as the coating thick-
ness is reduced, a number of problems in the fitting are observed.
Figure 5a shows the χ vs (Ef Hs /Es Hf ) fit for a 2-µm TiN coating on
stainless steel. Though a good fit close to the theoretical behavior is
maintained for low loads where the indenter penetration is small, this
is not true at higher loads. The hardness behavior as shown in Fig.
5b, and the parameters in Table 1 were determined from the low load
data. The reason for this deviation from theoretical behavior can be
seen from the scanning electron micrographs of Vickers indents at a
range of loads in Fig. 6. As the load increases, the coating undergoes
an increasing amount of fracture as it is pressed into the indentation
and in this case the deforming volume in the coating will no longer
be a simple slice through a hemispherical volume. In such a case, the
deforming volumes need to be modified to give closer to the theoreti-
cal behavior[21] but, it is not always possible to determine precisely
what the deforming volumes may be. Thus, the hardness and ISE
index of the coating are best determined from the low load data
where cracking within the indent is not such a problem, which
effectively imposes a minimum thickness of coating of 1 µm where
this modelling approach is useful.
Figure 3. Expected modifications to plastic zone morphology when the harder

material constrains deformation in the softer (where interfacial adhesion is good).
(a)
Figure 4. Variation of (a) χ with interface parameter and (b) measured hardness
of substrate and coating/substrate system and fitted coating hardness with load
for a 14-µm TiN coating on stainless steel. In (b) the volume law-of-mixtures
hardness model has been used to analyze the composite hardness data (o) and the
fits to experimental data are given by curve 3. The coating (curve 2) and substrate
(curve 1) hardness characteristics are also shown.
(b)
Table 1. Fitted Hardness Parameters for TiN Coatings on Stainless

Steel
Thickness Film Hardness, Film ISE Index n

Hf* mf
2 2300 1.67 0.51+

14 2600 1.6 0.5
*Hardness at 10 µm indentation diagonal.

+In low load region.
(a)
(b)
Figure 5. Variation of (a) χ with interface parameter and (b) measured substrate
and coating/substrate composite hardness with load together with fitted coating
hardness behavior for a 2 µm TiN coating on austenitic stainless steel. In (b) the
volume law-of-mixtures hardness model has been used to analyze the composite
hardness data (o) and the fits to experimental data are given by curve 3. The
coating (curve 2) and substrate (curve 1) hardness characteristics are also shown.
Figure 6. Scanning electron micrographs of Vickers hardness impressions in a 2

µm TiN coating on austenitic stainless steel, (a) 15 g load, (b) 25 g load, (c) 50 g
load, (d) 300 g load.
Experimental data for a range of systems gives fitted values of

the exponent n close to the theoretical range but this is not always the
case. Indeed, often good fits are achieved with n in the range 0.5 to
1.5. In these cases, the fitted ISE index of the coating is usually < 1.6,
a value which is not observed for bulk materials. This is most often
the case for very hard coatings, such as diamond-like carbon,
which are prone to fracture during indentation. Thus, the hardness
modeling approach must be used with caution, as the fitted coating
hardness is a sensitive function of what happens during the test, and

may thus not be a true hardness value.
The validity of the hardness modeling approach can be seen in
Fig. 7, where the hardness of a 2 µm titanium nitride film deposited
by sputter ion plating onto an austenitic stainless steel substrate is
plotted as a function of the substrate bias used during deposition. As
the bias is increased, the ion bombardment of the coating during
growth leads to a denser coating which can support higher stresses
and the hardness thus increases with bias.[22] Hardness data deter-
mined using direct measurements by the nanoindenter, and by con-
ventional microhardness testing (25 g Vickers indentation) is com-
pared with the modeled hardness. Though all methods show similar
trends, the magnitude of the modeled and nanoindenter values is
much higher, and the two show closer agreement than the direct
Vickers measurement.
Figure 7. Variation of coating hardness with applied substrate bias for sputter
ion plated TiN coatings on austenitic stainless steel. ● direct measurement
(nanoindenter); ▲ direct measurement (25 g Vickers); o indirect measurement
(volume law-of-mixtures hardness model).
Given the difficulty of measuring coating hardness directly, the

need for a hardness model is obvious. The volume law of mixtures
model has proved successful in a wide range of situations, but care must
be taken in its use as factors such as fracture during the indentation
process will affect its validity. The best fit coating ISE index pro-
vides a useful measure of whether the model is valid, but this should be
combined with careful post facto examination of the hardness impres-
sions in all cases where deviations from expected behavior are large.
3.0 MEASUREMENT OF COATING ADHESION
In any coating application, the minimum requirement for ac-

ceptable performance of a coated component must be a sufficient
level of coating/substrate adhesion. The slow take-up of surface
coating techniques stems, in part, from a lack of end-user confidence
in the quality of the supplied coatings, and thus, it is essential that a
simple and reliable adhesion test is available which should ideally be
of relevance to the intended application. In this section some of the
more successful methods for determining coating/substrate adhesion
are discussed and the advantages, and disadvantages of each are
briefly reviewed.
3.1 Definition of Adhesion
According to the ASTM definition (D907-70), adhesion is “the

state in which two surfaces are held together by interfacial forces
which may consist of valence forces or interlocking forces, or both.”
These bonding forces can be van der Waals forces, electrostatic
forces, or chemical bonding across the coating/substrate interface.
An important distinction must be made between this basic adhesion
(BA), which is the maximum possible attainable value, and experi-
mental or practical adhesion (EA), which also depends on other
factors such as the type of test, the residual stress levels in the film, or
the interfacial flaw distribution, such that EA >> BA. Pulker, et al.,[23]
further define the relationship of experimentally measured adhesion

to the basic adhesion as
Eq. (6) EA = BA - IS ± MSM
where IS is the internal stress and MSM is the method specific error
in measurement. From Eq. (6) it can be seen that the true value of
basic adhesion cannot generally be determined since the size of the
measurement error for each measurement technique cannot be esti-
mated. Experimentally adhesion can be measured in two ways:
1. In terms of forces. The force of adhesion is defined as
the maximum force per unit area to separate the
coating from its substrate.
2. In terms of energy. The work of adhesion WAB is
defined as the work done in separating the coating
(A), from its substrate (B) which is given by the
energy criterion
Eq. (7) WAB = γ A + γ B − γ AB
where γ A and γ B are the specific surface free energies of A and B and
γ AB is the specific interfacial free energy.
The total force of adhesion can be related to the work of
adhesion by making some assumptions about the changes in force
with separation of the surfaces,[24] but in general the work of adhe-
sion is a much more useful concept since it does not rely on so many
assumptions. If the break occurs at the interface AB then the failure is
adhesive, otherwise if it occurs within A or B it is a cohesive failure.
3.2 Requirements of an Ideal Adhesion Test
Though there are a large range of possible adhesion tests

available, none of them may be regarded as ideal, and thus, it is
interesting to consider those factors which would define the ideal test:
1. Non destructive.
2. Easily adaptable to the routine testing of complex
shapes.
3. Simple to perform and interpret.
4. Amenable to standardization and automation.
5. Reproducible.
6. Quantitative.
7. Directly related to coating reliability in specific
applications.
No test currently employed has all these attributes, and the
most commonly used tests are all destructive in nature. Indeed it is
difficult to see how such a non destructive test could be developed
given our current understanding, and thus, it is necessary to make the
best of what test we have available at the present time.
3.3 Measurement of Adhesion
Adhesion test methods have been categorized into three groups:

1. Nucleation methods.[25] Based on the measurement
of nucleation rate, island density, critical condensation,
and residence time of the depositing atoms. Such
tests require detailed electron microscopy, can be
difficult to perform or interpret, and are thus not
amenable to routine testing.
2. Mechanical methods. Adhesion is determined by the
application of a force to the coating/substrate system.
This force may be normal to the interface as in the
pull-off test, [26]–[29] or parallel to it, as in the shearing
test.[30] Alternatively, the forces may be introduced into
the coating substrate system by some other mechanical
stimulus as in the indentation test,[31]–[35] the scratch
test,[36]–[45] and laser spallation tests.[46]–[48]
3. Miscellaneous methods. In these tests, the adhesion is
inferred from the results of some indirect test method
such as x-ray diffraction,[49] thermal cycling,[50] or
capacity measurements.[24][51]
Of these tests, the mechanical methods are of most practical

interest. However, only the indentation and scratch tests have been
used extensively to characterize the adhesion of hard coatings and
these will be discussed in the next sections. The other adhesion tests
have been reviewed elsewhere.[52]
Indentation and Laser Spallation Methods. In the indenta-
tion adhesion test, a mechanically stable crack is introduced into the
coating/substrate interface by the use of conventional indentation
procedures; both Vickers and Brale indenters have been used.[53][35]
The resistance to propagation of the crack along the interface is then
used as a measure of adhesion, and by analogy with the fracture of
homogeneous brittle solids,[54][55] this may be characterized by both
a fracture resistance parameter, and a strength parameter. The frac-
ture resistance parameter relates uniquely to the bonding across the
interface, and is a more fundamental measure of adhesion, whereas
the strength is determined by the combined influences of the fracture
resistance, the strength controlling defects, and residual stresses
within the film. The test is based on the rationale that an interface (in
the vicinity of the plastic zone created during indentation) with a
lower toughness than that of either the film or the substrate material
will be a preferred site for lateral crack formation. When fracture
cannot be induced at the interface (but occurs in the film or substrate)
it can be concluded that the interface toughness is at least as large as
that of the weaker component.
The indentation adhesion test used by Jindal, et al.,[35] is shown
schematically in Fig. 8, in which a series of indentations are made,
and the changes in lateral cracking monitored as a function of applied
load. From the linear portion of the indentation load/lateral crack
length function a value of interface fracture toughness (K Ii) can be
derived (see also Fig. 8) according to
1
G E  2
Eq. (8) K Ii =  Ii c2 
 1−υ c 
where A is a constant and Ec and υ c are the Young’s modulus, and

Poisson’s ratio of the coating. A possible advantage of the indenta-
tion adhesion test over the more commonly used scratch adhesion
test is that the values of the indentation adhesion parameters PC and
KIi are relatively insensitive to substrate hardness, while the corre-
sponding scratch test parameter (Lc) is extremely sensitive to sub-
strate hardness.[35] To further illustrate the differences between these
two test methods, Fig. 9 shows indentation test results for carbide
inserts CVD coated with TiC/Al2O3 layers differentiated in terms of
the η-phase occurrence at the coating/substrate interface. On the
basis of the slope of the load/lateral crack diameter function the
coatings which are deposited onto a continuous layer of the brittle
η-phase have poorer adhesion, (toughness) compared with similar
coatings which are formed with a discontinuous η-phase layer, but
in both cases the scratch adhesion test indicated essentially identical
Lc values for adhesion failure at the coating/substrate interface.[35]
Figure 8. Schematic representation of the indentation coating adhesion test.

(After Jindal, Quinto, and Wolfe.)
Figure 9. Indentation load versus lateral crack diameter for multilayer TiC/
Al2 O3 coated samples with thin discontinuous η-phase (l) and thick continuous
η-phase (O). (After Jindal, Quinto, and Wolfe.)
The approach of Evans, and co-workers[32][33][53] is based on

the observation that in the absence of buckling, and for planar
interfaces, there is no driving force for growth of a delamination
which exists at the coating substrate interface; this initial delamina-
tion may arise due to interfacial contamination, or by void formation
and coalescence. Consequently, for such interfaces, buckling be-
comes a prerequisite for fracture propagation, and eventual spalling.
The critical stress for buckling of a circular delamination being given
by:[56]
Eq. (9) [ (
σ c = KE c 12 1 − υ c2 ) ](t a ) 2
where t is the coating thickness, a is the delamination radius, and K ≈

14.7. Once buckling occurs, a crack driving force (G) develops given
by:[57]
Eq. (10) ( ) (
G = 1 − υ c2 (1 − α )t σ 2 − σ c2 )E c
where σ is the net compressive stress in the coating and α = 0.38.

Further growth of the delamination occurs if G > Gc either for the
interface or for the coating and the delamination radius for coating
spallation, as, is given by: [57]
a s t ≈ 1.9 (E c σ )
1
Eq. (11) 2
For indentation-induced spalling[32][33][53]
Eq. (12) σ = σ R + E c V 2π (1 − υ c )t a 2
where σ R is the initial residual stress and V the indentation volume,

with
V = 0.24 (P H )3 cot ψ and a = λ P

3
Eq. (13) 4
where P is the indentation load, ψ is the indenter half-angle, H is the

hardness, and λ is an experimentally determined coefficient. The
critical indentation load for spalling, Ps, is given by
3
[
Eq. (14) Ps 2 = 3. 7 t 2 E c λ 2 σ R + 0.24 E c cotψ 2π (1 − υ c )t H
3
2
]
−1
Some results obtained (see Fig. 12) from indentation testing of

ZrO2-Y2O3 coatings,[31] and the data exhibit the expected trend (Eq.
15) that a ≈ P3\4 and the fracture toughness Gc along the delamination
path was found to be approximately 40 J m-2 comparable to literature
values for cubic zirconia. Similar results to those presented in Fig. 10

were obtained for indentation spalling tests on a series of ZnO/Si
samples, and excellent agreement was obtained between theoretical
values of Ps, (Eq. 14), and experimental measurements indicating
that the methods outlined above have some merit when it comes to
assessing the adhesion of thin films, although further work is needed
before the indentation method can be recommended as a technique
for determining interfacial adhesion.
Figure 10. Plot of trend in delamination radius with indentation load for ZrO2-
Y2O3 coatings on a NiCrAlY superalloy.
Recent work by Loh, et al.,[31] used laser impingement to

cause quasi-heating of the coating, and thereby introduce localized
compression which, added to the prior residual stress, attains suffi-
cient magnitude to initiate spallation. By selecting the power and
duration of the laser pulse, a prescribed compressive stress could be
induced in the coating, and the laser beam could be used to effec-
tively probe the relative spall resistance of brittle coatings. This
approach to the use of a laser to evaluate spallation is very different

from that reported by Vossen [48] wherein the laser was used to
generate a stress wave, and it is stress-wave reflections which cause
spallation. In the laser experiments reported by Loh, et al.,[31] spalla-
tion could not be induced in as-coated specimens of ZrO2-Y 2O3
using an incident power of 20 W and a pulse duration of ≈1 s.
However, spalling could be induced in the presence of an initial
delamination created by indentation. Clearly for spalling to occur
subject to constant laser power and pulse duration conditions, the
delamination must exceed a critical size; in the work reported by
Loh, et al., ac ≈ 250 µm. Above this critical delamination size, the
spall radius exceeds the delamination radius, indicating that crack
propagation preceded spalling, as discussed previously for the theo-
retical consideration of the indentation spalling criterion (see above).
In general, for well adhering coatings, the laser methods are non
destructive, and simulate the spall problems during thermal cycling
and impact damage, that is, they simulate real operating conditions
for coatings.
The Scratch Adhesion Test Method. In the scratch test, a
stylus is drawn over the sample surface under a stepwise or continu-
ously increasing normal force until the coating detaches.[45] In prac-
tice, the film is seldom removed entirely from the channel, so it is
convenient to define a critical load (related to coating adhesion); this
being the load at which the coating is removed in a regular way along
the whole channel length.[58] A range of failure modes has been
identified in the scratch test which do not depend on coating sub-
strate adhesion, but there are those which do, such as the “kidney-
shaped” failures in the track which arise due to detachment ahead of
the moving indenter.[21] Such adhesive failures are often associated
with chipping along the sides of the scratch track.
Coating detachment may be observed in a number of ways: (a)
optical or scanning electron microscopy, (b) acoustic emission,[42]
(c) frictional force measurement.[44][59][60] Acoustic emission and
friction force measurements enable electronic recording of the test
sensitivity of the scratch test method in measuring adhesion of thin

hard (< 1 µm) coatings is enhanced when using frictional measure-
ment.[44] A significant advance in scratch testing has been made by
Kinbara, et al.,[61] who propose the use of a vibrating stylus to
measure both friction coefficient, and adhesive force. This new
approach has been successfully applied to the adhesion determina-
tion of a wide variety of film/substrate combinations including
titanium nitride deposited on a glass substrate.
The initial analysis of the mechanics of the scratch test by
Benjamin and Weaver[40] used the theories developed for fully plas-
tic indentation theory to give an expression for the critical shearing
force for coating removal in terms of the scratch geometry, the
substrate properties, and the frictional force on the stylus.
Eq. (15) F = k AH R 2 − A2
where A = (w/πH)1/2, w is the critical load, R the indenter tip radius,

F the shearing force strength of the coating per unit area, A the radius
of the circle of contact, H the indentation hardness of the substrate
material, and k a constant varying between 0.2 and 1.0.[40][62] (see
Fig. 11) The assumption of fully plastic behavior is, in many cases,
inappropriate and further attempts have been made to modify the
expression to take account of elastic/plastic indentation behavior.[62]
However, there is some evidence that the process of coating removal
is more complex than generally assumed for these calculations being
dependent on the properties of the coating.[63][64] Thus, it seems
likely that this method of calculation is insufficient for most practical
coating systems.[65][66]
Figure 11. Schematic of the scratch adhesion tester and deformation geometry
assumed by Benjamin and Weaver.[40]
More recently, Laugier[67][68] has suggested that the adhesion

behavior can be modeled in terms of the strain energy released
during the removal of a coating. Using a Griffith energy balance
approach,[69] the energy so released is used to provide the surface
energy for a crack to form at the coating/substrate interface, and this
may be determined from the work of adhesion. By balancing the
released elastic energy to the surface energy of the crack, it is
possible to relate the local stress responsible for coating detachment
(σ ) to the work of adhesion (W) by[67][68]
σ 2t
Eq. (16) W AB = 1
2 Ec
where t equals coating thickness. Attempts have been made to

calculate σ expressed as a combination of the applied stresses due
to the sliding indenter and the internal stresses within the coating.
The analysis of Laugier[67][68] calculated the applied stresses from
the elastic equations of Hamilton and Goodman. [70] Clearly
this is insufficient to describe the stresses for materials where some

plasticity has occurred. Burnett, and Rickerby[71] have identified
three contributions to the stresses responsible for coating detach-
ment: (1) an elastic/plastic indentation stress, (2) an internal stress,
and (3) a tangential frictional stress. This analysis has been extended
by Bull, et al.,[21][72][73] where each of these contributions is ex-
pressed in terms of their effect on the measured friction coefficient as
shown by Fig. 12. The advantage of expressing all the stress contri-
butions in terms of their effects on the coefficient of friction is clear
when experimental measurements of the tangential force can be
made during scratching.[59]
Figure 12. The scratch adhesion test represented as the sum of three contribu-
tions; an indentation term, an internal stress term and a frictional term. These
may be represented as three frictional contributions; a ploughing component, an
internal stress component and a shear component. (After Bull, Rickerby, Matthews,
Leyland, and Pace.)
The total friction force is given by
Eq. (17) F = A1 p + A2 τ
where p is the ploughing flow stress, τ the tangential stress and A1

and A2 the cross-sectional area of the track and the contact area 2
respectively (see Fig. 12). Since A2 >> A1 the shear stress is small
compared to p and this dominates the stress responsible for coating
removal, σ. Assuming that for the moving stylus the load is sup-
ported on the front half of the contact, an expression relating the
critical load for coating detachment Lc, to the ploughing stress can be
derived
π 2
Eq. (18) Lc = d p
8 c
where dc is the track width at Lc. Combining Eqs. (16), and (18), an
expression for the work of adhesion in terms of Lc can be obtained.
Kendall[74] presents a similar analysis for elastic deformation during
scratch testing
1
π d c2  2 EW 2
Eq. (19) Lc =  
8  t 
Using this expression values for the work of adhesion for a range of
steel substrates have been calculated,[21] see Table 2. Such values are
somewhat smaller than those determined from fracture toughness
experiments for similar steels (~10 kJ m-2),[75] but are larger than the
surface energies of either coating or substrate material, or the interfa-
cial energy (which are generally in the range 1–10 J m-2 ). The work
of Bull and Rickerby[21] illustrates the point that crack-tip plasticity
plays an important role in scratch testing, but that failure occurs at or
near the weaker coating/substrate interface.
Table 2. Adhesion Parameter for 2 µm TiN Coatings Deposited onto

A Range of Steel Substrates
Substrate Work of Adhesion, w Interface Toughness
(J m-2) KI (MN m-3/2)
Stainless Steel 17 1.75

Quenched and Tempered B01 27 2.2
Quenched B01 37 2.6
M2 125 4.8
The value of the critical load for coating detachment, Lc, is

known to depend on the hardness of the substrate, the coating
thickness, and the level of residual stress within the coating.[45][73][81]
However, this does not generally represent variations in coating/
substrate adhesion but a change in the makeup of the stresses respon-
sible for coating detachment.[73] Typically as the coating thickness
increases or the substrate hardness increases the stress transfer to the
coating/substrate interface is reduced, and Lc increases, whereas as
the levels of residual stress in the film increase there is a reduction in
critical load. These effects are shown in Fig. 13.
Figure 13. Variation in critical load for coating detachment with (a) coating
thickness and (b) residual stress. Also shown in (a) is the effect of increasing
substrate hardness.
In addition to the extrinsic parameters which influence the

scratch test, a number of intrinsic parameters such as indenter tip
radius, indenter wear, etc., have an important bearing on the value of

critical load derived from the test. The more important of these are
given in Table 3, along with a listing of the extrinsic parameters, and
the interested reader is advised to read the excellent review of
Steinmann, et al.,[31] for a fuller discussion of these topics. Clearly,
although the scratch test appears at present to be the only available
practical method to study the adhesion of hard tribological coatings,
it seems that, except in a few cases, the test can only provide
comparative information about the adhesion at the coating/substrate
interface. Although the work of Bull, et al.,[21][72][73] illustrates that
the work of adhesion may be determined in some systems, this work
concentrated on one particular type of failure ahead of the indenter
while in practice several possible failure mechanisms have been
identified,[71] and the modeling approach can only apply to those
mechanisms where coating detachment occurs to relieve the com-
pressive stress ahead of the moving indenter. Therefore, although the
scratch test is potentially a useful technique for monitoring coating
adhesion, care is needed both in the performing of the test, due to the
influence of extrinsic and intrinsic parameters on the value of critical
load, and interpreting the results produced if reliable conclusions are
to be drawn regarding coating/substrate adhesion.
Table 3. Factors Affecting the Critical Load for Coating Detachment

in the Scratch Test Method
Intrinsic Parameters Extrinsic Parameters
Loading rate[59][73][76] Substrate properties[76][77]

Hardness
Scratching speed[59][73][76] Modulus
Thermal expansion coefficient
Indenter tip radius[60][62]–[65][73]
Coating properties[76][78]–[80]
Indenter wear[76] Hardness
Modulus
Machine factors Stress and interfacial properties
Thickness
Friction force and friction coefficient[44][72][76]
Surface condition and testing environment
Commercial scratch testers are basically of two types. In the

automatic scratch test, the normal load is continuously increased
along the length of the scratch track by a spring loading mechanism.
This test is very quick and simple to perform, but has the disadvan-
tage that catastrophic failure occurs at the first sufficiently large
flaw, and thus, the critical load may be an underestimate of the
practical adhesion in any application. Also, the critical load has been
found to be a sensitive function of the machine loading geometry and
this makes comparison between laboratories with various types of
machine quite difficult.[59][73][76] The alternative manual scratch tester,
uses dead weight loading, and hence, requires the performing of
many scratches to assess the critical load, a much more time consum-
ing process. However, it has the advantage that the interfacial flaw
distribution can also be assessed by counting the number of failures
that occur at each load.[21][81] The number of failures is found to
saturate at a certain load (usually just before it becomes possible to
distinguish individual failures in the track), as observed by both
reflected light microscopy, or acoustic emission detection where
each failure generates a small burst of acoustic emission. Thus, a
cumulative failure probability P(L) can be defined as
N ( L ) − N (0 )
Eq. (20) P( L) =
N sat
where N(L) is the number of failures at load L, N(0) is the number of

failures on low loads [usually N(0) = 0] and N is the saturation failure
number. P(L) can be given by the Weibull distribution
m
  L − Lo 
Eq. (21) P(L) = 1 − exp  −  
  La 
where La is the load at which the probability of failure is zero

(generally Lo = 0), La is a constant equal to the load at which there is
a 63.2% cumulative failure probability (which can be used as a
critical load related to the flaw distribution), and m is the Weibull
parameter. As m increases the distribution becomes narrower, and

the critical load becomes, a better measure of coating adhesion.
Details of the analysis are shown in Fig. 14. The advantage of this
analysis is, that it offers the possibility of providing a completely
automated way of determining the critical load for coatings which
produce well defined pulses of acoustic emission associated with
coating detachment failure events. The acoustic emission trace at
each load can be recorded and examined by computer, the number of
failures counted and stored, and the variation with load determined.
From this, a value of La can be calculated which involves no subjec-
tive decisions by the operator, and hence, is a much more reliable
adhesion criterion. However, much more work is needed to deter-
mine the viability of such an approach in a wide range of coating/
substrate systems.
Figure 14. Weibull statistics analysis of coating failures in the scratch test.
In summary, the scratch test is the most generally useful adhe-

sion test for thin hard coatings, though much care is needed in
interpreting its results. In the absence of a better test, it is important
that the reliability of the test is improved by developing a better
understanding of the mechanics of coating detachment, the origins of

the various failure modes, and how the design and use of the scratch
tester influences the magnitude of the critical load. Ultimately, it is
the operator which dictates the usefulness of the technique, and any
developments which reduce the operators subjective judgement can
only improve the results it produces.
4.0 X-RAY CHARACTERIZATION OF THIN FILMS
Internal stress is an important parameter in coating technology,

since it often relates to the maximum thickness which can be depos-
ited without spallation. Methods for measuring such stresses broadly
breakdown into two groups:
1. X-ray techniques,[82] in which an interplanar spacing
is used as an internal strain gauge.
2. Mechanical deflection techniques, such as cantilever
bending,[83] or tube slitting,[84] where the size of any
elastic deflections caused by the presence of residual
stresses is used to determine their magnitude.
Of the two techniques, the first is by far the most useful for thin
films (<10 µm) because, in addition to information on residual stress,
it also provides important structural information. In order to opti-
mize the levels of residual stress within a film, it is vitally important
to understand how this stress arises from the coating microstruc-
ture.[85][86] There are three types of lattice distortions which have
been identified in PVD coatings (see Fig. 15). [88]
(1) Macrostrains that are introduced into the coating
due to the high internal stresses that result from
growth and thermal mismatch factors. This
macrostrain corresponds to an average change of
the lattice parameter and can be derived from the
x-ray diffraction line shift.
(2) Microstrains that are introduced because distortions

such as dislocations, point defects, etc., bring about
local fluctuations of the lattice parameter around
the new average lattice parameter. These can be
determined from x-ray line broadening data.
(3) Microstrains that are caused by yield anisotropy
between grains, which give rise to a peak shift and
line broadening if they vary from point to point in
the coating.
Figure 15. Schematic of the three types of lattice distortions found in PVD
coatings.
All these may be monitored by careful x-ray diffraction experi-

ments.[95] The use of x-ray techniques to characterize coatings is
discussed in more detail in the following sections.
4.1 Measurement of Residual Stress by X-ray Techniques
In PVD titanium nitride coatings, x-ray techniques have been

used not only to measure the level of internal stress in a
coating, but also to characterize the microstructure of the coating.

In such systems, the types of lattice distortion can be related to the
properties of the coating, which in turn can be correlated with its
performance in engineering applications.[85][86] In this chapter, ex-
amples will be given on the measurement of residual stress levels in
PVD coatings, and perhaps more importantly, how these relate to the
microstructure of the coating and its properties.
Internal stress is an important parameter in coating technology
since it often relates to the maximum coating thickness which can be
deposited by PVD techniques without spallation. The importance of
internal stress to coating adhesion is highlighted by the work of
Chollet, et al.,[87] and Laugier.[67] A number of papers have been
published on the measurement, by X-ray techniques, of residual
stresses in PVD coatings such as titanium nitride, and all are based
on the so-called sin2ψ method.[87]–[90]
In this simple analysis, it is assumed that the stress state which
exists within a coating, is one of biaxial stress, which requires that
the measured value of strain is related to the principal stresses σ11
and σ22 according to[91]
Eq. (22) ε φψ = (S 2 2 )(σ 11 cos 2 φ + σ 22 sin 2 φ )sin 2 ψ + S1 (σ 11 + σ 22 )
where S1 and S2/2 are the X-ray elastic constants for a particular (hkl)
reflection. For an elastically isotopic solid then
Eq. (23) S 2 2 (hkl ) = (1 − υ ) E ; S1 (hkl ) = − υ E
where E is Young’s modulus and υ Poisson’s ratio. In the sin2ψ

method an interplanar spacing d serves as an internal strain gauge,
(Fig. 16) and the strain is measured along a certain direction.
Eq. (24) εφψ = (d φψ − d o ) d o

where dφψ is the interplanar spacing of crystal planes which lie

perpendicular to the ψ direction in the stressed condition and do the
interplanar spacing in the unstressed condition.
Figure 16. Schematic diagram showing the use of crystallographic planes to

determine residual stresses by x-ray techniques.
Figure 17 represents the idealized variation of lattice strain

with sin2ψ according to Eq. (22) from which values for the principal
stresses σ 11 and σ 22 can be derived or, as is more commonly
reported, values for σφ since σφ = σ11 cos 2φ + σ22 sin2φ. It can be
seen from Fig. 20 that at constant φ, the slope of the εφψ versus sin2φ
plot is only proportional to σφ when a biaxial stress state is assumed.
According to Eq. (22), a compressive stress acting in the plane of the
coating will give rise to an expansion of planes parallel to the surface
due entirely to the Poisson’s ratio effect which can be represented as:
Eq. (25) ε φψ = (a hkl − a o ) a o = s1 (σ 11 + σ 22 )

and hence
2υ
Eq. (26) a hkl = a o − σ φ ao
E
where ao is the lattice parameter of unstressed titanium nitride.

Figure 18 illustrates the variation of ahkl with internal stress (σφ )
according to Eq. (26).
Figure 17. Sin2 ψ method for stress determination by x-ray diffraction.
Figure 18. Variation of lattice parameter with internal stress for titanium nitride
coatings deposited by a number of physical vapor deposition processes.
In many instances, the true stress state is not one of biaxial

stress, and the idealized behavior depicted in Fig. 20 is not observed.
Several authors report behavior such as curvature and ψ-splitting in
the lattice strain versus sin2ψ plots which are inconsistent with the
biaxial approximation. In these cases, the sin2ψ method may be
extended from the simple biaxial approximation to allow for the
determination of the full stress tensor.[92]
Curvature of the εφψ Versus Sin 2ψ Plot. The approximation
of the true triaxial stress state to one of biaxial stress requires that the
stress components in the direction of the sample normal are negli-
gible in the volume sampled by the x-ray beam and that σ 11 and σ 22
do not vary with depth. This is not always the case since at each ψ-
inclination the average value of internal stress within the irradiated
volume will change as with increasing tilt, the stress averaging
occurs closer to the surface where stress relaxation and larger grains
sizes can lead to lower stress levels. The effect of this is to cause
curvature in the εφψ versus sin2ψ plot as shown in Fig. 19 for a
titanium nitride coating deposited onto a martensitic 420 substrate. It
is interesting to note that the degree of curvature is greater for
measurements made with Cr Κα radiation compared to Cu Κα
radiation since this is the less penetrating of the two x-radiations.
These observations are important since it has been shown by Noyan[93]
that even small curvatures in the sin2ψ plot can lead to an appreciable
error in the calculated stress state when this is approximated to that
of biaxial stress.
Psi- (ψ -) Splitting in the εφψ Versus Sin 2ψ Plot. If, in
addition to the normal stress ( σ 33), there exist stresses (σ 13, σ 23) in
the direction of the surface normal then ψ -splitting is observed in the
sin2ψ plot. As shown in Fig. 19 the plot separates into two branches
for positive and negative tilt with a large curvature near ψ = 0. If
the unstressed lattice parameter is known, then it is possible to
estimate the magnitude of the normal stress averaged over the pen-
etration depth (σ 33), and hence, the likely effects on σ φ.[92][94] Errors
in the value of σ φ can be as high as 50% if this is not taken into
consideration.
Figure 19. An example of curvature and ψ-splitting in the εφψ versus sin 2 ψ plot
for a 12 µm coating of titanium nitride deposited onto martensitic 420
stainless steel. Measurements were made with Cr Kα radiation and the broken
line represents the biaxial stress approximation (σ φ = 2770 MPa): n = ψ < 0;
l = ψ > 0.
4.2 Line Broadening Measurements
From the analysis of x-ray diffraction, line broadening the

contribution due to crystallite size and lattice strain can be deter-
mined.[96] The full width at half maximum height (FWHM) of the
peak can be used as a measure of the broadening if it is corrected for
the instrumental broadening.[95] The broadening due to crystallite
size, β c , is given by
Kλ
Eq. (27) βc =
L cosθ
where L is the grain size, λ the X-ray wavelength, θ the Bragg angle
and K is the shape factor for the crystallites which varies with
reflection (hkl). Similarly the broadening due to lattice strain, βc, is
given by
Eq. (28) β e = 4 e tan θ
where e is the lattice strain. The total broadening is the sum of these
contributions and if the constant K is set to unity we have
cosθ 1 4 e sin θ
Eq. (29) β = +
λ L λ
where β is the measured broadening corrected for instrument effects.

A plot of β cosθ/λ vs. 2sinθ/λ, the so-called “Hall-Williamson
plot” is a straight line with a slope of 2e and an intercept 1/L.[97] This
method for the separation of microstrain and crystallite size depends
on the fact that strain is a function of the order of the reflection, while
crystallite size is not. It is thus standard practice to use two orders of the
same set of crystallographic planes to separate L from e, though for
materials which are close to being isotropic such as titanium nitride
all available (hkl) reflections may be used in Eq. (29). A typical Hall-
Williamson plot for titanium nitride is shown in Fig. 20, comparing
the lattice strain, and hence, detect densities for films produced by
PVD and CVD processes. The strain is greater for PVD than CVD,
and decreases with deposition rate. In this case, the increased peak
broadening for the coating deposited by low rate PVD is due to the
incorporation of argon within the coating during deposition.[98]
The microstructure of the coating can have a marked effect on
the microstrain broadening through its effect on the stress levels
which the coating can bear. Titanium nitride films deposited by CVD
are fully dense (Fig. 21a), but for coatings deposited onto cemented
carbide substrates the thermal mismatch stresses are low, and these
can be accommodated elastically within the coating; low levels of
stress and microstrain broadening are observed for these materials.
In contrast, the PVD films (Fig. 21b) can accommodate the internal
stress elastically at first, but as the total internal stress exceeds the
yield strength of the coating it is accommodated by plastic flow (i.e.,
generation of dislocation loops), with a corresponding increase in
microstrain broadening.[88] The defects created can be clearly seen in
TEM images of PVD titanium nitride films (see Fig. 22).
(a)
(b)
Figure 20. Hall-Williamson plots comparing microstrain broadening for tita-

nium nitride coatings deposited by PVD and CVD processes: (a) reactive ion
plating with deposition rates of 0.54 µm min-1 and 0.05 µm min-1; (b) CVD.
(a)
(b)
Figure 21. Scanning electron micrographs of titanium nitride coatings deposited

by CVD and PVD processes: (a) CVD, (b) reactive ion plating at a deposition rate
of 0.05 µm min-1.
Figure 22. Bright field transmission electron micrographs of a titanium nitride

film removed from a martensitic 420 stainless steel substrate.
In summary, the use of x-ray techniques provides residual

stress and defect information which can enable the performance of
coatings in any end-user application to be better understood.
5.0 SUMMARY AND CONCLUSIONS
The tests described in this paper for measuring the hardness,

adhesion, and internal stress state of thin films are all easy to
perform, but the results obtained require careful interpretation if
reliable conclusions about the properties of the coating in tribologi-
cal applications are to be drawn.
The measurement of the hardness of thin films requires a
modeling approach to be adopted if the coating properties are to be
obtained independent of the substrate. In the absence of a reliable
theoretical model for the plastic deformation associated with the

indentation testing, a modified volume-law-of mixtures offers the
best method of apportioning deformation between the coating and
the substrate. The model is widely applicable but breaks down when
coating fragmentation is well developed. This difficulty, together
with a need to simplify the complex curve-fitting processes neces-
sary to implement the model has led to the development of an
alternative model based on indentation[99] but some work is neces-
sary to determine its validity.
The measurement of coating adhesion using the scratch test
can be performed, if a coating failure mechanism can be identified
that involves detachment at the coating/substrate interface. Though
in many cases such a failure may be identified by acoustic emission,
there are many other coating failure modes which can contribute to
the acoustic emission signal, and thus this technique for detecting the
onset of coating failure is not reliable. It is possible to obtain a work
of adhesion for the coating in some cases from the scratch test but
this is affected by a number of factors specific to the test such as;
loading rate, surface contamination, and stylus wear. Further, there is
a distribution of flaws at the coating/substrate interface which results
in the occurrence of coating detachment and an increasing number of
sites along a scratch track as the load is increased. Such observations
make it difficult to obtain absolute measurements of adhesion but the
test is still valuable for making comparisons between coatings.
The residual stresses in coatings can be determined by x-ray
diffraction using the sin2ψ method. The linear behavior in the εφψ
versus sin2ψ plots predicted on the basis of a biaxial stress state is not
always observed, and plots often show ψ -splitting and curvature due
to the presence of shear stresses and stresses normal to the surface. It
is important to identify these effects, since if they are ignored and the
stress state approximated to biaxial then appreciable errors in the
stress values determined can result. Similarly, the broadening of
individual X-ray peaks can be used to determine the grain size, and
lattice strain within the coating which gives important information
about the defect state of the films.
In order for the properties of a coating to be more fully under-

stood in any application, it is essential that good data on fundamental
properties is available. The techniques described in this chapter go
some way to providing this information, but tests are needed that
measure properties which are more relevant to the intended applica-
tion (such as hardness, adhesion, and residual stress at the operating
temperature), if good predictions about performance are to be made.
REFERENCES
1. Matthews, A., Surface Engineering, 1:93 (1985)

2. Hinterman, H. E., J. Vac. Sci. Technol., 2B:816 (1984)
3. Sargent, P. M., and Page T. F., Proc. Brit. Ceram. Soc., 26:209
(1978)
4. Brookes, C. A., in: Science of Hard Materials, (R. K. Viswanadhan,
D. J. Rowecliffe, and J. Gurland, eds.), pp. 181–198, Plenum, New
York (1983)
7. Newey, D., Wilkins, M. A., and Pollock, H. M., J. Phys. E, 15:119
(1982)
8. Stone, D., LaFontaine, W., Ruoff, S., and Li, C. Y., Mater. Res. Soc.
Symp. Proc., 72:43 (1986)
9. Pethica, J., Hutchings, R., and Oliver, W. C., Phil. Mag., A48:593
(1983)
10. Bull, S. J., Yoffe, E., and Page, T. F., Phil. Mag. Lett., 59:281 (1989)
11. Doerner, M. F., and Nix, W. D., J. Mater. Res., 1:601 (1986);
Oliver, W. C., and Pharr, G. M., J Mater. Res., 7:1564 (1992)
12. Wu, T. W., Burn, R. A., Chen, M. M. , and Alexopoulos, P. S., Mat.
Res. Soc. Symp. Proc., 130:117 (1989)
13. Oliver, W. C., McHargue, C. J., and Zinkle, S. J., Thin Solid Films,
153:185 (1987)
14. O’Hern, M., Oliver, W. C., McHargue, C. J., Rickerby, D. S., and
Bull, S. J., Mat. Res. Soc. Symp. Proc., 188:139–143 (1990)
15. Sargent, P. M., Ph.D., Thesis, University of Cambridge (1979)
16. Burnett, P. J., and Page, T. F., J. Mat. Sci., 19:845–860 (1984)
17. Jonsson, B., and Hogmark, S., Thin Solid Films, 114:257–269
(1984)
18. Johnson, K. L., J. Mech. Phys. Sol., 18:115–126 (1970)
19. Lawn, B. R., Evans, A. G., and Marshall, D. B., J. Am. Ceram. Soc.,
63:574–581 (1980)
20. Chiang, S. S., Marshall, D. B., and Evans, A. G., J. Appl. Phys.,
53:298–311 (1982)
21. Bull, S. J., and Rickerby, D. S., Surf. Coat. Technol., 42:149–164
(1990)
22. Rickerby, D. S., and Burnett, P. J., Thin Solid Films, 157:195–222
(1988)
23. Pulker, K. H., and Berger R., Surf. Technol., 14:25 (1981)
24. Mittal, K. L., Electrocomp. Sci. Technol., 3:21 (1976)
25. Campbell, D. S., in: Handbook of Thin Film Technology, (L. I.
Maissel and R. Glang eds.), McGraw-Hill, New York (1970)
26. Jacobsson, R., Thin Solid Films, 34:191 (1976)
27. Katz, G., Thin Solid Films, 33:99 (1976)
28. Bodö, P., and Sundgren, J. E., J. Appl. Phys., 60:1161 (1986)
29. Jankowski, A. F., Thin Solid Films, 154:183 (1987)
30. Lin, D. S., J. Phys. D: Appl. Phys., 4:1977 (1971)
31. Loh, R. L., Rossington, C., and Evans, A. G., J. Am. Ceram. Soc.,
69:139 (1986)
32. Marshall, D. B., and Evans, A. G., J. Appl. Phys., 56:2632 (1984)
33. Rossington, C., Marshall, D. B., Evans, A. G., and Khuri-Yakub, B.
T., J. Appl. Phys., 56:2639 (1984)
34. Mehrota, P. K., and Quinto, D. T., J. Vac. Sci. Technol. A, 3:2401
(1985)
35. Jindal, P. C., Quinto, D. T., and Wolfe, G. J., Thin Solid Films,
154:361 (1987)
36. Heavens, O. S., J. Phys. Radium, 11:355 (1950)
37. Weaver, C., and Hill, R. M., Phil. Mag., 3:1402 (1958)
40. Benjamin, P., and Weaver, C., Proc. R. Soc. London, Ser. A, 254:177
(1960)
41. Benjamin, P., and Weaver, C., Proc. R. Soc. London, Ser. A, 261:516
(1961)
42. Perry, A. J., Thin Solid Films, 78:77 (1981)
43. Valli, J., Makela, U., and Matthews, A., Surf. Eng., 2:49 (1986)
44. Valli, J., Makela, U., Matthews, A., and Murawa, V., J. Vac. Sci.
Technol. A, 3:2411 (1985)
45. Perry, A. J., Surf. Eng., 3:183 (1986)
46. Stephens, A. W., and Vossen, J. L., J. Vac. Sci. Technol., 13:38
(1976)
47. Cooper, J. A., Dewhurst, R. J., and Palmer, S. B., in: Proc. Ultrasonic
International, Brighton, (July 2–4, 1985), Butterworths, London
(1985)
48. Vossen, J. L., Adhesion Measurement of Thin Films, Thick Films,
and Bulk Coatings, (K. L. Mittal, ed.), in: Special Tech. Publ.,
640:122, ASTM, Philadelphia (1978)
49. Chopra, K. L., Thin Films Phenomena, p. 313, McGraw-Hill, New
York (1969)
50. Chapman, B. N., J. Vac. Sci. Technol., 11:106 (1974)
51. Bullet, T. R., and Prosser, J. L., Proc. Org. Coat., 1:45 (1972)
52. Rickerby, D. S., Surf. Coat. Technol., 36:541–557 (1988)
53. Chiang, S. S., Marshall, D. B., and Evans, A. G., in: Surfaces and
Interfaces in Ceramics and Ceramic-Metal Systems, (J. Pask, and A.
G. Evans, eds.), p. 603, Plenum Press, New York (1981)
54. Lawn, B. R., and Wilshaw, T. R., Fracture of Brittle Solids, p. 53,
Cambridge University Press, Cambridge, England (1975)
55. Evans, A. G., J. Am. Ceram. Soc., 65:127 (1982)
56. Timoshenko, S., and Gere, J. M., Theory of Elastic Stability, p. 390,
McGraw-Hill, New York (1961)
57. Evans, A. G., and Hutchinson, J. W., Int. J. Solid. Struct., 20:455
(1984)
58. Perry, A. J., Thin Solid Films, 107:167 (1983)
59. Valli, J., J. Vac. Sci. Technol. A, 4:3007 (1986)
60. Jacobson, S., Jonsson, B., and Sundquist, B., Thin Solid Films,
107:89 (1983)
61. Kinbara, A., Baba, S., and Kikuchi, A., J. Adhesion Sci. Technol.,
2:1 (1988)
62. Weaver, C., J. Vac. Sci. Technol., 12:18 (1975)
63. Butler, D. W., Stoddard, C. T. H., and Stuart, P. R., J. Phys. D, 3:877
(1970)
64. Je, J. H., Guarmati, E., and Naoumidis, A., Thin Solid Films, 136:57
(1986)
65. Steinmann, P. A., Laeng, P., and Hintermann, H. E., Mater. Tech.,
13:85 (1985)
66. Laeng, P., and Steinmann, P. A., in: Proceedings of the 8th
International Conference on CVD, p. 723, The Electrochemical
Society, Pennington, NJ (1981)
67. Laugier, M. J., Thin Solid Films, 117:243 (1984)
68. Laugier, M. J., J. Mater. Sci., 21:2269 (1986)
69. Griffith, A. A., Philos. Trans. R. Soc. London, 221A:163 (1920)
70. Hamilton, G. M., and Goodman, L. E., J. Appl. Mech., 33:371
(1966)
72. Bull, S. J., Rickerby, D. S., Matthews, A., Leyland, A., and Pace, A.
R., Surf. Coat. Technol., 36:503 (1988)
73. Bull, S. J., Rickerby, D. S., Matthews, A., Pace, A. R., and Leyland,
A., Proc. 1st Int. Conf. Plasma Surface Engineering, p. 1226,
Garmusch-Partenkirchen (September 1989)
74. Kendall, K., J. Phys. D: Appl. Phys., 4:1186 (1971)
75. Wyatt, O. H., and Dew-Hughes, D., Metals Ceramics and Polymers,
p. 411, Cambridge University Press, Cambridge, England (1974)
76. Steinmann, P. A., Tardy, Y., and Hintermann, H. E., Thin Solid
Films, 154:333 (1987)
77. Hammer, B., Perry, A. J., Laeng, P., and Steinmann, P. A., Thin
Solid Films, 96:45 (1982)
79. Lopez, E., Beltzung, F., and Zambelli, G., J. Mat. Sci. Lett., 8:346
(1989)
80. Kingswell, R., Rickerby, D. S., Bull, S. J., and Scott, K. T., Thin
Solid Films, 198:139–148 (1991)
81. Bull, S. J., and Rickerby, D. S., Proc. 3rd Int. Conf. Surface
Modification Technologies, Neuchtel, Switzerland, TMS (1990)
82. Wagner, C. N. J., Boldrick, M. S., and Keller, L., in: Advances in X-
ray Analysis, (C. S. Barrett, J. V. Hilfrich, R. Jenkins, J. C. Russ, J.
W. Richardson , Jr., and P. K. Predecki, eds.), 31:129, Plenum Press,
New York (1988)
83. Doerner, M. F., and Nix, W. D., Crit. Rev. Solid State Mat. Sci.,
14:225 (1988)
84. Rickerby, D. S., Eckold, G., Scott, K. T., and Buckley-Golder, I. M.,
Thin Solid Films, 154:125–141 (1987)
85. Rickerby, D. S., and Burnett, P. J. , Surf. Coat. Technol.,
33:191–211 (1987)
86. Bull, S. J., Rickerby, D. S., Robertson, T., and Hendry, A., Surf.
Coat. Technol., 36:743–754 (1988)
87. Chollett, L., Boving, H., and Hintermann, H. E., J. Mater. Energy
Systems, 6:293–299 (1985)
88. Rickerby, D. S., J. Vac. Sci. Technol. A, 4:2809–2814 (1986)
89. Chollett, L., and Perry, A. J., Thin Solid Films, 123:223–234 (1984)
90. Perry, A. J., and Chollett, L., J. Vac. Sci. Technol. A, 4:2801–2808
(1986)
91. Cohen, J. B., Dôlle, H., and James, M. R., Proc. Symp. on Accuracy
in Powder Diffraction, National Bureau of Standards Special
Publication, 567:453 (1980)
92. Rickerby, D. S., Jones, A. M., and Bellamy, B. A., Surf. Coat.
Technol., 36:661–674 (1988)
93. Noyan, I. C., Metall. Trans., 14A:249 (1983)
94. Dôlle, H., and Hauk, V., Haerterei Tech. Mitt., 31:165 (1976)
95. Rickerby, D. S., Jones, A. M., and Bellamy, B. A., Surf. Coat.
Technol., 37:111 (1989)
96. Klug, H. P., and Alexander, L. E., X-ray Diffraction Procedures for
Polycrystalline and Amorphous Materials, 2nd Ed., pp. 618–708,
Wiley Interscience, New York (1974)
97. Williamson, G. K., and Hall, W. H., Acta. Met., 1:22 (1953)
98. Rickerby, D. S., Jones, A. M., and Perry, A. J., Surf. Coat. Technol.,
36:631 (1988)
99. Korsunsky, A. M., McGurk, M. R., Bull, S. J. and Page, T. F., Surf.
Coat. Technol., 99:171–183 (1998)
Macro- and Micromechanical and Tribological Properties 229
Macro- and
Micromechanical and
Tribological Properties
Bharat Bhushan and Bal K. Gupta
1.0 INTRODUCTION
Mechanical properties of the solid surfaces and thin films are

of interest as the mechanical properties affect friction and wear
performance of interfaces. Among the mechanical properties of
interest, one or more of which can be obtained using commercial and
specialized hardness testers are: elastic-plastic deformation behav-
ior, hardness, Young’s modulus of elasticity, scratch resistance,
film-substrate adhesion, residual stresses, time-dependent creep and
relaxation properties, fracture toughness, and fatigue. For macro
devices, bulk properties are important. However, physical contacts at
sliding interfaces in microelectromechanical systems (MEMS), such
as microsensors, microactuators, micromotors, microgear trains,
microvalves, and magnetic recording heads typically occur at very
low loads, thus, friction and wear of sliding interfaces performance is
primarily controlled by the physical and chemical properties of a few
surface atomic layers.[38][48][56][114][145][275] To protect these devices
against wear and corrosion and to achieve a low coefficient of
friction, ultrathin coatings are sometimes deposited on the moving
components. For understanding and/or estimating the functional
behaviors of mechanical devices of small sizes ranging from a
229
couple of tens of microns to a few millimeters, and involving very

low loads on the orders of a few tens of nanonewtons, to a few
millinewtons, measurements of mechanical and tribological proper-
ties of thin coatings or the near-surface region of a bulk material on
micro- and nanoscales are of crucial importance. In the last decade,
a variety of apparatuses have been developed to measure the me-
chanical and tribological properties on micro- to nanoscales. Some
of these common apparatuses are: depth-sensing mechanical proper-
ties microprobe (Nanoindenter), atomic force microscope (AFM),
and friction force microscope (FFM). These apparatuses operate at
very low loads and yield properties of the near surface region.
In this chapter, we will present an overview of techniques, and
apparatuses used for measurement of mechanical and tribological
properties on the macro- to nanoscale. The diamond and amorphous
carbon coatings used extensively for controlling friction, wear, and
corrosion are chosen to exemplify the mechanical and tribological
measurements made on the micro- and nanoscale.
2.0 MEASUREMENT OF MECHANICAL PROPERTIES
Hardness implies the resistance to local deformation. Most

commonly used hardness measurements are: scratch hardness and
static indentation hardness.[272] Scratch hardness depends on the
ability of one material to scratch another, or to be scratched by
another solid. The solid and thin film surfaces are scratched by a
sharp stylus made of hard material typically diamond, and either the
loads required to scratch or fracture the surface, or delaminate the
film or the normal/tangential load-scratch size relationships are used
as a measure of scratch hardness and/or interfacial adhe-
sion.[2][30][37][56][76][131]–[133][139][147][199][223][224][261][272][273] [287][302]
The methods most widely used in determining the hardness of
materials are (quasi) static indentation methods. Indentation hard-
ness is essentially a measure of their plastic deformation properties
and only to a secondary extent with their elastic properties. In the
indentation methods, a spherical, conical, or pyramidal indenter is

forced into the surface of the material which forms a permanent
(plastic) indentation in the surface of the material to be exam-
ined. The hardness number (GPa or kg/mm2) equivalent to the
average pressure under the indenter, is calculated as the applied
normal load divided by either the curved (surface) area (Brinell,
Rockwell, and Vickers hardness numbers), or the projected
area (Knoop and Berkovich hardness numbers) of the contact be-
tween the indenter and the material being tested, under
load.[15][33][39][46][60][151][204][219][272][273][298]
In a conventional indentation hardness test, the contact area is
determined by measuring the indentation size by a micro-
scope after the sample is unloaded. At least, for metals, there is a
little change in the size of the indentation on unloading, so that the
conventional hardness test is essentially a test of hardness under
load, although it is subjected to some error due to varying elastic
contraction of the indentation. [262] More recently, in depth-
sensing indentation hardness tests, the contact area is determined
by measuring the indentation depth during the loading/unloading
cycle.[38][39][60][65][101][207][211][218][228][230][282][304] Depth measure-
ments have, however, a major weakness arising from “piling-up,”
and “sinking-in” of material around the indentation. The measured
indentation depth needs to be corrected for the depression (or the
hump) of the sample around the indentation, before it can be used for
calculation of the hardness.[99][100][111][210][218][230][304] Young’s modu-
lus of elasticity is the slope of the stress-strain curve in the elastic
regime. It can be obtained from the slope of the unloading
curve.[210][218][230] Hardness data can be obtained from a depth sens-
ing instrument without imaging the indentations with high reproduc-
ibility. This is particularly useful for small indents required for
hardness measurements of extremely thin films.
In addition to measurements of hardness and Young’s modulus
of elasticity, static indentation tests have been used for measure-
ments of a wide variety of material properties such as elastic-plastic
deformation behavior,[100][111][127][130][218][228][263] flow stress,[272]
scratch resistance and film-substrate adhesion, [2][30][37][39][41][50]
[56][57][76][131]–[133][139][147][199][223][224][261][263][272][287][301][302][305][306] re-
sidual stresses,[168][183][270] creep,[19][84][144][178][206][290][297] stress re-

laxation,[85][135][136][166][167][191][207][234][235][302] fracture toughness and
brittleness,[79][172][174][194][220][232] and fatigue.[177][307]
The extended load range of static indentation hardness testing is
shown schematically in Fig. 1. We note that only the lower micro- and
ultra-microhardness, or nanohardness load range can be employed
successfully for measurements of extremely thin (submicron thick)
films. The intrinsic hardness of surface layers or thin films becomes
meaningful only if the influence of the substrate material can be
eliminated. It is therefore generally accepted that the depth of inden-
tation should never exceed 30% of the film thickness.[15] The mini-
mum load for most commercial microindentation testers available is
about 10 mN. Loads on the order of 50 µN to 1 mN are desirable if
the indentation depths are to remain few tens of a nanometer. In this
case, the indentation size sometimes reaches the resolution limit of a
light microscope and it is almost impossible to find such a small
imprint if the measurement is made with a microscope after the
indentation load has been removed. Hence, either the indentation
apparatuses are placed in situ a scanning electron microscope (SEM), or
in situ indentation depth measurements are made. The latter mea-
surements, in addition, would offer the advantages to observe the
penetration process itself. In viscoelastic/viscoplastic materials, since
indentation size changes with time, in situ measurements of the
indentation size is particularly useful, which can, in addition, provide
more complete creep and relaxation data of the materials.
Figure 1. Extended load range of static indentation hardness testing.

In this section, we review various hardness test apparatuses for

measurements of mechanical properties of surface layers of bulk
materials and extremely thin films (submicron in thickness) on
macro- to nanoscale.
2.1 Apparatuses for Hardness Measurement on Macroscale
Macrohardness measurements are generally used to determine

bulk hardness. These include Brinell and Rockwell hardness
testers.[46][272]
Brinell hardness is determined by forcing a hardened sphere
under a known load into the surface of a material, and measuring the
diameter of the indentation left after the test. The load is maintained
for about 30 seconds, and the diameter of indentation at the surface is
measured with an optical microscope (10×) after the ball has been
removed. The Brinell hardness number (BHN) or simply the Brinell
number, is defined in kilograms per square millimeter, as the ratio of
the load used to the actual surface area of the indentation, which is, in
turn, given in terms of the imposed load W, ball diameter D, and the
indentation diameter d or depth of indentation h:
2W W
BHN = =
Eq. (1) π D( D − D 2 − d 2 ) π Dh
Hardened steel bearing balls may be used up to 450 BHN, but

beyond this hardness, especially treated steel or tungsten carbide
balls should be used to avoid flattening of the indenter. The standard
size ball is 10 mm, and the standard loads are 3000, 1500, and 500
kg, with 250, 125, and 100 kg sometimes used for softer materials,
(for more details see Method of Brinell Hardness Testing, ASTM
E10–61T). The load and the ball diameter should be adjusted to keep
the ratio d/D within the range of 0.3 to 0.5. The nearest edge of the
specimen should be no closer than 2.5 impression diameters, and the
thickness should be more than one diameter. Since d/D is normally
less than 0.5, this means that for a 10-mm ball, the uninterrupted
width and depth of the specimen may have to be as great as 25 and 10

mm, respectively, to avoid spurious side and bottom effects.
Use of the curved area in the Brinell test was originally intro-
duced to try to compensate for the effects of work-hardening. It is
cumbersome to use the surface area of indentation, though, so to
overcome this disadvantage, Meyer hardness is defined in kilograms
per square millimeter, as the load used divided by the projected area of
indentation:
4W
Eq. (2) Meyer hardness =
π d2
In the Rockwell Hardness test, the indenter may be either a

steel ball of some specified diameter or a spherical-tipped conical
diamond of 120° angle with a 0.2-mm tip radius, called a Brale. A
minor load (or perload) of 10 kg is first applied, and this causes an
initial penetration that holds the indenter in place. At this point, the
dial is set to zero, and the major load is applied. The standard loads
are 150, 100, and 60 kg. Upon removal of the major loads, the
reading is taken while the minor load is still on. The hardness number
may then be read directly from the scale that measures penetration.
The Rockwell hardness number is defined by an arbitrary equation of
the following form:
Eq. (3) R = C1 − C 2 ∆t
where C1 and C2 are constants for a given indenter size, shape, and
hardness scale, and ∆ t is the penetration depth in millimeters be-
tween the major and minor loads. Although Rockwell hardness
increases with Brinell hardness, the two are not proportional and the
dimensions of the Rockwell hardness are not force per unit area. In
fact, the Rockwell hardness number cannot be assigned any dimen-
sions, since it is defined in an arbitrary Eq. (3).
A variety of combinations of indenters and major loads are
possible; the most commonly used are HRB or R B, which uses
1.59-mm diameter (1/16 inch) ball as the indenter and a major load
of 100 kg; HRC, or RC, which uses a cone as the indenter and a major
load of 150 kg; and HRA or RA, which uses a cone as the indenter
and a major load of 60 kg. Rockwell B is used for soft metals and
Rockwell C and A are used for hard metals.
2.2 Apparatuses for Hardness Measurement on Microscale
Microhardness measurements allow the indenter to be shallow

and of small volume so as to measure the hardness of brittle materials
or thin materials or coatings. The apparatuses include Vickers and
Knoop hardness testers.[272] Both testers use highly polished dia-
mond pyramidal indenters. The Vickers indenter is a diamond in the
form of a square pyramid with face angles of 136° (corresponding to
edge angles of 148.1°) (Fig. 2a), and relatively low loads varying
between 1 and 120 kg are used. The Knoop indenter is a rhombic-
based pyramidal diamond with longitudinal edge angles of 172.5°
and 130° (Fig. 2b). In general, the loads used in Knoop testers vary
from about 0.2 to 4 kg. Smaller loads (as low as 1 to 25 g) also may
be used in Vickers and Knoop hardness measurements, the lengths of
the diagonals of the indentation are measured using a medium
powered microscope after the load is removed. If d is the mean value
of a diagonal in millimeters and W is the imposed load in kilogram,
Vicker hardness number (V or HV), sometimes called diamond-
pyramid hardness (DPH), is given by the load divided by the actual
surface area, that is,
1.8544W
Eq. (4) HV =
d2
If l is the long diagonal in millimeters, Knoop hardness number

(K or HK) is given by the load divided by the projected area of the
indentation, that is,
14.229W
Eq. (5) HK =
l2
Figure 2. Geometry and indentation with (a) a Vickers indenter and (b) a Knoop
indenter.
The depth to the diagonal for the Vickers indenter is about 7

and depth to the long diagonal (used for the Knoop indenter) is about
30.5. The advantage of the Knoop indenter over the Vickers indenter
in microhardness testing, lie in the fact that a longer diagonal is
obtained for a given depth of indentation or a given volume of
material deformed. The Knoop indenter is thus advantageous when
shallow specimen or thin, hard layers must be tested. The Knoop
indenter is also desirable for brittle materials (such as glass or dia-
mond), in which the tendency for fracture is related to the area of
stressed material.
2.3 Apparatuses for Hardness Measurement on Nanoscale
In this section, we review nanoindentation hardness appara-

tuses in which the indent is imaged after the load has been removed,
as well as the depth-sensing indentation apparatuses in which the
load-indentation depth are continuously monitored during the load-
ing and unloading processes. Earlier work by Alekhin, et al.,[6]
Ternovskii, et al.,[279] and Bulychev, et al.,[72] led to the development
of depth-sensing apparatuses. Prototype depth-sensing hardness ap-
paratuses developed by several research groups are reviewed by
Bhushan.[38][39] A commercial depth-sensing nanoindentation hard-
ness test apparatus manufactured by Nano Instruments Inc., is exten-
sively used and is described in detail.
Nanoindentation Hardness Apparatuses With Imaging of
Indents After Unloading. For completeness, we first describe a
commercially available microindentation hardness apparatus (Model
No. Micro-Duromet 4000), that uses a built-in light optical micro-
scope for imaging of indents after the sample is unloaded. It is
manufactured by C. Reichert Optische Werke AG, A-1171, Vienna,
Box 95, Austria.[233] The indenter’s case is of the size of a micro-
scope objective mounted on the objective revolver. The load range for
this design is from 0.5 mN to 2 N; therefore it is used for thicker films.
A commercial nanoindentation hardness apparatus for use in-
side a scanning electron microscope (SEM), (Model No. UHMT-3)
for imaging the indents after the sample is unloaded, is manufactured
by Anton Paar K.G., A-8054, Graz, Austria. The apparatus is mounted
on the goniometer stage of the SEM. In this setup, the indenter is
mounted on a double-leaf spring cantilever, and is moved against the
sample by an electromagnetic system to attain the required indenta-
tion load, which is measured by strain gages mounted on the leaf
springs.[22][23] After the required load has been reached and the dwell
time has elapsed, the sample is unloaded, and the indentation diago-
nal is measured by an SEM.
Depth-Sensing Nanoindentation Hardness Apparatus and
Its Modifications. General Description and Principle of Operation.
The most commonly used commercial depth-sensing nanoindentation
hardness apparatus is manufactured by Nano Instruments Inc., 1001

Larson Drive, Oak Ridge, Tennessee 37830. Ongoing development
of this apparatus have been described by Pethica, et al.,[228] Oliver, et
al.,[216] Oliver and Pethica,[217] Oliver and Pharr,[218] and Pharr and
Oliver.[230] This instrument is called Nanoindenter. The most recent
model is Nanoindenter II.[16] The apparatus continuously monitors
the load, and the position of the indenter relative to the surface of the
specimen (depth of an indent), during the indentation process. The
area of the indent is then calculated from a knowledge of the geom-
etry of the tip of the diamond indenter. The load resolution is about
±75 nN, and position of the indenter can be determined to ±0.1 nm.
Mechanical properties measurements can be made at a minimum
penetration depth of about 20 nm (or a plastic depth of about 15
nm).[216] Description of the instrument which follows, is based on
Ref. 16.
The Nanoindenter consists of three major components: (1) the
indenter head, (2) an optical microscope, and (3) a X-Y-Z motorized
precision table for positioning and translating the sample between
the optical microscope and indenter, Fig. 3a. The loading system
used to apply the load to the indenter consists of a magnet and coil in
the indenter head, and a high precision current source, Fig. 3b. The
displacement sensing system consists of a special three-plate capaci-
tive displacement sensor, used to measure the position of the in-
denter, Fig. 3b. The indenter column is attached to the moving plate.
This plate-and-indenter assembly is supported by two leaf springs
cut in such a fashion as to have very low stiffness. At the bottom of
the indenter rod, a three-sided pyramidal diamond tip (Berkovich
indenter, discussed next) is generally attached. The indenter head
assembly is rigidly attached to the “U” beam, and sample rides on the
X-Y-Z table, Fig. 3a. The optical microscope is also attached to the U
beam. The position of an indent on a specimen is selected using the
microscope (max. magnification of 1500×). The spatial resolution of
the position of the table in the X-Y plane is ±400 nm.
(a)
(b)
Figure 3. Schematics of the Nanoindenter II (a) showing the major components-

the indenter head, an optical microscope and a X-Y-Z motorized precision table
and (b) showing the details of indenter head and controls (microscope which is
directly behind indenter and massive U bar are not shown for clarity).[16]
The Nanoindenter also comes with a continuous stiffness mea-

surement device.[217][229] This device makes possible the continuous
measurement of the stiffness of a sample, which allows the elastic
modulus to be calculated, as a continuous function of time (or
indentation depth). Useful data can be obtained from indents with
depths as small as 20 nm. Because of the relatively small time
constant of the measurements, the device is particularly useful in
studies of time-dependent properties of materials.
Weihs, et al.,[295] used acoustic emission (AE) sensors to detect
cracking during indentation tests using Nanoindenter. Acoustic emis-
sion (AE), measurement is a very sensitive technique to monitor
cracking of the surfaces and subsurfaces. The nucleation and growth
of cracks result in a sudden release of energy within a solid, then
some of the energy is dissipated in the form of elastic waves. These
waves are generated by sudden changes in stress, and in displace-
ment that accompany the deformation. If the release of energy is
sufficiently large and rapid, then elastic waves in the ultrasonic
frequency regime (acoustic emission) will be generated, and these
can be detected using piezoelectric transducers (PZTs), via expan-
sion and compression of the PZT crystals.[38][249][312] The energy
dissipated during crack growth can be estimated by the rise time of
the AE signal. Weihs, et al.,[295] mounted commercial transducer
with W impregnated epoxy backing for damping underneath the
sample. The transducer converts the AE signal into voltage, that is
amplified by an oscilloscope, and used for continuous display of AE
signal. Any correlation between the AE signal, and the load-dis-
placement curves can be observed (also see Refs. 302 and 306).
The Berkovich Indenter. The main requirements for the in-
denter are high elastic modulus, no plastic deformation, low friction,
smooth surface, and a well defined geometry that is capable of
making a well defined indentation impression. The first four require-
ments are satisfied by choosing the diamond material for the tip. A
well defined perfect tip shape is difficult to achieve. Berkovich is a
three-sided pyramid, and provides a sharply pointed tip compared to
the Vickers or Knoop indenters, which are four-sided pyramids and
have a slight offset (0.5–1 µm).[38][39][273] Because any three nonpar-

allel planes intersect at a single point, it is relatively easy to grind a
sharp tip on an indenter if Berkovich geometry is used. However, an
indenter with a sharp tip suffers from a finite, but an exceptionally
difficult to measure tip bluntness. Experimental procedures have
been developed to correct for the tip shape, to be described later.
Berkovich indenter is a three-sided (triangular-based), pyrami-
dal diamond, with a nominal angle of 65.3° between the (side) face,
and the normal to the base at apex, an angle of 76.9° between edge
and normal, and with a radius of the tip less than 0.1 µm (Fig. 4a).[33]
Typical indenter is shaped to be used for indentation (penetration)
depths of 10–20 µm. The indents appear as equilateral triangles (Fig.
4b), and the height of triangular indent l is related to the depth h as
h  2 1
Eq. (6a) =   cot 76.9 =
l  3 6.44
(a)
(b)
Figure 4. (a) Schematic of a Berkovich tip and (b) impression of a indentation

made by Berkovich tip.
The relationship h(l ) is dependent on the shape of the indenter. The

height of the triangular indent l is related to the length of one side of
the triangle a as
Eq. (6b) l = 0.866a
and
h 1
Eq. (6c) =
a 7.44
The projected contact area (A) for the assumed geometry is given as
Eq. (7) A = 0.433a2 = 23.97h2
The exact shape of the indenter tip needs to be measured for

determination of hardness, and Young’s modulus of elasticity. Since
the indenter is quite blunt, direct imaging of indentations of small
size in the scanning electron microscope is difficult. Determination
of tip area function will be discussed later.
Analysis of Indentation Data. An indentation curve is the
relationship between load W and displacement (or indentation depth or
penetration depth) h, which is continuously monitored and recorded
during indentation. Stress-strain curves, typical indentation curves,
the deformed surfaces after tip removal, and residual impressions of
indentation for ideal elastic, rigid-perfectly plastic, and elastic-per-
fectly plastic, and real elastic-plastic solids are shown in Fig. 5. For
an elastic solid, the sample deforms elastically according to Young’s
modulus, and the deformation is recovered during unloading. As a
result, there is no impression of the indentation after unloading. For
a rigid-perfectly plastic solid, no deformation occurs until yield
stress is reached, when plastic flow takes place. There is no recovery
during unloading, and the impression remains unchanged. In the case
of elastic-plastic solid, it deforms elastically according to Young’s
modulus, and then it deforms plastically. The elastic deformation is
recovered during unloading. In the case of elastic-perfectly plastic
solid, there is no work hardening.
Figure 5. Schematics of stress-strain curves, typical indentation curves, de-

formed surfaces after tip removal, and residual impressions of indentation, for
ideal elastic, rigid-perfectly plastic, elastic-perfectly plastic (ideal) and real
elastic-plastic solids.
All engineering surfaces follow real elastic-plastic deforma-

tion behavior with work hardening.[151] The deformation pattern of a
real elastic-plastic sample during and after indentation is shown
schematically in Fig. 6. In this figure, we have defined the contact
depth (hc) as the depth of indenter in contact with the sample under
load. The depth measured during the indentation (h) includes the
depression of the sample around the indentation in addition to the
contact depth. The depression of the sample around the indentation
(hs = h-hc) is caused by elastic displacements and must be subtracted
from the data to obtain the actual depth of indentation or actual
hardness. At peak load, the load and displacement are Wmax and hmax,
respectively, and the radius of the contact circle is a. Upon unload-
ing, the elastic displacements in the contact region are recovered and
when the indenter is fully withdrawn, the final depth of the residual
hardness impression is h f.
Figure 6. Schematic representation of the indenting process illustrating the

depression of the sample around the indentation and the decrease in indentation
depth upon unloading.[218]
Schematic of a load displacement curve is shown in Fig. 7.

Based on the work of Sneddon[255] to predict the deflection of the
surface at the contact perimeter for a conical indenter and a paraboloid
of revolution, Oliver and Pharr[218] developed an expression for hc at
the maximum load (required for hardness calculation) from hmax,
Eq. (8a) hc = hmax − ε W max / S max
where ε =0.72 for the conical indenter, ε =0.75 for the paraboloid of
revolution, and ε =1 for the flat punch, and Smax is the stiffness (=1/
compliance), equal to the slope of unloading curve (dW/dh) at the
maximum load. Oliver and Pharr[218] assumed that behavior of
Berkovich indenter is similar to that of conical indenter, since cross-
sectional areas of both types of indenters varies as the square of the
contact depth, and their geometries are singular at the tip. Therefore,
for Berkovich indenter, ε ~ 0.72. Thus hc is slightly larger than
plastic indentation depth (hp) which is given by
Eq. (8b) h p = hmax − Wmax / S max

We note that Doerner and Nix[100] had underestimated hc by assum-

ing that hc = hp. Based on the finite element analysis of the indenta-
tion process, Laursen and Simo[171] showed that hc cannot be as-
sumed equal to hp for indenters which do not have flat punch
geometry.
Figure 7. Schematic of load-displacement curve.
For a Vickers indenter with ideal pyramidal geometry (ideally

sharp tip), projected contact area to depth relationship is given
as[38][100]
Eq. (9a) A = 24 .5hc2
Since the area to depth relationship is equivalent for both typical

Berkovich and Vickers pyramids, Eq. (9) holds for Berkovich in-
denter as well. Though we have derived a slightly different expres-
sion for A(h) presented in Eq. (7) for the assumed Berkovich indenter
geometry, we use Eq. (9) for A(h) in this chapter, as this relationship
is most commonly used in the analysis of the indentation hardness

data.
As shown in Fig. 6, the actual indentation depth, hc, produces
a larger contact area than would be expected for an indenter with an
ideal shape. For the real indenter used in the measurements, the
nominal shape is characterized by an area function F(hc) which
relates projected contact area of the indenter to the contact depth
(Eq. 9a),
Eq. (9b) A
1/ 2
= F ( hc )
The functional form must be established experimentally prior to the

analysis.
Determination of Load Frame Compliance and Indenter
Area Function. As stated earlier, measured displacements are the
sum of the indentation depths in the specimen, and the displacements
of suspending springs, and the displacements associated with the
measuring instruments, referred to as load frame compliance. There-
fore, to accurately determine the specimen depth, load frame compli-
ance must be known. The exact shape of the diamond indenter tip
needs to be measured because hardness and Young’s modulus of
elasticity depend on the contact areas derived from measured depths.
The tip gets blunt and its shape significantly affects the prediction of
mechanical properties (Figs. 8–10).
The method used in the past for determination of area function
has been to make a series of indentations at various depths in
materials in which the indenter displacement is predominantly plas-
tic, and measure the size of the indentations by direct imaging.[100][228]
Optical imaging cannot be used to accurately measure submicron
size impressions. Because of the shallowness of the indent impres-
sions, SEM images result in poor contrast.
(a)
(b)
Figure 8. (a) Predicted projected contact area as a function of indentation depth

curves for various tip radii and measured data, (b) predicted load as a function of
indentation depth curves for various tip radii and measured data.[254]
Oliver and Pharr [218] proposed an easier method for determin-

ing area functions that requires no imaging. Their method is based
only on one assumption, that Young’s modulus is independent of
indentation depth. They also proposed a method to determine load-
frame compliance. We first describe the methods for determining of
load frame compliance followed by the method for area function.
They modeled the load frame and the specimen as two springs in
series, thus
Eq. (10) C = C s + Cf
where C, Cs, and Cf are the total measured compliance, specimen
compliance and load frame compliance, respectively. The measured
compliance C is given by
Eq. (11a) C = dh/dW
The relationship for the sample compliance Cs (inverse of
stiffness S) for an (Vickers, Knoop, and Berkovich) indenter is given as
π 
1/ 2
1
Eq. (11b) Cs =  
2 Er  A
where
1 1 − vs 1 − vi
2 2
= +
Er Es Ei
and dW/dh is the slope of the unloading curve at the maximum load
(Fig. 7), Er, Es, and Ei are the reduced modulus and elastic moduli of
the specimen and the indenter, and νs and νi are the Poisson’s ratios
of the specimen and indenter. C (or S) is the experimentally mea-
sured compliance (or stiffness) at the maximum load during unload-
ing and A is the projected contact area at the maximum load.
From Eqs. (10) and (11), we get
1/ 2
1 ð 
Eq. (12) C = Cf +  
2 Er  A 
From Eq. (12), we note that if the modulus of elasticity is constant, a

plot of C as a function of A1/2 is linear and the vertical intercept gives
Cf. It is obvious that most accurate values of Cf are obtained when
the specimen compliance is small, i.e., for large indentations.
Using the measured Cf value, they calculated contact areas for
indentations made at shallow depths on the aluminum with measured
Er, and/or on a harder fused silica surface with published values of
Er, by rewriting Eq. (12) as
π 1 1
Eq. (13) A=
4 E r (C − C f ) 2
2
from which an initial guess at the area function was made by fitting A
as a function hc data to an eighth order polynomial
Eq. (14) A = 24.5hc2 + C1 hc + C 2 hc1 / 2 + C 3 h c1 / 4 + ⋅ ⋅ ⋅ ⋅ + C 8 hc1 / 128
where C1 through C8 are constants. The first term describes the

perfect shape of the indenter; the others describe deviations from the
Berkovich geometry due to blunting of the tip.
2.4 Mechanical Properties by Nanoindentation
Hardness And Modulus Of Elasticity. Berkovich hardness

HB (or HB ) is defined as the load divided by the projected area of the
indentation. It is the mean pressure which a material will support
under load. From the indentation curve, we can obtain hardness at the
maximum load as,
Eq. (15) HB = Wmax / A
where Wmax is the maximum indentation load and A is the projected

contact area at the peak load. The contact area at the peak load is
determined by the geometry of the indenter and the corresponding
contact depth hc using Eq. (3a) and (4b). Plot of hardness as a
function of indentation depth for polished single-crystal silicon (111),
with and without tip shape calibration, is shown in Fig. 9. We note

that, for this example, tip shape calibration is necessary and the
hardness is independent of corrected depth.
Figure 9. Hardness as a function of indentation depth for polished single-crystal

silicon (111) calculated from the area function with and without tip shape
calibration.[100]
Figure 10. Compliance as a function of inverse of indentation depth for tungsten

with and without tip shape calibration. A constant modulus with 1/depth would
be indicated by the straight line. The slope of the corrected curve is 480 GPa,
which compares reasonably well to the known modulus of tungsten (420 GPa).
The small y-intercept of about 0.3 nm/mN is attributed to load frame compliance,
not removed.[100]
It should be pointed out that hardness measured using this

definition may be different from that obtained from the more con-
ventional definition, in which the area is determined by direct mea-
surement of the size of the residual hardness impression. The reason
for the difference is that, in some materials, a small portion of the
contact area under load is not plastically deformed, and as a result,
the contact area measured by observation of the residual hardness
impression may be less than that at peak load. However, for most
materials, measurements using two techniques give similar results.
Even though during loading, a sample undergoes elastic-plas-
tic deformation, the initial unloading is an elastic event. Therefore,
the Young’s modulus of elasticity or simply called elastic modulus
of the specimen can be inferred from the initial slope of the unload-
ing curve (dW/dh) called stiffness (1/compliance) (at the maximum
load) (Fig. 7). The modulus of elasticity is calculated from Eq. (12).
It should be noted that the contact stiffness is measured only at the
maximum load and no restrictions are placed on the unloading data
being linear during any portion of the unloading.
Hardness And Modulus of Elasticity of Thin Films From
The Composite Response of Film and Substrate. As mentioned
previously, for a thin film on a substrate, if the indentation exceeds
about 30% of the film thickness, then measured hardness is affected
by the substrate properties. A number of researchers have attempted
to derive expressions that relate thin-film hardness to substrate
hardness, composite hardness (measured on the coated substrate),
and film thickness, and so allow the calculations of these quantities
give the remaining three.[34][35][69][70][74][152][244] Here we discuss
two models based on volume law of mixtures, (volume fraction
model) and finite element simulation.
Sargent[244] suggested that the hardness of a film/substrate
composite is determined by a weighted average of the volume of
plastically deformed material in the film (Vf), and that in the sub-
strate (Vs),
Vf Vs
Eq. (16) H =Hf + Hs
V V
where V = Vf + Vs. The deformed volumes of film and substrate can

be calculated using expanding spherical cavity model.[151] Burnett
and Rickerby[73][74] found it necessary to incorporate a further weigh-
ing factor to deforming volume, to obtain a reasonable fit to experi-
mental data. This factor accounts for the differences in relative sizes
of the plastic zones in the film and substrate. Equation (16) is
modified as for a soft film on the hard substrate,
Vf Vs
Eq. (17a) H =Hf X3 + Hs
V V
and for the hard film on the soft substrate,
Vf Vs
Eq. (17b) H = Hs + Hs X 3
V V
where X is the ratio of plastic zone volumes given as,
n
 Ef Hs 
Eq. (17c) X =  

 H f Es 
They found that n, determined empirically, ranged from 1/2 to 1/3.

Bhattacharya and Nix[34] modeled the indentation process us-
ing finite element method to study the elastic-plastic response of
materials. Bhattacharya and Nix[35] calculated elastic and plastic
deformation associated with submicron indentation by a conical
indenter of thin films on substrates, using the finite-element method.
The effects of the elastic and plastic properties of both the film and
substrate on the hardness of the film/substrate composite were stud-
ied by determining the average pressure under the indenter as a
function of the indentation depth. They developed empirical equa-
tions for film/substrate combinations for which the substrate is either
harder or softer than the film. For the case of a soft film on a harder
substrate, the effect of substrate on film hardness can be described as
 σ f  
  σ  2
  s   hc 
= 1 +  f − 1 exp  − 
H H
Eq. (18a)  t  
Hs  Hs   Ef   f  
  E s  
where Ef and Es are the Young’s moduli, σf and σs are the yield
strengths and Hf and Hs are the hardnesses of the film and substrate,
respectively. H is the hardness of the composite, hc is the contact
indentation depth, and tf is the film thickness. Similarly for the case
of a hard film on a softer substrate, the hardness can be expressed as
 Hf  
   
 H   H s  hc 
− 1 exp −
H
= 1 +  1  
f
Eq. (18b) Hs  Hs   σ  E   tf
2

  f σ  f  
  s  Es  
Composite hardness results were found to depend only very

weakly on Poisson’s ratio (v), and for this reason, this factor was not
considered in the analysis. In Fig. 11, they show the composite
hardness results as a function of (hc/tf), for cases in which the film
and substrate have different yield strengths. We note that hardness is
independent of the substrate for indentation depths less than about
0.3 of the film thickness, after which the hardness slowly increases/
decreases because of the presence of the substrate. In Fig. 12, they
show the composite hardness results for cases in which the film and
substrate have different Young’s moduli. It is observed that the
variation of hardness with depth of indentation in these cases is
qualitatively similar to cases in which the film and substrate have
different yield strengths, although the hardness changes more gradu-
ally than in the previous cases. Burnett and Rickerby[75] and Fabes et
al.[111] have applied Eqs. (19)–(21) to calculate the film hardness
from the measured data for various films and substrates.
(a)
(b)
Figure 11. Effect of relative yield strengths of the film and the substrate on the
composite hardness for (a) a soft film on a hard substrate and (b) a hard film on a
soft substrate.[35]
(a)
(b)
Figure 12. Effect of relative Young’s moduli of the film and the substrate on the
composite hardness for (a) a soft film on a hard substrate and (b) a hard film on a
soft substrate.[35]
Doerner and Nix[100] empirically modeled the influence of the

substrate on the elastic measurement of very thin film in an indenta-
tion test using the following expression for the compliance
Eq. (19)
1 − v 2f   − αt f   
1  1 − exp 
 A    
dh 1  π  2  E f     
C= =    +b
dW 2  A   1 − v f 2
 − αt f  1 − v i 2
+ E exp  +
 E i 
 s  A 
where the subscripts f, s, and i refer to the film, substrate, and

indenter, respectively. The term A is equal to (24.5)1/2hc for the
Vickers or Berkovich indenter. The film thickness is tf, and b is the y
intercept for the compliance versus 1/depth plot, obtained for the
bulk substrate, which can be neglected in most cases. The weighing
factors [1-exp(-αtf / A )] and exp (-αtf / A )] have been added to
account for the changing contributions of the substrate and film to
the compliance. The factor α can be determined empirically.
King’s analysis[157] verified that Eq. (19) is an excellent func-
tional form for describing the influence of the substrate, and theoreti-
cally determined the values of α for various indenter shapes. The
value of α was found to depend on the indenter shape and size and
film thickness and was found to be independent of Ei/Es. The values
of α as a function of A /tf for Berkovich (triangular) indenters are
shown in Fig. 13. The values of α are found to be similar for square
and triangular indenters. Bhattacharya and Nix[35] analyzed the de-
formations of a layered medium in contact with a conical indenter
using the finite-element method. Their analysis also verified the
relationship given in Eq. (19).
Figure 13. Parameter α as a function of normalized indenter size for Berkovich

indenter indenting a layered solid surface.[157]
Viscoelastic/Plastic Properties. Most materials including ce-

ramics, and even diamond are found to creep at temperatures well
below half their melting points, even at room temperature. Indenta-
tion creep and indentation load relaxation (ILR) tests are used for
measurement of the time-dependent flow of materials. These offer an
advantage of being able to probe the deformation properties of a thin
film as a function of indentation depth and location.
In the indentation creep test, the hardness indenter maintains its
load over a period of time under well controlled conditions, and
changes in indentation size are monitored.[19][144][178][206][290][297]
Nanoindenters are also used for indentation creep stud-
ies.[178][182][234][235] Indentation creep is influenced by a large num-
ber of variables, such as the material’s plastic deformation proper-
ties, diffusion constants, normal load of indenter, duration of the
indentation, and the test temperature.
In a typical ILR test, the indenter is first pushed into the sample
at a fixed displacement rate until a predetermined load or displace-
ment is achieved, and the position of the indenter is then fixed. The
material below the indenter is elastically supported, and will con-
tinue to deform in an nonelastic manner, thereby tending to push the
indenter farther into the sample. Load relaxation is achieved by
conversion of elastic strain in the sample into inelastic strain in the
sample. During the test, the load and position of the indenter, and the
specimen are continuously monitored. Normally the indenter motion
is held constant, and the changes in the load are monitored as a

function of time. It is possible to obtain the plastic indentation rate
from the indentation load and total depth information during the
relaxation run.[85][136][207] The resulting load relaxation data are
reported in the form of log (indentation pressure) as a function of log
(plastic indentation strain rate).[135][166][167][302]
The indentation pressure is calculated by dividing the load by
the projected area of the indenter. Once the plastic indentation depth
is known as a function of time, the projected area is determined
experimentally as described earlier. The plastic indentation strain
rate [(1/h)(dh/dt), where h is the current indentation depth] is calcu-
lated in the manner similar to that for bulk relaxation data.
The strain rate sensitivity of materials is measured in terms of
stress exponent, n which is defined by the equation,
Eq. (20) Plastic indentation rate = A (indentation pressure) n
where A and n are the constants. The stress exponent is found as a

slope of a log-log plot of plastic indentation rate (or strain rate) and
indentation pressure. In the ILR test, the continuous change in the
contact area results in continuous changes in both plastic indentation
rate and pressure. Thus, data from a single indentation test, which may
span several orders of magnitude in both strain rate and pressure, are
sufficient to determine the stress exponent. Stress exponent can be used
to define the superplasticity of a material. The variations in stress
exponent reflect the changes that may take place when the substructure
generated at high strain rate approaches equilibrium condition.[190]
Nanofracture Toughness. Fracture toughness, KIc of a mate-
rial is a measure of its resistance to the propagation of cracks, and the
ratio H/KIc is an index of brittleness, where H is the hardness.
Indentation fracture toughness is a simple technique for determina-
tion of fracture toughness.[18][79][81][82][88][134][172]–[174][220][232] In-
dentation cracking method is especially useful for measurement of
fracture toughness of thin films, or small volumes. This method is
quite different from conventional methods, in that no special
specimen geometry is required. Rather, the method relies on the fact

that when indented with a sharp indenter, most brittle materials form
radial cracks, and the lengths of the surface traces of the radial cracks
(for definition of crack length, see Fig. 14) have been found to
correlate reasonably well with fracture toughness. Using simple
empirical equations, fracture toughness can then be determined from
simple measurement of crack length.
Figure 14. Schematic of Vickers indentation with radial cracks.
In microindentation, cracks at relatively high indentation loads

of several hundred grams are on the order of 100 µm in length, and
can be measured optically. However, to measure toughness of very
thin films or small volumes, much smaller indentations are required.
However, a problem exists in extending the method to nanoindentation
regime in that there are well defined loads, called cracking thresh-
olds, below which indentation cracking does not occur in most brittle
materials.[169] For a Vickers indenter, cracking thresholds in most
ceramics are about 25 g. Pharr, et al.[232] have found that Berkovich
indenter, (a three-sided pyramid) with the same depth-to-area ratio as
a Vickers indenter, (a four-sided pyramid) has a cracking of the
thresholds very similar to that of the Vickers indenter. They showed
that cracking thresholds can be substantially reduced by using sharp
indenters, i.e., indenters with smaller included tip angles, such as a
three-sided indenter with the geometry of the corner of a cube.
Studies using a three-sided indenter with the geometry of a corner of
a cube have revealed that cracking thresholds can be reduced to loads

as small as 0.5 g, for which indentations and crack lengths in most
materials are submicron in dimension.
Based on fracture mechanics analysis, Lawn, et al.[174] devel-
oped a mathematical relationship between fracture toughness and
indentation crack length given as,
1/ 2
E  W 
Eq. (21) K IC = B   3/2 
H c 
where W is the applied load and B is an empirical constant depending

upon the geometry of the indenter, (also see Refs. 172 and 232).
Antis, et al.[18] conducted a study on a number of brittle materials
chosen to span a wide range of toughness. They indented with a
Vickers indenter at several loads, and measured crack length opti-
cally. They found a value of B = 0.016 to give good correlation
between the toughness values measured from the crack length, and
the ones obtained using more conventional methods. Mehrotra and
Quinto[196] used a Vickers indenter to measure fracture toughness of
the coatings. Pharr, et al.[232] tested several bulk ceramics listed in
Table 4 using Vickers, Berkovich, and cube corner indenters. They
found that the fracture toughness equation can be applied for the data
obtained with all three indenters, provided a different empirical
constant was used for a cube corner indenter. The constant B for
Vickers and Berkovich indenter was found to be about 0.016 and for
cube corner it was about 0.032. Pharr, et al.[232] further reported that
predominant cracks formed with Vickers or Berkovich indenters are
cone cracks, and with cube corner indenter, predominant cracks were
radial cracks, Fig. 15.
Figure 15. Indentations in fused quartz made with the cube corner indenter
showing radial cracking at indentation loads of (a) 12 g and (b) 0.45 g.[232]
Chantikul et al.[79] developed a relationship between fracture

toughness and the indentation fracture strength and the applied load
1 3
 E  8  1
4
Eq. (22) K Ic = c    σ f W 3 
H   
where σ is the fracture strength after indentation at a given load and

c is an empirical constant (0.59). Advantage of this analysis is that
the measurement of crack length is not required. Mecholsky et al.[194]
used this analysis to calculate fracture toughness of diamond films
on silicon. They indented the films at various indentation loads of 3
to 9 kg, and then fractured in four point flexure to measure fracture
strength. The data was then used to get fracture toughness. Equation
(22) was found to hold for the measurements. They reported a
fracture toughness of 6 and 12 µm thick diamond films on silicon on
the order of 2 MPa m .
Chaing et al.[81] developed a relationship for fracture toughness
of the coating/substrate interface, which can be used as a measure of
indentation adhesion of the coatings. This analysis is presented later
on in the section on adhesion measurements.
Nanofatigue. Delayed fracture resulting from extended service

is called fatigue. Fatigue fracturing progresses through a material via
changes within the material at the tip of a crack, where there is a high
stress intensity. There are several situations: cyclic fatigue, stress
corrosion, and static fatigue. Cyclic fatigue results from cyclic load-
ing of machine components, e.g., the stresses cycle from tension, and
compression occurs in a loaded rotating shaft. Fatigue also can occur
with fluctuating stresses of the same sign, as occurs in a leaf spring,
in a dividing board. In a low flying slider in a head-disk interface,
isolated asperity contacts occur during use, and the fatigue failure
occurs in the multilayered thin-film structure of the magnetic disk.[38]
Asperity contacts can be simulated using a sharp diamond tip in an
oscillating contact with the thin-film disk.
Li and Chu[177] developed a indentation fatigue test, called
impression fatigue. In this test, a cylindrical indenter with a flat end
was pressed onto the surface of the test material with a cyclic load
and the rate of plastic zone propagation was measured.
Wu et al.[307] developed a nanoindentation fatigue test by modi-
fying their Nanoindenter. The cyclic indentation was implemented
by servo controlling the PZT stack to drive the indenter so that the
loadcell output followed a 0.1 Hz sinusoidal loading pattern, and the
latter was specified by the cyclic frequency, and the lower and upper
limits for the desired load range. The lower limit for all the tests was
set at about 0.2 mN to perform a full load cycle indentation fatigue. A
nonzero limit was required in order to activate the load cell servo
control mode. Several maximum loads, namely 4, 16, and 24 mN
were used. Following results can be obtained: (i) endurance limit,
i.e., the maximum load below which there is no coating failure for a
preset number of cycles, (ii) number of cycles at which the coating
failure occurs, and (iii) changes in contact stiffness measured by
using the unloading slope of each cycle which can be used to monitor
the propagation of the interfacial cracks during cyclic fatigue pro-
cess. They used a conical diamond indenter with a nominal 1 µm tip.
Applied load and penetration depth were simultaneously monitored
during the entire test.
Typical nanoindentation fatigue results for a 0.11-µm DC

planar magnetron sputtered amorphous carbon films on (100) silicon
substrate deposited at argon pressure of 30 m torr is shown in Fig.
16. [307] The test was run at a maximum cyclic load of 4 mN and 0.1
Hz for a total of 105 cycles with fracture in the 93rd cycle. SEM
micrograph of the damaged zone (Fig. 16c) shows that the plastic
deformation attributed primarily to the silicon substrate, occurred in
the central indent area, and moreover, the carbon coating spalled
around the indent, and resulted in several isolated carbon flakes as
well as cantilevered flakes. Wu et al. reported that in a single
indentation test, the critical indentation load required to crack the
carbon coating was about 6 mN. The critical indentation load was
extracted by using the criterion of the applied load at which a load
drop appeared along a loading curve. Evidently, the endurance limit
can be significantly lower than the critical load of a single indenta-
tion. This scenario is analogous to the fatigue strength versus tensile
strength in macrotests. Wu et al. further reported that the carbon
films deposited at a argon pressure of 6 m torr exhibited endurance
limit of about 24 mN, about a factor of six higher than the film
deposited at 30 m torr. Scratch resistance of the film at 30 m torr was
also poorer as compared to that deposited at 6 m torr.[302][305][306]
Adhesion Strength Measurements. For measurement of ad-
hesion strength of the coating-substrate interface using indentation
method, the coating sample is indented at various loads. At low
loads, the coating deforms with the substrate. However, if the load is
sufficiently high, a lateral crack is initiated and propagated along the
coating-substrate interface. The lateral crack length increases with the
indention load. The minimum load at which the coating fracture is
observed is called the critical load, and is employed as the measure of
coating adhesion (Fig. 17). For relatively thick films, the indentation
is generally made using Brinell hardness tester with a diamond sphere
of 20 µm radius,[278] Rockwell hardness tester with a Rockwell C
120o cone with a tip radius of 200 µm,[196] or a Vickers pyramidal
indenter.[5][81][180] However, for extremely thin films, a Berkovich
indenter,[263] or a conical diamond indenter with a tip radius of 5 mm
and 30° of included angle,[285] is used in a nanoindenter.
(a)
(b) (c)
Figure 16. Typical microindentation fatigue results from 0.11-µm thick dc

sputtered amorphous carbon on (100) Si, (a) the direct outputs from a strip chart
recorder of applied load (LC) and indenter position (IND) (Maximum load = 4
mN, frequency = 0.1 Hz), (b) a plot of the indentation fatigue loading curve
versus total penetration depth (the plot includes only the load cycles from 91 to
100; note the abrupt depth increase started from the 93rd cycle) and (c) SEM
micrograph of the fatigue indent.[307]
Figure 17. Schematic illustration of the indentation method for adhesion

measurement.
It should be noted that the measured critical load Wcr is a

function of hardness and fracture toughness in addition to the adhe-
sion of coatings. Chiang et al.[81] have related the measured crack
length during indentation, the applied load and critical load (at which
coating fracture is observed) to the fracture toughness of the sub-
strate-coating interface. A semianalytical relationship derived be-
tween the measured crack length c and the applied load W:
1/ 2
 W 
Eq. (23) c = α  1 − cr  W 1 / 4
 W 
α 1t c3 / 2 H 1 / 2
α2 =
where (K Ic )interface
α1 is a numerical constant, tf is the coating thickness, H is the mean

hardness and (KIc )interface is the fracture toughness of the substrate-
coating interface. Mehrotra and Quinto[196] used this analysis to
calculate fracture toughness of the interface.
Marshall and Oliver[184] estimated adhesion of composites by
measuring the magnitude of shear (friction) stresses at fiber/matrix
interfaces in composites. They used a Berkovich indenter to push on
the end of an individual fiber and measured the resulting displace-
ment of the surface of the fiber below the matrix surface (due to
sliding). The shear stress was calculated from the force-displacement
relation obtained by analysis of the frictional sliding. The force and
displacement measurements were obtained only at the peak of the
load cycle, and the sliding analysis was based on sliding at constant
shear resistance at the interface. These experiments provided mea-
surements of average shear stresses at individual fibers.
2.5 Scratch Resistance/Adhesion Measurements
Macro-, micro-, and nanoscratch techniques are used to mea-

sure scratch resistance of surfaces from bulk materials to a few
nanometer thick films. Adhesion describes the sticking together of
two materials. Adhesion strength, in practical sense, is the stress
required to remove a coating from a substrate. Indentation, described
earlier, and scratch on micro-and nanoscales are the two commonly
used techniques to measure adhesion of thin hard films with good
adhesion to the substrate (>70 MPa).[37][46][60][76][199]
Scratching the surface with a fingernail, or a knife, is probably
one of the oldest methods for determining the adhesion of paints and
other coatings. Scratch tests to measure adhesion of films was first
introduced by Heavens.[139] A smoothly round chrome-steel stylus
with a tungsten carbide, or Rockwell C diamond tip (in the form of
120° cone with a hemispherical tip of 200 mm radius),[196][223]
[224][261][287] or Vickers pyramidal indenter,[70][73] for macro-and
microscratching a conical diamond indenter (with a 1 µm or 5 µm of

tip radius and 60° of included angle), for nanoscratching[56][131]
[132][302][305][306] is drawn across the coating surface. A vertical load is
applied to the scratch tip, and is gradually increased during scratch-

ing until the coating is completely removed. The minimum or critical
load at which the coating is detached or completely removed is used
as a measure of adhesion.[2][30][70][73][76][80][119][147][149][153][170][196]
[199][223]–[225][253][261][287][288][299][301][302][305][306] It is a most com-
monly used technique to measure adhesion of hard coatings with

strong interfacial adhesion (>70 MPa).
For a scratch geometry shown in Fig. 18, surface hardness H is
given by
Wcr
H=
Eq. (24) π a2
and adhesion strength t is given by [30]

Eq. (25a) t = H tan è
Wcr  a 
=  2 
π a2 (
 R − a 2 )
1/ 2

or
Wcr
Eq. (25b) τ= if R >> a
π aR
where Wcr is the critical normal load, a is the contact radius and R is
the stylus radius.
Figure 18. Geometry of the scratch.
Burnett and Rickerby,[73] and Bull and Rickerby [70] analyzed

scratch test of a coated sample in terms of three contributions: (i) a
ploughing contribution, which will depend on the indentation stress
field, and the effective flow stress in the surface region, (ii) an
adhesive friction contribution due to interactions at the indenter-
sample interface, and (iii) an internal stress contribution since any
internal stress will oppose the passage of the indenter through the
surface, thereby effectively modifying the surface flow stress. They
derived a relationship between the critical normal load Wcr and the
work of adhesion W ad
1
π a 2  2 EWad  2
Eq. (26) Wc =  
2  t 
where E is the Young’s modulus of elasticity and t is the coating

thickness. Plotting of Wc as a function of a2/t1/2 should give a straight
line of the slope π(2EWad /t)1/2/2 from which Wad can be calculated.
Bull and Rickerby suggested that either the line slope (interface
toughness) or Wad could be used as a measure of adhesion.
An accurate determination of critical load Wcr sometimes is
difficult. Several techniques, such as (i) microscopic observation
(optical or scanning electron microscope) during the test, (ii) chemi-
cal analysis of the bottom of the scratch channel, (with electron
microprobes) and (iii) acoustic emission, have been used to obtain
the critical load.[149][223][224][253][261][287][302][306] The acoustic-emis-
sion technique has been reported to be very sensitive in determining
critical load. Acoustic emission starts to increase as soon as cracks
begin to form perpendicular to the direction of the moving stylus. In
some instruments, tangential (or friction), force is measured during
scratching.[16][55]–[57][131]–[133][147][287][301][302][306] An instant increase
in the friction force during scratch has also been used as an indicator
of a damage event.
Apparatuses For Scratch Measurements On Macroscale.
Several macroscratch testers are commercially available, such as the
Taber shear/scratch tester model 502 with a no. 139–58 diamond
cutting tool (manufactured by Teledyne Taber, North Tonawanda,
NY) for thick films, Revetest automatic scratch tester (manufac-
tured by Centre Suisse d’Electronique et de Microtechnique S.A.,
CH-2007, Neuchatel, Switzerland), and scratch tester with friction
force attachment (manufactured by VTT, Technical Research Center
of Finland, Espoo, Finland) for thin films.[223][224][253][261][288] Of all
the macroscratch testers available commercially, Revetest is used
most extensively by numerous research groups to study the adhesion
characteristics of hard coatings.
Revetest automatic scratch tester is shown schematically in

Fig. 19. This tester has a Rockwell C diamond tip with a cone angle
of 120°, and a tip radius of about 200 µm. Normal load is applied
through a spring loaded arm in steps of 1 N up to a maximum of 200
N. This tester was further modified by replacing-loaded arm with an
electromagnetic coil to operate at lower load of 10 mN to 30 N with
a detection sensitivity of 10 mN. This tester is equipped with an
acoustic emission detector, and a device which enables the tangential
force in the direction of displacement to be measured. Damage at the
coating-substrate interface usually results in an increase in friction
force, and a generation of an acoustic emission signal.
Figure 19. Schematic illustration of the scratch method for adhesion measurement.
An example of a scratch test, optical micrographs of typical

failure mode of sputtered TiN coating on a steel substrate are pre-
sented in Fig. 20a. The acoustic emission curves showing the maxi-
mum and minimum signals are plotted as a function of stylus load in
Fig. 20b. As the stylus load was increased to 7.8 N, the coating
started to crack perpendicular to the steel surface at the edge of the
channel. The onset of coating damage was accompanied by a sudden
increase in acoustic emission at loads greater than 7.8 N (region b in
Fig. 20b). The very high normal load can cause the detached coating
particles either to the completely pressed back into the substrate or be
partially removed, thus leaving a smooth surface within the channel.
(a)
(b)
Figure 20. (a) Optical micrographs of channels produced during scribing from
right to left under various stylus loads in Crofer 1700 coated with sputtered
TiN about 1.5 µm thick. (b) Acoustic emission signal maxima and minima
taken from the curves recorded within the scratch channel length between 0.5
and 2.5 mm.[149]
Apparatuses For Scratch Measurement On Microscale. In

principle, the microscratch testers are quite similar to those used for
macroscratch but only with a difference in the normal load and size
and radius of the diamond tip. In the microscratch testers, usually the
normal load ranges from a few hundred µN to a few tens of mN and
tips radii varies from about 100 nm to a couple of microns. A
prototype microscratch tester was developed by Wu et al. [307] at IBM
Almaden Research Center by modifying a nanoindenter. With this
instrument, following measurements can be made simultaneously
during a scratch test: applied load and tangential load along the
scratch length (coefficient of friction), critical load, i.e., applied
normal load corresponding to an event of coating failure during a
scratch process, total and plastic depth along the scratch length, the
accumulated acoustic emission (AE) counts versus the scratch length.
The commercially available Nanoindenter, described earlier, has
also been modified for making microscratch measurements.
We describe the microscratch and tangential force attachments
of Nanoindenter which allows making of the scratches of various
lengths at programmable loads. Tangential (friction) forces can also
be measured simultaneously.[16][56] The additional hardware for the
tangential force option includes a set of proximity (capacitance)
probes for measurement of lateral displacement or force in the two
lateral directions along x and y, and a special “scratch collar” which
mounts around the indenter shaft with hardness indenter, Fig. 21.
The scratch collar consists of an aluminum block, mounted around
the indenter shaft, with four prongs descending from its base. Two of
these prongs hold the proximity probes, and the set screws set them
in place, while the other two prongs hold position screws (and
corresponding set screws). The position screws serve a dual purpose;
they are used to limit the physical deflection of the indenter shaft,
and they are used to lock the indenter shaft in place during tip change
operations. A scratch block is mounted on the end of the indenter
shaft, in line with the proximity probes and the positioning screws.
Finally, the scratch tip itself is mounted on the end of the indenter
shaft, covering the scratch tip. The scratch tip is attached to the
scratch block with two Allen head screws. The scratch tip can be a
Berkovich indenter, or a conventional conical diamond tip with a tip
radius of about 1 to 20 µm, and an included angle of 60 to 90°
(typically 1 µm or 5 µm of tip radius with 60° of included
angle.)[56][131]–[133][302][306] The tip radius does not have to be very
small as it will get blunt readily.
Figure 21. Schematic of the tangential force option hardware (not to scale and
the front and rear prongs not shown).[16]
During scratching, a load is applied up to a specified indenta-

tion load or up to a specified indentation depth, and the lateral motion
of the sample is measured. In addition, of course, load and indenta-
tion depth are monitored. Scratches can be made either at the con-
stant load, or at ramp up load. Measurement of lateral force allows
the calculations of coefficient of friction during scratching. The
resolution of the capacitance proves to measure tangential load is
about 50 µN, therefore, a minimum load of about 0.5 mN can be
measured, or a minimum normal load of about 5 mN should be used
for a sample with coefficient of friction of about 0.1. Microscopy of
the scratch produced at ramp up load allows the measurement of
critical load required to break up of the film (if any) and scratch
width, and general observations of scratch morphology. Additional
parameters which are used to control the scratch are scratch length
(µm), draw acceleration (µm/s2), and draw velocity (µm/s). The
latter parameters control the speed with which the scratch is per-
formed. The default values of 10 µm/s2 and 10 µm/s provide safe
rates for performing the scratch. Draw velocity is limited by the
maximum rate of data acquisition (during a scratch the maximum
rate is approximately 2/s) and the length of the desired scratch. Thus,
a scratch with a desired 20 points over 1 mm must have a draw
velocity no greater than 100 µm/s.
Wu[302] has used the scratch technique to study the adhesion of
diamondlike carbon and zirconia films deposited on Si(100) sub-
strates. Figure 22 shows the scratch morphology at increasing nor-
mal loads, and typical scratch data (normal load, tangential load and
acoustic emission, as well as calculated apparent coefficient of
friction). We note that all three monitored outputs (LC, TG, and AE),
detected the first spallation event of the carbon coating by showing
sudden changes in their output signals. Correlation between the
delamination pattern and the sudden change in the scratch loading is
clearly observed.
Figure 22. Scratch morphology and scratch loading curves of 0.11-µm thick
d-c sputtered diamondlike carbon film on a Si substrate.[302]
Gupta and Bhushan[132] measured scratch depths during

scratching and residual depth after scratching. The surface profile of
the coated surface was first obtained by translating the sample at a
low load of about 1.5 mN, which is insufficient to damage the sample
surface. The actual scratches were made by translating the sample
while ramping the loads on the conical tip over different loads, for
instance ranging from 1.5 mN to 45 mN. The actual depth during
scratching was obtained by subtracting the initial profile from the
scratch depth measured during scratching. In order to measure per-
manent depth, the scratched surface was profiled at a low load of 1.5
mN and was subtracted from the surface profile before scratching. A
typical scratch experiment consisted of seven subsequent steps:
1. Approaching the surface.
2. Indent into sample surface by loading the tip to
1.5 mN.
3. Translating the sample at a constant load of 1.5 mN
at a speed of 5 microns per second.
4. Translating the sample in the opposite direction at
ramping load to a load ranging from 1.5 to 45 mN at
a speed of 5 micron per second.
5. Unloading of the tip to 1.5 mN.
6. Translating the sample at constant load of 1.5 mN
at a speed of 5 micron per second.
7. Final unloading of the tip.
The 500-micron long scratch at ramping normal load was made
during the fourth step, and surface profiles before and after scratch
were obtained during the third and sixth steps, respectively. The
scratch depth profiles obtained during and after the scratch of 20 nm
thick carbon coatings deposited on silicon by cathodic arc, and
sputtering are plotted with respect to initial profile after the
cylindrical curvature is removed, Fig. 23. Reduction in scratch
depth after scratching is observed in Fig. 23. The reduction in
scratch depth after scratching is attributed to an elastic recovery
after the removal of the normal load. It appears that the scratch depth
after scratching indicates the final depth which reflects the extent of
permanent damage and ploughing of the tip into the sample surface.
We believe that the scratch depths after scratching are probably more
relevant for visualizing the damage that can occur in real applica-
tions. The abrupt increase in the coefficient of friction and scratch
depth is associated with damage to the coating. The cathodic arc
coating exhibits an almost constant low coefficient of friction of
about 0.1–0.15 during the initial stages and an abrupt increase in
friction when normal load exceeds the critical load, the load suffi-
cient to damage the coating. Sputtered carbon coating exhibits a
gradual increase in the coefficient of friction with increasing normal
load from the beginning of the scratch.
Figure 23. Coefficient of friction profile during scratching and scratch depth
during and after scratching as a function of normal load for scratches made on
20 nm thick carbon coatings deposited on silicon by cathodic arc and sputtering
deposition techniques.[132]
Bulk Si exhibits a very low coefficient of friction of 0.05 at the

beginning of a scratch at 2 mN, Fig. 24. The coefficient of friction
remains constant up to 4 mN. The coefficient of friction increases
abruptly from 0.05 to 0.15, and then gradually increases to 0.25, as
the normal load increased from 4 to 18 mN, and to 0.9 when the load
exceeded 18 mN. Particulate debris of submicron size was observed
when the normal load exceeded 18 mN. First abrupt increase in
friction corresponds to an initiation of ploughing of the silicon
surface by the tip, whereas the second abrupt increase in friction
corresponds to a catastrophic damage of the surface.
Figure 24. SEM images of various regions and coefficient of friction profiles as
a function of normal load for 500 µm long scratches made on single-crystal
silicon (111) at 2 to 20 mN.[56]
The dependence of coefficient of friction on increasing normal

load and shape and size of debris generated during scratching can be
used to obtain important information regarding the adhesion of the
coating with the substrate, and how the coating or sample surface is
damaged during scratching.
Apparatuses For Scratch Measurement On Nanoscale.
Nanoscratch measurements were made by Bhushan et al.[53] and
Bhushan[40] using a modified commercial AFM/ FFM (Nanoscope
III from Digital Instruments, Inc. Santa Barbara, Calif.). The modi-
fied AFM/ FFM will be described later on in the next section, on the
measurements of friction and wear on nanoscale. The scratches were
made by a three-sided pyramidal diamond tip with a tip radius of
about 100 nm at normal loads ranging from 10 to 100 µN. Sample
surfaces were scanned before and after the scratch to obtain the
initial and the final surface topography at a load of about 0.05 µN,
over an area larger than the scratches region. AFM images of scratches
made on (111) silicon, PECVD-oxide coated silicon, dry-oxidized
silicon, and C+ implanted (1 × 10-17 ions cm-2 at 100 keV) silicon are
shown in Fig. 25.[51] As expected, the scratch depth increases with an
increase in normal load. The depths of scratches at 40 µN on
PECVD-oxide coated silicon and (111) silicon are about 5 and 20
nm, respectively. PECVD-oxide coated silicon has the largest scratch
resistance followed by dry-oxidized silicon and ion-implanted sili-
con. Ion implantation showed no improvements on scratch resis-
tance because the depth of the scratches is lower than the depth of
implanted zone beneath the surface.
This study demonstrates that scratches with a depth of a few
nanometers can be made at very low loads using AFM/ FFM. This
help characterizing the scratch resistance/ adhesion of top few top
tens of monolayers of bulk materials or coating of the order of a few
nanometers thickness.
Figure 25. Surface profiles for scratched (a) (111) single-crystal silicon, (b)
PECVD-oxide coated Si, (c) dry-oxidized Si and (d) C+-implanted Si. The loads
used for various scratches at ten cycles are indicated in the plot.[51]
3.0 MEASUREMENT OF FRICTION AND WEAR
Friction and wear between two moving surfaces depends on

mechanical properties of the mating materials such as: hardness, elastic
modulus, fracture toughness; other physical and chemical properties
such as: thermal conductivity, surface energy, adsorption character-
istics, chemical reactivity; surface conditions such as: roughness and
apparent area of contact; and operating conditions such as: load, speed,
interface temperature, and environment. Various friction mecha-
nisms such as adhesion and ploughing, that contribute to friction and
wear, are strongly affected by the properties of mating surfaces and
operating conditions. Surface roughness of mating surfaces and area
of contact between these surfaces play a key role in deciding friction.
For instance, rigid thin-film magnetic disk are deliberately rough-
ened to reduce the area of contact to obtain lower friction between a
head slider and the magnetic disk. The friction and wear tests should
be performed under the close-to-ideal conditions. The conditions and
test geometry for friction and wear tests should be selected on the
basis of the actual operating conditions of the components.
With the advent of new surface imaging tools like atomic force
microscope (AFM), friction force microscope (FFM), and point
contact microscope (PCM), it is possible to perform friction tests at
ultra small loads of a couple of µN against a sharp tip of about 10 to
100 nm tip-radius which simulates a single-asperity contact condi-
tions.[39][40][58] Sliding tests with a single-asperity contact are more
close-to-ideal conditions of micromechanical devices, and help un-
derstand the failure mechanisms.
In this section, we present the apparatuses used for friction and
wear measurements on macro- to nanoscales.
3.1 Friction and Wear Measurements on Macroscale
Accelerated friction and wear tests on macroscale are con-

ducted to rank the friction and wear resistance of coatings, or bulk
materials to optimize their selection or development for specific
applications. After these coatings or bulk materials have been ranked
by accelerated friction and wear tests, the most promising candidates
(typically from one to three), should be tested in the actual machine

under the actual operating conditions (functional tests). Accelerated
friction and wear tests should accurately simulate the operating
conditions to which the material pair will be subjected. If these tests
are properly simulated, an acceleration factor between the simulated
test, and the functional tests can be empirically determined so that
the subsequent functional tests can be minimized, saving consider-
able test time. Standardization, repeatability, short testing time, and
simple measuring and ranking techniques are desirable in these
accelerated tests. The coefficient of friction of a material, or a
coating depends not only on the counterface materials, but on the
operating conditions such as speed, load, lubrication, and environ-
ment. Shown in Table 1 are a couple of examples of coefficients of
friction and wear rate for various bulk material and coating combina-
tions. We note that with a proper selection of materials and operating
conditions, one can achieve as low coefficient of friction as 0.02
between two solid materials without using any lubricant. In this
section, we will present the design methodology and typical test
geometries for friction and wear tests.
Design Methodology. Design methodology of a friction and
wear test consists of four basic elements: simulation, acceleration,
specimen preparation, and friction and wear measurements. Simula-
tion is most critical, but no other elements should be overlooked.
Proper simulation ensures that the behavior experienced in the
test is identical to that of the actual system. A successful simulation
requires the similarity between the functions of actual system, and
those of the test system, i.e., similarity of inputs and outputs, and of
the functional input-output relations. To obtain this similarity, first
the mating materials, the lubricant, and the operating conditions of
the test should be the same as the actual system requirements.
Selection of the contact geometry depends on the geometry of the
function to be simulated. Other factors besides contact type that
significantly influence the success of a simulation include type of
motion, load, speed, and operating environment (contamination,
lubrication, temperature, and humidity). Specimen preparation plays
a key role in obtaining repeatable/reproducible results.
Table 1. Typical Values of Coefficients of Friction for Selected

Materials and Coatings in the Ambient Environments (~ 22°C, 50%
RH) Unless Otherwise Specified[46][64][142][227]
Materials pair
Bulk Counterface material Coefficient of friction
material/Coating
and Substrate
Bulk materials
Aluminum Aluminum 1.0

Copper Copper 0.8
Silver Silver 0.9
Brass Brass 0.4
Cast iron Cast iron 0.6
Mild steel 0.4
Brass 0.25
Bronze 0.25
Copper 0.3
Mild steel Mild steel 0.8
Aluminum 0.5
Nickel 0.7
Silver 0.5
Copper 0.8
Brass 0.3
Bronze 0.3
Babbitt 0.3
Graphite 0.15
PTFE 0.1
Tool steel Tool steel 0.4
Brass 0.25
Polyethylene 0.25
PTFE 0.1
Nickel Nickel 1.1
Chromium Chromium 0.4
Silicon nitride Silicon nitride 0.3
(Cont’d.)
Table 1. (Cont’d.)
Materials pair
Bulk Counterface material Coefficient of friction
material/Coating
and Substrate
Tungsten carbide Tungsten carbide 0.35
Steel 0.4
Natural diamond Natural diamond 0.05
PTFE PTFE 0.05
Coatings/Surface
treatments
MoS2 (Sputtered) Steel 0.05–0.1 (Ambient)
WC-Co 0.02 (Vacuum)
Graphite (air sprayed) Steel 0.1–0.2 (Ambient)
0.4–0.6 (Dry)
PTFE (Air sprayed) Steel 0.03–0.1
Silver/Gold (Sputtered) Steel 0.1–0.25
Al2O3 (CVD) Steel 0.2–0.5
TiN (Sputtered) Steel 0.15–0.5
TiC(Sputtered) Steel 0.2–0.5
Diamond(HFCVD)/Si Steel 0.1–0.2
a-C:H (Sputtered) Steel 0.15–0.3
a-C:H (PEPVD) Steel 0.15–0.3
The coefficient of friction is generally measured during a wear

test. It is calculated from the ratio of friction force to applied normal
force. The stationary member of the material pair is mounted on a
flexible member, and the frictional force is measured using the strain
gauges or displacement gauges.[36]–[38] Common wear measurements
are weight loss, volume loss, or displacement scar width or depth, or
other geometric measures, and indirect measurements such as, time
required to wear through a coating or load required to cause severe

wear, or a change in the surface finish, size of indentation marks, or
width and depth of scratches. Scanning electron microscopy of worn
surfaces is commonly used to measure microscopic wear. Other less
commonly used techniques include radioactive decay, scanning tun-
neling microscopy (STM), and atomic force microscopy (AFM).
The resolution of various macroscopic wear measurement tech-
niques are compared in Table 2.
Table 2. Resolution of Several Macroscopic Wear Measurement

Techniques
Typical Test Geometries. The choice of a test geometry

depends on the wear situation to be simulated. In this section, we will
present a few most commonly used test geometries to rank coatings
and materials in terms of their resistance to sliding wear, and rolling
contact fatigue wear. However, depending on the requirements, tests
can be performed with replicas and facsimiles of the actual devices.
Many accelerated test apparatuses are commercially available that
allow control of such factors as sample geometry, applied load,
sliding velocity, ambient temperature, and humidity. Bayer,[24]–[27]
Benzing et al.[31] Bhushan,[37][38] Bhushan and Gupta,[46] Clauss,[87]

Nicoll,[209] and Yust and Bayer[313] have reviewed the various fric-
tion and wear testers that have been used in various tribological
applications.
Sliding Friction And Wear Tests. In the sliding wear test
configurations, ball, pin, cylinder, or ring of one material slides over
the disk, block, or cylinder of another material in the presence or
absence of a lubricant. The most commonly used interface geom-
etries used to rank coatings and materials in terms of their resistance
to sliding wear are schematically shown in Fig. 26, and are compared
in Table 3. With these test geometries, static, or dynamic loading can
be applied, and the tests can be performed either in the presence or
absence of a lubricant. In the pin-on-disk test apparatus (Fig. 26a),
the pin is held stationary, and the disk rotates or oscillates. The pin
can be a non rotating ball, a hemispherically tipped rider, a flat-ended
cylinder, or even a rectangular parallelepiped. In the pin-on-flat test
apparatus (Fig. 26b), a flat moves relative to a stationary pin in
reciprocating motion, such as in a Bowden Leben apparatus. The
pin-on-cylinder test apparatus (Fig. 26c), is similar to the pin-on-disk
apparatus, except that loading of the pin is perpendicular to the axis
of rotation or oscillation. In the thrust-washer test apparatus (Fig.
26d), the flat surface of a washer (disk or cylinder) rotates or
oscillates on the flat surface of a stationary washer, such as in the
Alpha model LFW-3. In the pin-into-bushing test apparatus (Fig. 26e),
the axial force necessary to press an oversized pin into a bushing is
measured, such as in the Alpha model LFW-4. The normal (axial)
force acts in the radial direction, and tends to expand the bushing; the
radial force can be calculated from the material properties, the
interference, and the change in the bushing’s outer diameter. Divid-
ing the axial force by the radial force gives the coefficient of friction.
In the rectangular flat, on a rotating cylinder test apparatus (Fig. 26f),
two rectangular flats are loaded perpendicular to the axis of rotation
or oscillation of the disk. The crossed cylinders test apparatus (Fig.
26g), consists of a hollow (water cooled) or solid cylinder as the
stationary wear member and a solid cylinder as the rotating or
oscillating wear member that operated at 90o to the stationary mem-

ber, such as in Reichert wear tester. The four ball test apparatus (Fig.
26h), also called the Shell four-ball tester, consists of four ball in the
configuration of an equilateral tetrahedron. The upper ball rotates
and slides against the lower three balls, which are held in a fixed
position. This test configuration is extensively used to study liquid
lubricants.
Figure 26. Schematic illustration of typical interface geometries used for sliding
friction and wear tests: (a) pin-on-disk, (b) pin-on-flat, (c) pin-on-cylinder, (d)
thrust washer, (e) pin-into-bushing, (f) rectangular flats on rotating cylinder, (g)
crossed cylinders and (h) four ball.[46]
Table 3. Details of Typical Test Geometries for Sliding Friction and

Wear Testing
Rolling-Contact Fatigue Wear Tests. A number of rolling-

contact fatigue (RCF) tests are used for testing materials and lubri-
cants for rolling-contact applications such as rolling element bear-
ings, and gears. In RCF test apparatus, basically a pair of driven
rollers are pressed against one another, and the surface damage on
roller surfaces is monitored with the number of cycles. The surface
damage could be the appearance of cracks, change in surface texture
or roughness, or spalling of material. In general, rolling-contact
fatigue (RCF) wear is compared in terms of number of cycles
sufficient to result a specific damage on the rollers. The most com-
monly used interface geometries used to rank coatings and materials
in terms of their resistance to rolling-contact fatigue wear are sche-
matically shown in Fig. 27. The disk-on-disk test apparatus (Fig.
27a) uses two disks or a ball-on-disk rotating against each other
on their outer surfaces (edge loaded). The rotating four ball test
apparatus (Fig. 27b) consists of four balls in the configuration of an

equilateral tetrahedron. The rotating upper ball is deadweight loaded
against the three supporting balls (positioned 120° apart), which
orbit the upper ball in rotating contact. The rolling-element-on-flat
test apparatus consists of three balls or rollers equispaced by a retainer
that are loaded between a stationary flat washer and a rotating
grooved washer (Fig. 27c). The rotating washer produces ball mo-
tion, and serves to transmit load to the ball and the flat washer.[42]
Figure 27. Schematic illustration of typical interface geometries used for

rolling-contact fatigue wear tests: (a) disk-on-disk, (b) rotating four ball, (c)
balls-on-flat.[46]
3.2 Friction and Wear Measurements on Micro- and Nanoscale
Friction and wear measurements on a micro-to nanoscale can

be performed using a nanoindenter in scratch mode which was
described earlier, or an atomic force microscope, or friction force
microscope.[41][58][185] The coefficient of friction can be measured

using a nanoindenter in the scratch mode by monitoring the friction
force during scratching at low loads on the order of a few hundred
microNewtons. The selection of the normal load should be such that
it is insufficient to cause damage to the sample surface, or to result in
ploughing of the tip into the sample surface. In these experiments the
diamond tip can be replaced with tips of different materials to study
the friction of a coating against a variety of materials.
The atomic force microscope (AFM), was developed by Bin-
ning, et al.,[59] to measure ultrasmall forces on the order of 1 nN or
less present between the AFM tip surface, and a sample surface.
Binning and co-workers measured these small forces by measuring
the motion of a cantilever beam having an ultrasmall mass. For
imaging surfaces, AFM can be thought of as a nanometer scale
profiler. A sharp tip at the end of a cantilever is brought into contact
with a sample surface by moving the sample with piezoelectric
transducers. During initial contact, the atoms at the end of the tip
experience a very weak repulsive force due to electronic orbital
overlap with the atoms in the sample surface. The force acting on
the tip causes a lever deflection which is measured by tunnel-
ing, [59] capacitive detectors,[192] or optical detectors such as laser
interferometry.[109][185] The optical techniques are believed to be
reliable and easily implementable detection methods. Mate et al.[185]
were the first to modify an AFM in order to measure both normal and
friction forces; the modified instrument is generally called the
friction force microscope (FFM), or lateral force microscope
(LFM). This group measured the atomic-scale friction of a tung-
sten tip sliding on a basal plane of a single grain of highly oriented
pyrolytic graphite (HOPG). Since then, several research groups have
developed FFMs of various design. Ruan and Bhushan[242] Bhushan
et al.[53][58] and Bhushan[40][41] have modified a commercial AFM/
FFM (Nanoscope III from Digital Instruments, Inc., Santa Barbara,
Calif.) to conduct studies of friction, scratching, wear, indentation,
lubrication, and nanofabrication.
A modified AFM/FFM is shown in Fig. 28. Simultaneous

measurements of friction force, and surface roughness can be made
using this instrument. The sample is mounted on a piezoelectric tube
(PZT) scanner which can precisely scan the sample in the horizontal
(xy) plane, and can move the sample in the vertical (z) direction. A
sharp tip at the free end of a cantilever is brought in contact with the
sample. A laser beam from a laser diode is focused on to the back of
a cantilever near its free end. The cantilever is tilted downwards at
about 10o with respect to the horizontal plane. The beam is reflected
from the cantilever, and is directed through a mirror on a split
photodetector with four quadrants. Two quadrants (top and bottom)
of the detector are used for topography measurements. As the sample
is scanned under the tip, topographic features of the sample cause the
tip to deflect in the vertical direction. This tip deflection will change
the direction of reflected laser beam, changing the intensity differ-
ence between the top and bottom photodetectors (AFM signal). For
topographic imaging, applied normal force is kept constant, a feed-
back current is used to modulate the voltage applied to the PZT
scanner to adjust height of the PZT. The PZT height variation is a
direct measure of surface roughness of the sample. For measurement
of friction force being applied at the tip surface during sliding, the
other two (left and right) quadrants of the photodetector (arranged
horizontally) are used. The sample is scanned back and forth in a
direction orthogonal to the long axis of the cantilever beam. Friction
force between the sample and the tip will produce a twisting of the
cantilever. As a result, the laser beam will be reflected out of the
plane defined by the incident beam, and the beam reflected vertically
from an untwisted cantilever. This produces an intensity difference
between the left and right detectors (FFM signal), and is directly
related to the degree of twisting, and hence, to the magnitude of
friction force. One problem associated with this method, is that any
misalignment between the laser beam, and the photodetector axis
would introduce error in the measurement. However, by following
the procedures developed by Ruan and Bhushan[242] in which the
average FFM signal as the sample is scanned in two opposite direction,
and is subtracted from the friction profiles of each scan, the misalign-
ment effect can be eliminated. By following procedures developed
by Ruan and Bhushan[242] voltage corresponding to normal and
friction forces can be converted to force units. By making measure-
ments at various normal loads, the average value of the coefficient of
friction is obtained which can then be used to convert the friction
profile to the coefficient of friction profile. Thus, any directionality
and local variation of friction can be measured. Surface topogra-
phy data can be measured simultaneously with the friction data, and
a local relationship between the two profiles can be established.
Figure 28. Schematic of AFM/FFM instrument.[53]
Bhushan et al.[53][58] have measured microfriction data for

magnetic tapes and magnetic disks by FFM, by following the afore-
mentioned procedure of microfriction measurements. Figure 29 are
the surface roughness profile, the slope of surface roughness profile
taken along the sample sliding direction, and friction profile of an
unlubricated, textured, rigid disk, respectively. No direct correlation
between the surface profiles, and the corresponding friction profiles
is observed in these figures, e.g., high and low points on the friction
profile do not correspond to high and low points on the roughness

profile, respectively. However, there is a correlation between the
slope of the roughness profiles, and the corresponding friction pro-
files. Bhushan and Ruan[52] have presented evidence of similar
correlation on a microscale of 500 nm × 500 nm for the tape and disk
samples. A ratchet mechanism explains the correlation of friction to
the local surface slope, with the ascending edge of an asperity having
a larger friction force than that at the descending edge. They also
reported that the macroscale coefficient of friction is about a factor of
five larger than that of the average microscale coefficient of friction
of the corresponding samples.[52] Larger macro coefficient of fric-
tion may be the result of plowing effect associated with the macroscale
measurement, since visible wear scars were observed on the samples
after the measurement.[58] Bhushan and Ruan[52] had also observed
the directionality in the local variation of microscale friction data as
the samples were scanned in either direction, resulting from the
scanning direction, and the anisotropy in the surface topography.
Bhushan et al.[53] have also made wear measurement on
microscale by scanning the sample (in 2D) using a FFM. Surface
profiles of the wear scars generated in a single cycle on an as-
polished, unlubricated disk at different loads are shown in Fig. 30.
Wear depths as a function of normal load for the as-polished disks
(lubricated and unlubricated) are shown in Fig. 31. The normal force
for the imaging was about 0.5 µN, and the loads used for the wear
were 20, 50, 80, and 100 µN. They reported that wear takes place
relatively uniformly across the disk surface, and appears to be essen-
tially independent of the lubrication for the disks studied. For both
lubricated and unlubricated disks, the wear depth increases slowly
with load at low loads with almost the same wear rate. As the load is
increased to about 60 µN, wear increases rapidly with load. The wear
depth at 50 µN is about 14 nm, slightly less than the thickness of the
carbon film. The rapid increase of wear with load at loads larger than
60 µN is an indication of the breakdown of the carbon coating on the
disk surface. Wear depths as a function of number of cycles for the
as-polished disks (lubricated and unlubricated) are shown in Fig. 32.
Again, for both unlubricated and lubricated disks, wear initially

takes place slowly with a sudden increase between 40 and 50 cycles at
10 µN. The sudden increase occurred after 10 cycles at 20 µN. This
rapid increase is associated with the breakdown of the carbon coating.
The preceding microfriction and microwear results illustrate
that friction force microscopy can be used for various tribological
studies on microscale. By this technique it is possible to identify
normal load and number of cycles sufficient to initiate damage to the
20-nm thick hard carbon overcoat on a thin-film magnetic disk.
Figure 29. A 200 nm × 200 nm scan of an unlubricated, textured disk. (a)

Surface roughness profile (σ = 2.9 nm), (b) slope of the roughness profile (mean
= -0.03, σ = 0.22), and (c) friction profile (mean = 5.7 nN, σ = 1.6 nN) for a
normal load of 140 nN.[53]
Figure 30. Surface profile of an as-polished unlubricated disk showing the worn
region (center 2 µm × 2 µm) after one cycle of wear. The normal load and the
number of test cycles are indicated in the figure.[53]
Figure 31. Wear depth as a function of load for both lubricated and unlubricated
as-polished disks after one cycle.[53]
Figure 32. Wear depth as a function of number of cycles for as-polished,

lubricated and unlubricated disks at 10 µN and for an as-polished, unlubricated
disk at 20 µN.[53]
3.3 Friction and Wear Measurements in MEMS
Reduction of friction and wear are critical to successful perfor-

mance of micromechanical elements requiring relative motion. Since
electrostatic forces, van der Waals forces, and viscous drag forces in
liquids change in proportion to the surface area of the element, these
become dominant over electromagnetic forces, and inertial forces,
which are in proportion to the volume of the element. In MEMS, the
electrostatic and van der Waals forces are the major sources of the
normal force which directly affect the friction force. Wear of sliding
components also need to be minimized. One of the MEMS devices in
which friction and wear issues are critical, is a micromotor. Friction
and wear studies have been conducted on actual or simulated MEMS
components, as well as on the coupons in conventional accelerated
tribological tests.
To in-situ measure static friction of rotor-bearing interface in a
micromotor, Tai and Muller[274] measured the starting torque (voltage)
and pausing position for different starting positions under a constant
bias voltage. A friction-torque model was used to obtain the coefficient
of static friction. To in-situ measure kinetic friction of the turbine

and gear structures, Gabriel et al.[114] used a laser-based measure-
ment system to monitor the steady-state spins and decelerations. Lim
et al.[179] designed and fabricated a polysilicon microstructure to in-
situ measure static friction of various films. The microstructure
consisted of shuttle suspended above the underlying electrode by a
folded beam suspension. A known normal force was applied, and
lateral force was measured to obtain coefficient of static friction.
Gabriel et al.[114] reported wear of turbine after spinning at 2500 rps
for 5 minutes. Mehregany et al.[195] developed a quantitative method
for in-situ wear measurements in micromotors. They used a wobble
micromotor under electric excitation for quantitative wear measure-
ment. Since the gear ratio of the wobble micromotor depends on the
bearing clearance, changes in the gear ratio can be a direct measure
of wear in the motor bearing.
Conventional accelerated friction and wear test apparatuses
have been used on a coupon level to measure friction and wear of
various thin films. To measure static friction, Deng et al.[93]–[95] used
a measurement technique shown in Fig. 33, with two millimeter-
sized flat coupons in controlled environment. A polyvinylidene
difluoride (PVDF) bimorph cantilever was used to generate a repeat-
able tangential force from 0 to 0.8 mN. The normal force was applied
by electrostatic attraction through a bias voltage, and it was in the
range of 1 mN. A capacitance-voltage (C-V) meter that supplied the
clamping voltage to the samples was also used to detect the displace-
ment of the samples. When the samples slid over each other, the
charge equilibrium at the interface was disturbed, and the air gap
between the two surfaces was changed. The change in the charge
corresponded to the displacement of the sample. During an experi-
ment, the electrostatic clamping force was applied using the bias
voltage of a high frequency CV meter, and the capacitance was
measured simultaneously. Next, an adjustable DC voltage was ap-
plied to the PVDF bimorph cantilever to generate the tangential force
that pushed the sample forward until a displacement was observed as
a sudden change of capacitance indicated by the CV meter. The
coefficient of static friction was then obtained from a plot of tangen-

tial force as a function of normal force. Figure 34 shows the tangen-
tial force as a function of normal force for the sputtered SiO2 coating
from a fused silica target against itself in different environments. The
tangential force versus normal force plots for the SiO2/ SiO2 pair in
air before baking is 0.54, in air after baking is 0.21, and in high
vacuum of 10-5 torr is 0.36.[95]
Figure 33. Schematic of the setup used to measure static friction between two
flats samples at very low loads.[95]
Figure 34. Tangential force versus normal force plots for SiO2/ SiO 2, pair in
ambient air as well as in high vacuum (10-5 torr). The coefficient of friction is
0.54 before baking the sample and 0.21 after. The coefficient of friction is 0.36 at
10-5 torr after the sample was baked.[95]
Deng et al.[93][95] developed another setup to measure static

friction between a flat coupon, and a three-millimeter radius alumi-
num bullet coated with various materials in controlled environment,
Fig. 35. The normal force on the bullet was applied by placing a 10 N
weight rotating on a threaded rod. The arm holding the bullet was
translated horizontally by a translation table using a piezoelectric
positioning device. The push of the piezoelectric positioning device
was controlled by a high voltage power supply, supervised by a
programmable controller. The tangential force on a bullet was mea-
sured by a charge mode piezoelectric force sensor. The sliding test
was performed by lowering the bullet onto the substrate, adjusting
the normal force on the bullet, and pushing the bullet forward using
the translational device until slippage was observed on the force-
displacement curve. The bullet was then raised. Deng and Ko[95]
used this setup to measure friction between various material/ coating
pairs in ambient air, in UHV (8 × 10-10 torr), argon, and oxygen
environments. The coefficients of friction obtained by using this
setup are compared in Table 4. The values of coefficients of friction
obtained by using this setup were found to be very close to those
values obtained from in situ friction measurements in micromotors.[95]
Figure 35. Schematic of the setup used to measure static friction between a
coated bullet and a flat in controlled environment.[95]
Table 4. Measurement Results from the Setup Shown in Fig. 35.

SiNx: PECVD Silicon Nitride[95]
Noguchi et al.[212] measured static and kinetic friction between

a millimeter-sized flat, and coated coupons using an experimental
setup shown in Fig. 36. The millimeter sized flat mounted on a
mover was driven electrostatically on the surface of a fixed flat to
measure static friction. The gap between the mover and the fixed
plate was 0.5 mm. The driving electrode was attached to the backside
of the fixed plate. The AC voltage was applied between the mover and
the driving electrode. The static friction was calculated from the mini-
mum voltage required to start the flat moving. The kinetic friction was
measured from the transient measurement of the flat traveling across
the gap. When a voltage was applied between the flat and the driving
electrode, the flat moved toward the insulating glass pulled by the
electrostatic force. The transit time during which the mover crosses the
gap was measured using a timer, a video tape recorder, and a monitor.
(a)
(b)
Figure 36. Schematic of the setup used to measure static and kinetic friction
between two flat samples.[212]
For friction and wear studies, Suzuki et al.[268] used a magnetic

disk drive with a cylindrical rider sliding against a disk surface (see
Ref. 38). They mounted the rider on a strain-gauged I beam to
measure coefficient of static and kinetic friction. Beerschwinger et
al.[28] conducted wear studies by sliding a micromachined sample
against a flat disk in a conventional pin-on-disk arrangement. To
better simulate MEMS interactions, they used specially fabricated
microstructures sliding against a disk surface. Microstructures with
different apparent area of contact and heights were micromachined

on single-crystal silicon and polycrystalline silicon by using photoli-
thography. The SEM images of micromachined surfaces, sample
holder with circularly arranged holes for the placement of triangles,
and pin-on-disk type of experimental setup for sliding tests are
shown in Fig. 37.[28] Different geometry, and size of micromachined
surface gives different contact pressures ranging from 50 to 1700 Pa.
The wear rate was calculated from the changes in the dimensions of
a microstructure on a sample measured from the surface profiles
before and after the wear tests. The wear rate of LPCVD diamondlike
carbon (DLC), PECVD silicon nitride (SiN), silicon dioxide (SiO)
by wet hydrogen at 1050°C; and p-type (100) single-crystalline
silicon (SCS) microstructures sliding against a DLC coated silicon
substrate as a function of sliding distance at a contact pressure of 170
Pa, and a normal load of 30 µN, are compared in Fig. 38. The DLC
coated microstructure exhibits the lowest wear rates.
4.0 MECHANICAL AND TRIBOLOGICAL PROPERTIES

OF DIAMOND AND AMORPHOUS CARBON
COATINGS
Mechanical, friction and wear properties of hard coatings vary

over a wide range with their physical and metallurgical properties,
which are strongly influenced by the deposition conditions and the
interfacial chemistry. Among the most widely used hard coatings are
polycrystalline diamond, amorphous carbon nitrides, and carbides of
various metals. Lately, more attention has been paid to boron nitride,
and carbon-nitrogen compound coatings because of their outstand-
ing hardness (> 50 GPa). Coatings of amorphous carbon with and
without hydrogen have been produced by numerous researchers to
protect surfaces of mechanical components such as magnetic thin-
film rigid disk and recording heads, and MEMS devices against
wear. In this section, we present the mechanical and tribological
properties of polycrystalline diamond, and amorphous carbon

coatings in detail because of their potential applications in various
devices. For completeness, a brief overview of various CVD, and
PVD techniques used to deposit polycrystalline diamond and amor-
phous carbon coatings is also given.
Figure 37. (a) SEM images of triangles micromachined on silicon by photoli-

thography; total area of contact of triangle A and B is 1.88 × 10-1 mm-2 and 2.36
× 10 -2 mm -2, respectively. (b) Sample holder with circular arranged holes for the
placement of test triangles. (c) Schematic of the set-up used for sliding triangles
on a rotating disk. [28]
Figure 38. Comparison of wear rates of various coatings pairs at a contact

pressure of 170 Pa and a normal force of 30 µN obtained by using the set-up
shown in Fig. 37.[28]
4.1 Deposition Processes
Polycrystalline diamond, and amorphous carbon coatings have

been deposited by various deposition conditions, and have been
extensively studied by numerous researchers for their applications in
tribology, optics, and electronics. Depending on the deposition con-
ditions, one can synthesize coatings of almost pure diamond with
very high hardness, as well as very soft coatings of carbon poly-
mers.[12]–[14][46][86][176][189][238][239][241][256][257][284][286] The sp3-bond-
ing in amorphous carbon results in many of the outstanding proper-
ties of diamond such as high hardness, elastic modulus, low friction,
optical transparency, and chemical inertness. Atomic hydrogen is
believed to be responsible for formation of sp3-bonding.[12][14]
A wide variety of CVD techniques have been used to synthe-
size pure diamond films at low pressures on a variety of non-
diamond substrates since the breakthrough of diamond deposition on
nondiamond substrates.[96][154][259] These techniques have been re-
viewed from time to time by many researchers. [10][11][13][20]
[77][97][98][148][159][208][256] [260][310][311] Not much success has been
achieved in synthesizing diamond coatings by PVD. In general,

CVD polycrystalline diamond coatings are deposited from a carbon-
containing precursor gas, and formation of nondiamond carbon is

prevented by the presence of a selective enchant, such as atomic
hydrogen. To obtain high deposition rates, energetic and efficient
decomposition of precursor gas and diamond deposition at lower
substrate temperatures; CVD techniques have been assisted by a
wide variety of thermal-hot filament,[175][187] electrons,[247] laser,[158]
plasma-DC/RF discharge,[266] DC arc jet plasma,[4] DC plasma
jet, [113][165][214] RF thermal plasma, [162][186][188] microwave
plasma,[154] electron cyclotron resonance,[103][150][155][267] microwave
plasma torch,[198] and combustion flame processes.[141] In plasma-
assisted CVD techniques, precursor gases are activated by plasma
generated by electrical discharges at various frequencies or induc-
tion heating. The role of plasma is to produce atomic hydrogen and
proper carbon precursor for the growth of diamond film. In plasma
processes, the atomic hydrogen is produced by electron impact
dissociation of molecular hydrogen, and its kinetic energy depends
on the impact energy of electrons. The kinetic energy of atomic
hydrogen produced by plasma is much higher than that produced by
thermal cracking.
These assisted CVD techniques, although different in their
deposition reactions and deposition configurations, have a few com-
mon features: the presence of a gas-phase nonequilibrium in the
immediate vicinity of the substrate generated by some sort of gas-
phase activation, and growth of diamond on elevated substrate tem-
peratures of 500 to 1200°C. The deposition rates of diamond coat-
ings deposited by various CVD techniques have been observed to vary
from 0.1 µm·h-1 to approximately 1 mm/h. Usually higher growth
rates are associated with higher pressure processes such as, combus-
tion and thermal plasma, which run at pressures up to 1 atm. It has
been reported that DC plasma jet CVD process resulted in the highest
deposition rate of 930 µm·h-1, whereas the microwave-assisted CVD
resulted in high quality diamond films on non-diamond substrates such
as Si, Mo, and WC over an area of 40 cm2. However, deposition rates
in the microwave-assisted CVD were low (0.1 µm·h-1).
The nucleation density, growth mechanisms, and quality of
diamond films (characterized by sp3-bonding fraction), grain size,
and presence of nondiamond component are strongly influenced by
numerous deposition parameters in a very complex manner. It is

difficult to set up any trend towards the dependence of deposition
parameters on the properties of diamond films. The parameters of
particular importance in CVD depositions are: appropriate com-
bination of carbon, hydrogen and/or oxygen in precursor gas mix-
ture, substrate temperature, total pressure of precursor gas mixture,
plasma density, and substrate biasing. The substrate temperature
from 500 to 1200°C is essential for diamond deposition. Films
deposited above 1200°C are found to be graphitic, whereas films
deposited below 500°C are amorphous. The substrate temperature is
also found to affect the deposition rates,[258][260] and crystallite orien-
tation of diamond grains. The total pressure of hydrocarbon-hydro-
gen gas mixture determines the recombination length, the life time,
and the drift distance of atomic hydrogen.[17] The precursor gas
temperature affects the deposition rates, higher gas temperatures
yield higher deposition rates. In the low-pressure plasma-assisted
CVD, the deposition rates were found to increase linearly with the
plasma density.[292] The substrate biasing which affects the energy of
electron and/or ion bombardment on the growing film controls the
kinetics of growth, nucleation density, which in turn affects diamond
quality, and physical properties of the films.[205][247][264]
Diamond films synthesized by different CVD techniques can
range from rough well-faceted films with a crystallite size of a couple of
microns, to optically smooth films with a crystallite size of a couple of
nanometers. Thermochemical, chemomechanical, and plasma/ion beam/
laser techniques have been explored to polish diamond films.[1][48][54][61]–
[63][110][126][140][250][280][281][286][314] A summary of these polishing tech-
niques is given in Table 5.[54] Thermochemical and chemomechanical

techniques are contact techniques, so these can be used for surfaces that
are planar. Plasma, ion beam, and laser techniques do not require bulk
sample heating, and are noncontact techniques that can be used on
nonplanar surfaces, although these are line-of-sight processes. The
material removal rates using these techniques have thus far been
small. High-energy ion beams have the potential for high material
removal rates. However, heat generated during sputtering results in
graphitization of the diamond film, so the laser technique emerges as
the most desirable among the noncontact techniques.
Table 5. Summary of Various Polishing Techniques[54]

Like diamond deposition, amorphous carbon coatings have

also been deposited CVD techniques by using a hydrocarbon gas
mixed with hydrogen as a starting materials, and the gas mixture is
activated by similar thermal or plasma activation processes. RF glow
discharge-assisted CVD followed by microwave discharge-assisted
CVD are the most widely used for deposition of dense hydrogenated
carbon coatings.[32][68][78][160][186][200][215] Depending on the deposi-
tion technique, a-C or a-C:H films can have carbon phases of differ-
ent chemical composition (%H), atomic arrangement (especially the
coordination of the nearest neighbors and the degree and scale of order),
and electronic structure (binding type). It is believed that deposition
of sp3-bonded carbon requires the depositing species to have kinetic
energies on the order of 100 eV or higher, well above those obtained
in thermal processes like evaporation (~ 0.1 eV). The species must
then be quenched into the metastable configuration via rapid energy
removal. Excess energy, such as that provided by substrate heating,
is detrimental to the achievement of a high sp3-fraction.
In PVD techniques, amorphous carbon coatings are deposited
by thermal evaporation, DC/RF glow discharge sputtering, or single/
dual ion beams sputtering,[3][309] ion plating,[296] direct ion beam
(DIB),[106][107][271][293] and vaporizing by arc discharge, [7]–[9][67]
[91][181] or a laser source (such as CO laser),[112] in argon or argon
2
and hydrogen gas mixture, and the growing carbon film is bom-
barded with energetic ions of 20–500 eV energy, which is believed to
be necessary to promote sp3-bonding. To achieve good adhesion of
a-C and a-C:H films on nondiamond substrates, and materials that do
not readily form carbides, thin silicon interlayers of a couple of
nanometers are commonly applied.[121][122][125]
The common feature of various PVD techniques is that the
carbon species arrive with an energy significantly greater than that
represented by the substrate temperature. Generically, high kinetic
energy of carbon species, which depends on the deposition configu-
ration and substrate biasing, results in higher degree of sp3-bonding,
and improved adhesion and mechanical properties of the coatings.
The extent to which the properties of a film are affected by ion
bombardment is dependent on both the kinetic energy, and the flux

density of bombarding ions, as well as the ions to be deposited atom
ratio to the mass ratio of the ions relative to the surface atoms. Ion-
surface interactions during deposition include several processes:
increased adatom migration, desorption of adsorbed surface impuri-
ties, displacement of atoms in the near surface region, recoil implan-
tation, sputtering, re-sputtering of deposited species, working gas
entrapment. The severity of these interactions depends on the ion
energy and dose which in turn affect the film density, stoichiometry,
growth kinetics (rates), residual stresses, hardness, and other me-
chanical properties.[237][276]
The structure and properties of a-C or a-C:H films are strongly
influenced by the deposition conditions, deposition configurations,
and the constituents of precursor gas in a very complex manner. The
energy of carbon species which depends on the kinetic energy
involved in a particular PVD technique, and substrate biasing in the
CVD techniques, is one of the most sensitive deposition parameters.
The films deposited at low kinetic energy of a few tens of electronvolts
by ion plating technique using benzene were found to be polymer
like; their physical properties and IR spectra resemble those of
polystyrene.[296] The formation of amorphous carbon occurred over
an energy range of several hundred electronvolts. The features of
carbon coatings deposited at different kinetic energies by ion plating
using benzene are shown in Fig. 39.[296] The concentration of hydro-
gen was found to be decreasing exponentially up to an ion energy of
1 keV. The films deposited at 150 eV energy did not exhibit any C-O
bonds, even when the film was exposed to air. Furthermore, some
graphite traces were found in the films deposited at energies over 1
keV. Weissmantel et al.[296] believe that secondary radiation damage
due to impinging energetic ions is responsible for the formation of
graphitic carbon. The films were nonuniform in thickness when
kinetic energy exceeded 3 keV. This study suggests that the proper-
ties of amorphous carbon coatings are very sensitive to the kinetic
energy of the carbon species which can be manipulated through
substrate biasing, ion beam energy, and laser irradiation.
Figure 39. General properties of amorphous carbon coatings deposited at

various acceleration voltages (or kinetic energies) by ion plating from ben-
zene.[296]
The sp3-bonded carbon fraction in a-C:H films deposited by

RF capacitively coupled plasma-assisted CVD, using methane-based
precursor gas as a function of substrate biasing is shown in Fig. 40.[277]
The sp2-bonded carbon fraction predominates at the highest biasing
where the structure is highly ordered graphitic.
Figure 40. Effect of substrate biasing on the sp3-bonded carbon fraction of

diamond films deposited by RF capacitively coupled plasma-assisted CVD using
methane based precursor gas.[277]
The influence of composition of the precursor gas (Ar+CH4

and H2+CH4) on the film density and hydrogen content of a-C:H
films deposited by MW/RF PACVD is shown in Fig. 41. We note
that the density profile as a function of Ar concentration passes
through a peak (1.57 g cm-3 ) at about 20–40 vol.% Ar. The
hydrogen content increases with H2 concentration in MW/RF
PACVD a-C:H films, whereas it decreases with H2 concentration in
RF PCVD a-C:H films.[236]
Figure 41. Effect of composition of precursor gas on the film density of diamond
films deposited by MW/RF plasma-assisted CVD using methane mixed with
argon and hydrogen.[236]
The dependence of ion energy on the film density of a-C films

deposited by ion beam deposition is illustrated in Fig. 42.[146][193][240]
McKenzie et al.[193] observed that the density profile of a-C films as
a function of ion energy passes through a peak (highest density
~3.15 g cm-3) at 30 eV. Ishikawa et al.[146] also found a peak in
density profile and other properties with ion energy but at much
higher ion energy of 150–300 eV, Fig. 42.
Figure 42. Effect of ion energy on the film density of amorphous carbon
films deposited by ion beam deposition with e/m selection.[146][193][240]
The dependence of substrate biasing on the film density and

hydrogen content of a-C:H films deposited by RF capacitively coupled
plasma-assisted CVD using benzene-based precursor gas,[163] meth-
ane-based precursor gas,[277] and hydrocarbon vapor, [300] and RF
inductively coupled plasma-assisted CVD using methane-based pre-
cursor gas[222] is shown in Fig. 43. We note that in most of the cases,
film density increases and hydrogen content decreases linearly with
the substrate biasing. However, the density profile as a function of
substrate biasing of a-C:H films deposited by methane-based precur-
sor gas[277] passes through a peak at 200–400 V, similar to the
variation observed for ion beam deposited a-C coatings, Fig. 43.
Figure 43. (a) Effect of substrate biasing on the film density of amorphous
carbon films deposited by rf capacitively coupled plasma-assisted CVD using
benzene-based precursor gas,[163] methane based precursor gas,[207] and hydro-
carbon vapor,[300] and (b) effect of substrate biasing on hydrogen content of
amorphous carbon films deposited by rf capacitively coupled plasma-assisted
CVD using benzene-based precursor gas,[163] methane based precursor gas,[207]
and hydrocarbon vapor;[300] and by rf inductively coupled plasma-assisted CVD
using methane based precursor gas.[222]
4.2 Mechanical Properties
The mechanical properties of carbon coatings depend on their

physical properties, growth conditions, sp3- sp2-bonded carbon ratio,
the amount of hydrogen in the coating, and adhesion of the coating
to the substrate, which are influenced by the precursor material,
kinetic energy of the carbon species prior to deposition, deposition
rate, substrate temperature, and substrate biasing.[11][47][83][89]–
[91][221][236][238][243][245][246][248][265][283][284][289] In this section, we
present some representative results on the mechanical and tribologi-

cal properties on the micro- and nanoscale, of diamond, a-C, and a-
C:H coatings reported by various authors.
Hardness And Young’s Modulus of Elasticity. High cohe-
sive energy, short bond length and covalent bonding in diamond
result in the highest hardness and Young’s modulus of elasticity. In
polycrystalline diamond films, even in the presence of high strength
of interatomic forces, the mechanical properties significantly vary
with the operative deformation mechanisms, which vary with the
microstructure and other structural features like voids, impurities,
defects, and texture.[265] The micromechanical characterization stud-
ies of a-C and a-C:H films deposited by various PVD and CVD
techniques have been done by various research groups. [56][83]
[116][132][133][246] Since these coatings are much smoother than poly-
crystalline diamond films, it is comparatively easy to obtain repro-

ducible hardness, and elastic modulus values by nanoindentation. In
this section, we focus on the micromechanical properties of diamond
and a-C and a-C:H coatings. The information on the hardness and
elastic modulus measured by conventional methods can be found in
various reviews.[236][283][284]
Very few studies have been conducted on the micromechanical
properties measurements of diamond films. Beetz et al.[29] and O’Hern
and McHargue[213] have made nanoindentation measurements on
polycrystalline diamond films deposited on silicon by HFCVD using
methane and hydrogen mixture precursor gas. Beetz et al.[29] depos-
ited films at two methane concentrations (0.11 and 0.99% CH4 in
H2), 1 kPa partial pressure, and at about 800–900°C substrate tem-
perature. The films deposited at lower methane concentration exhibit
larger size grain size of about 5–8 µm; and the films deposited at
higher methane concentration exhibit smaller size grain size of
about 0.5 µm. The load-displacement plots of nanoindentation, and
SEM images of two polycrystalline diamond films deposited by
Beetz et al. [29] are shown in Fig. 44.[29] We note that the coarse-
grained films exhibit almost an elastic deformation, whereas
the fine-grained diamond films exhibit a little hysteresis in the
unloading curve, indicating some amount of plastic deformation, the

consequence of which is a lower hardness. Beetz et al.[29] reported
that the higher methane concentration in the precursor gas led to a
decrease in grain size, and incorporation of sp2-bonded carbon,
indicated by Raman Studies, which results into a decrease in hard-
ness of the diamond films. O’Hern and McHargue[213] deposited
films at a methane concentration of 1% CH4 in H2 and a substrate
temperature of about 1100°C. The load-displacement plots of a
HFCVD polycrystalline diamond film and type II a natural diamond
are compared in Fig. 45.[213] We note that loading and unloading
curves of natural diamond overlap each other indicating a perfect
elastic deformation during nanoindentation. However, the unloading
curve of the diamond film shows a little plastic deformation.
(a)
(b)
Figure 44. Load-displacement plots of nanoindentation and SEM images of

HFCVD diamond films deposited by HFCVD using (a) 0.11% CH4 in H 2 gas and
(b) 0.99% CH4 in H 2 gas.[29]
Figure 45. Comparison of load-displacement plots of nanoindentations made on

a type II a natural diamond and a HFCVD diamond film.[213]
The load-displacement plots of three nanoindentations and the

SEM image of a 3 µm thick polycrystalline diamond film having
average grain size of about 2–3 µm, deposited on (100) silicon by
MPCVD using 1% methane in a H2 mixture, are shown in Fig.
46.[246] The loading and unloading curves of these indentations show
a very little hystereses during unloading consistent with the observa-
tions of Beetz et al.[29] and O’Hern and McHargue.[213] Savvides and
Bell[246] made indentations at two different indentation loads of 67
and 100 mN (corresponding to 0.06 and 1.07 mN contact force,
respectively), and obtained hardnesses and elastic moduli of 80 and
533 GPa and 56–102 and 1050 GPa, respectively. The hardness and
elastic modulus values of polycrystalline diamond films are quite
close to those of natural diamond which varies from 80 GPa for
polycrystalline diamond to 104 GPa for single-crystal diamond.[66]

It is important to mention that since in nanoindentation the depths of
indentation are very shallow, rough surface morphology with large
grain size result in a wide scatter in values; thus for accurate and
reproducible measurements, rough polycrystalline films need to be
polished.
Figure 46. Load-displacement plots of nanoindentations and SEM images of

HFCVD diamond films deposited by MPCVD using 1% CH4 in H2 gas.[246]
Seino et al.[251] and Seino and Nagai[252] deposited polycrystal-

line diamond films by MPCVD with 0.5–1 micron grain size at 4 kPa
partial pressure, 0.1 to 5% methane in a H2 mixture, and 500 W
microwave power, and measured the hardness by Vickers indenta-
tion at 2–10 mN, and elastic modulus of a free standing film by a
vibrating reed apparatus. The effect of methane concentration in
precursor gas on the hardness and elastic modulus is shown in Fig.
47.[251] We note that films deposited at lower methane concentration
of 0.1% exhibit the highest hardness of about 94 GPa and elastic
modulus of about 960 GPa and hardness and elastic modulus de-
crease with increasing methane concentration resulting in an in-
crease in sp2-bonded carbon fraction in diamond films. The high
hardness, and elastic modulus values of MPCVD diamond films
deposited at lower methane concentration are very close to those of
natural diamond ranging from 80 to 100 GPa.[66]
Figure 47. Effect of methane concentration on the hardness and elastic modulus
of diamond films deposited by MPCVD using 0.1 to 5% CH4 in H2 gas. [251]
The data reported by various authors indicate that the polycrys-

talline diamond films exhibit hardness and elastic modulus values
very close to those of natural diamond. More systematic study on
micromechanical characterization is needed to better understand the
dependence of deposition parameters, effects of diamond crystal-
lite orientation, surface morphology, grain size, and role of atomic
hydrogen in the growth process on the mechanical behavior of
diamond films.
The load-displacement plots of three indentations made on a 5-µm

thick a-C film deposited on silicon by ion-assisted magnetron sput-
tering are shown in Fig. 48.[246] Compared to polycrystalline dia-
mond films (Fig. 46), these films exhibit much higher indentation
depths because of their lower hardness. The dependence of ion
energy of carbon species in magnetron sputtering, which conse-
quently affect the sp3-to-sp2-bonded carbon ratio on the hardness is
shown in Fig. 49. We note that the hardness decreases with increas-
ing ion energy and sp3-to-sp 2-bonded carbon ratio.
Figure 48. Load-displacement plots of three nanoindentations made on 5-µm

thick amorphous carbon films deposited by low-energy ion-beam assisted
magnetron sputtering.[246]
Anders et al.[9] deposited hydrogen free amorphous carbon

films by cathodic arc deposition technique by changing substrate
pulsed bias voltage, and duty bias cycle %. The variation in hardness
with pulsed bias voltage, and pulsed bias duty cycle is shown in Fig.
50. We note that the hardness profile as a function of pulsed bias
voltage attains a maxima of 35 GPa at about -100 V and as a function
bias duty cycle % attains a maxima of 45 GPa at 55%. The authors
believe that the deposition conditions resulting in higher hardness
probably formed increased fraction of sp3-bonded carbon.
Figure 49. The dependence of hardness of amorphous carbon films on ion

energy per depositing carbon atom and the sp3/sp2 bonding ratio. The films were
deposited by low-energy ion-beam assisted magnetron sputtering.[246]
Cho et al.[83] deposited a-C:H films with varying hydrogen

content by magnetron sputtering of a graphite target at a pressure of
10 mtorr by changing the power density of hydrogen-argon plasma
from 0.1 to 10 watt cm-2. The variation in the hardness, measured by
Nanoindenter, with hydrogen content in the film is shown in Fig. 51.
We note that the hardness decreases from 14 to 3 GPa as hydrogen
content increases from about 0.4 to 6%. Cho et al.[83] found from
Raman and electron energy loss spectrometry (EELS) measurements
that hydrogen incorporation in the a-C:H films enhances the crystal-
linity of the sp2-bonded carbon clusters, as well as increases the sp2-
bonded carbon fraction. The lower hardness at higher hydrogen
content is attributed to the higher sp2-bonded carbon fraction in the
a-C:H films.
Figure 50. The dependence of hardness of amorphous carbon films deposited by

cathodic arc deposition (a) on substrate pulsed bias voltage for 10 and 33% and
pulsed duty cycles, and (b) on pulsed bias duty cycle for pulsed and dc bias
voltages of -100.[9]
Figure 51. Effect of hydrogen content of the hardness of a-C:H films deposited
by magnetron sputtering using a graphite target and by varying hydrogen concen-
tration from 0.5 to 15 % in Ar + H2 gas mixture.[83]
The variations in hardnesses with substrate biasing of a-C:H

films deposited by RF capacitively coupled plasma-assisted CVD
under different deposition conditions: (i) methane-based precursor
gas biasing from -100 to -1250 V,[116] (ii) 75% CH4 and 25% Ar,
biasing from -50 to -600 V, substrate temperature 200°C,[92] and (iii)
benzene-based precursor gas at 3 Pa partial pressure[163] are com-
pared in Fig. 52. We note that films deposited from methane-based
precursor gas exhibit a peak maxima in the hardness profile at about
-200 V, whereas the films deposited from benzene-based precursor
exhibit a continuing increase in hardness with substrate biasing
voltage. Furthermore, low biasing voltage of about -100 V resulted
in soft polymeric films and high biasing voltage of about -600 V
resulted in the graphitic films.[116]
Figure 52. The effect of substrate bias voltage on the hardness of amorphous
carbon films deposited by rf capacitively coupled plasma-assisted CVD using
methane based precursor gas and biasing from -100 to -1250 V,[116] and 75% CH4
and 25% Ar precursor gas and biasing from -50 to -600 V,[92] and benzene-based
precursor gas at 3 kPa partial pressure.[163]
Pappas and Hopwood[222] studied the effect of various deposi-

tion parameters on the hardness, measured by Knoop indentation, of
a-C:H films deposited on silicon by RF induction plasma-assisted
CVD. Shown in Fig. 53 are the effects of substrate biasing, RF
induction power, Ar concentration in methane-based precursor gas,
and precursor gas flow on the hardness of a-C:H films. Like RF
capacitively coupled plasma-assisted CVD deposited a-C:H films,

the hardness profile of these films as a function of substrate biasing
also passes through a maxima of about 30 GPa peak at -100 V. The
hardness decreases with RF induction power and methane gas flow
and increases with argon concentration in methane-based precursor.
The variation in hardness with deposition conditions can be related
to the variations in sp3-to-sp2-bonded carbon ratio resulting from
changes in the deposition conditions.
(a)
(b)
(c)
(d)
Figure 53. The dependence of hardness of amorphous carbon films on various

deposition parameters (a) substrate biasing, (b) rf induction power, (c) argon
percentage in methane, and (d) methane gas flow rate. These films were depos-
ited by rf inductively coupled plasma-assisted CVD technique.[222]
The hardness and elastic modulus of metal-containing amor-

phous carbon films (Me-C:H) were measured using Nanoindenter by
Wang et al.[291] Figure 54 shows the dependence of metal concentra-

tions on the hardness, and elastic modulus of Ti-C:H films deposited
on silicon by DC magnetron sputtering and Ta-C:H films deposited
by combining RF capacitively coupled plasma-assisted CVD at three
substrate temperatures, i.e., 20, 270, and 20–270°C in the presence
of methane-based plasma. Both the hardness and elastic modulus
profiles as a function of metal concentration in films passes through
a peak at about 50 to 60% metal. A model for explaining the increase
in hardness and elastic modulus of Me-C:H with metal concentration
need to be developed.
Figure 54. The dependence of hardness and elastic modulus Me-C:H films
deposited by rf plasma CVD combined with reactive rf glow discharge sputtering
using argon+acetylene precursor gas at different percentages of acetylene and
substrate temperatures. Films shown by S1 were deposited at 270°C, S2 at 20°C,
and S3 at temperatures between 20 to 270°C.[291]
To study the effect of deposition technique on the mechanical

properties of amorphous carbon coating, Gupta and Bhushan[132][133]
deposited 20- to 400-nm thick amorphous carbon coatings on silicon
by four deposition processes: (i) cathodic arc, (ii) ion beam deposi-
tion, (iii) RF-plasma enhanced chemical vapor deposition (RF-
PECVD), and (iv) RF sputtering. Hardness and elastic modulus
profiles, measured by Nanoindenter, as a function of depth of carbon
coatings deposited by various techniques are compared in Fig. 55.
Cathodic arc carbon coatings exhibit the highest hardness of 38 GPa,
and elastic modulus of 350 GPa of all coatings examined in this
study. The high hardness and elastic modulus of cathodic arc carbon
coatings are followed by ion beam carbon and PECVD/ sputtered
carbon coatings. The high hardness and elastic modulus of cathodic
arc carbon are attributed to the high kinetic energy of carbon species
involved in the cathodic arc deposition. The difference in hardness
and elastic modulus of carbon coatings deposited by various deposi-
tion techniques is attributed to their varying sp3- to sp2-bonding
ratio, and the amount of hydrogen. The reduction in hardness, and
elastic modulus of cathodic arc carbon coatings with the indentation
depth is attributed to the increased contributions of the silicon
substrate (having a lower hardness of 11 GPa) at larger depths.
The hardness and elastic modulus data on a-C and a-C:H films
reported by several research groups suggest that mechanical behav-
ior of these films is very sensitive to sp3- to sp2-bonding ratio and
hydrogen content which can be optimized by changing the deposition.
Adhesion. Usage of the diamond, and amorphous carbon
coatings to protect against wear and corrosion, and to serve as a solid
lubricant coating for reduced friction, requires high adhesion of these
coatings. The adhesion forces between the coating and the substrate
must be sufficiently high so that the forces due to residual stresses
can be neutralized. Diamond coatings are found to adhere well with
materials that have a capability of carbide and silicide formation,
such as, Si, SiC, Si3N4, quartz, and iron alloys.[11][105][296] On the
contrary, almost no adhesion of PECVD deposited amorphous car-
bon coatings was observed on some materials like Co, Cr, Ni, Al2O3,
and ZrO2 and their alloys.[105][121][143] In general, high adhesion of a

coating can be achieved by increasing the nucleation density, form-
ing a compositionally graded junction between the coating and
substrate, and applying a thin interlayer.
Figure 55. The hardness and elastic modulus, measured by nanoindentation, of

400-nm thick amorphous carbon films deposited on silicon by cathodic arc,
direct ion beam, RF plasma-enhanced CVD, and dc magnetron sputtering.[132]
A variety of surface pretreatments to achieve a high nucleation

density which in turn enhances adhesion of diamond coatings for
various substrates have been discussed in an earlier section. In the
case of thin amorphous carbon coatings (~ 20 nm thick), a post
deposition treatment by ion bombardment of inert gas species result-
ing into ion beam mixing at the interface can also be used to enhance
the adhesion. Grill et al.[121][122][125] studied the effect of silicon and
metallic interlayers on the adhesion of a-C:H coatings deposited by
RF plasma CVD, and concluded that a 2 to 4 nm thick film of silicide
forming metal like Si is very effective in improving the adhesion of a
carbon coating on nondiamond substrate.
To study the effect of deposition techniques on the adhesion of

amorphous carbon coating, Gupta and Bhushan[132] deposited 20-nm
thick amorphous carbon coatings on silicon by four deposition pro-
cesses: (i) cathodic arc, (ii) ion beam deposition, (iii) RF-plasma
enhanced chemical vapor deposition (RF-PECVD), and (iv) RF
sputtering. The adhesion of these coatings was compared by scratch-
ing at ramping loads ranging from 1 to 15 mN using Nanoindenter,
and monitoring the friction and scratch depth while scratching.[132]
The coefficient of friction and scratch depth profiles as a function of
increasing normal load and SEM images of two regions over scratches:
at the beginning of the scratch (indicated by ‘A’ on friction profile),
and at the point of initiation of damage at which the coefficient of
friction increases abruptly to a very high value (indicated by ‘B’ on
friction profile), made on various 20-nm thick carbon and SiC
coatings and uncoated substrate are shown in Fig. 56. The abrupt
increase in the coefficient of friction and scratch depth is associated
with damage to the coating. Most of the coatings exhibit a steady low
coefficient of friction of about 0.1–0.15 during the initial stages and
an abrupt increase in friction when normal load exceeds the critical
load, the load sufficient to damage the coating. Only sputtered
carbon coating exhibits a gradual increase in the coefficient of
friction with increasing normal load from the beginning of the
scratch. In a separate study, Bhushan and Gupta[55] have shown that a
steady low coefficient of friction during the initial stages of a scratch
is associated with the sliding of the tip without significantly plough-
ing into the sample, and abrupt increase in friction is associated with
catastrophic failure, as well as significant ploughing of the tip into
the sample. The gradual increase in friction with normal load during
scratching is associated with the ploughing of the tip into the sample;
and the depth of ploughing increases with increasing normal load.[55]
The friction profiles shown in Fig. 56 suggests that most of the
coatings except sputtered carbon coating did not get damaged before
the critical load, and got damaged catastrophically as the normal load
exceeded the critical load limit. In contrast, the sputtered carbon
coatings were damaged right from the beginning of the scratch.
Figure 56. The adhesion, measured by microscratching, of 20-nm thick amor-

phous carbon films deposited on silicon by cathodic arc, direct ion beam, rf
plasma-enhanced CVD, and DC magnetron sputtering.[132]
The scratch depth profiles obtained during and after the scratches
on all samples are plotted with respect to initial profile after cylindri-
cal curvature removed, Fig. 56. Reduction in scratch depth after
scratching is observed. The reduction in scratch depth after scratch-
ing is attributed to an elastic recovery after the removal of the normal
load. It appears that the scratch depth after scratching indicates the
final depths which reflect the extent of permanent damage and
ploughing of the tip into the sample surface. We believe that the
scratch depths after scratching are probably more relevant for visual-
izing the damage that can occur in real applications.
The higher scratch resistance/adhesion of cathodic arc carbon
coatings is attributed to an atomic intermixing at the coating-sub-
strate interface because of high kinetic energy (2 keV) plasma
formed during the cathodic arc deposition process.[9] The atomic
intermixing at the interface provides a graded compositional transi-
tion between the coating and the substrate materials. In all other
coatings used in this study, the kinetic energy of the plasma was
insufficient for atomic intermixing.
4.3 Tribological Properties
In this section, we present the friction and wear characteristics

of a few selected diamond and amorphous carbon coatings. The
friction and wear properties of coating are strongly influenced by
their physical properties and the chemistry in the interface of the
mating surfaces. In many cases environment plays a crucial role in
affecting the friction and wear characteristics of a material. Friction
and wear properties of polycrystalline diamond films are found to be
strongly influenced by their surface morphology, grain size of crys-
tallite size, and amount of nondiamond carbon phase in the diamond
matrix.[21][115][117][118][129][161][164][226] Surface roughness plays an
important role in friction and wear properties of CVD diamond
films.[49][54][71][129][137][138][202][203] In this section, we will show how
surface morphology and surface roughness affect the friction of
diamond films, and the wear of the counterface materials.
Miyoshi et al.[203] studied friction and wear characteristics of
diamond films of different surface morphologies deposited on sili-
con, and a-SiC by high-pressure MPCVD at varying partial pres-
sures ranging from 5 to 40 torr, substrate temperatures ranging from
850–1000°C, microwave power ranging from 500 to 1000 watts, and
deposition times ranging from 10 to 21 h. Shown in Fig. 57 are the
SEM images of these films indicating that the grain size of diamond
crystallites varies from 20–100 nm to 3.3 mm. The fine-grained films
(grain size ~ 20–100 nm) are smooth, and contain a significant
amount of nondiamond carbon, whereas the coarse-grained films
(grain size ~ 1.1–3.3 mm) are rough and contain a little amount of
nondiamond carbon. The coefficients of friction of these films against
bulk diamond pins at 1N in humid air and dry nitrogen are shown in
Fig. 58. As expected, smooth films exhibit low coefficients of
friction of 0.03–0.04 in humid air, as well as in dry nitrogen. Rough
films exhibit slightly higher coefficients of friction of 0.05–0.07.
Measurements in high vacuum of 10-7 torr show very high coeffi-
cient of friction approaching 1.5–1.8.[203] The authors believe that in
humid air and nitrogen, friction arises from abrasion from a diamond
pin sliding on the diamond film, and high vacuum adhesion between
the sliding surfaces. They also observed that the friction in air and
nitrogen increases with the surface roughness, whereas in vacuum it
is independent of surface roughness.
(a) (b) (c)
Figure 57. SEM images of diamond films deposited by MPCVD. (a) Fine-grain
(20-100 nm) diamond film on (100) silicon; surface roughness rms = 15 nm. (b)
Medium-grain (1100 nm) diamond film on (100) silicon; surface roughness rms
= 63 nm. (c) Coarse-grain (3300 nm) diamond film on α-SiC silicon; surface
roughness rms = 160 nm.[203]
Figure 58. Coefficient of friction as a function of sliding distance of diamond

films whose surface morphologies are shown in Fig. 57, slid against a bulk
diamond pin (a) in humid air with 40% RH and (b) in dry nitrogen at 1 N normal
load, 86 mm min-1 sliding speed, and 25°C ambient temperature.[203]
Gangopadhyay and Tamor[115] studied the friction and wear

properties of diamond films deposited on silicon by MPCVD with
three different surface morphologies, Fig. 59 (i.e., faceted,
cauliflowered, and smooth) by changing the composition of CO+H2
precursor gas. The grain size and surface roughness (cla value) of
these films were measured as 0.2–1.5 mm and 29 nm; 1–3 mm and
21 nm; 0.1–0.3 mm and 11 nm, respectively. Shown in Fig. 60a are
the coefficients of friction of these diamond films sliding in recipro-
cating mode against themselves, silicon nitride, zirconia, and steel at
4.2 N at ambient temperature and relative humidity of 55% to 85%.
The average wear rates of counterface materials are compared in Fig.
60b. All material combinations except diamond film sliding against
another diamond film, exhibit high coefficients of friction ranging
from 0.38 to 0.90. Furthermore, in all cases faceted films exhibit
higher friction and wear than those of smooth films. The smooth
films also exhibit formation of a wear track during sliding tests. The
wear of the diamond film sliding against steel occurred by the
fracture of the tips of the grains, although evidence of delamination
was also observed. Similar wear mechanisms were observed for a
diamond film sliding against another diamond film.
(a) (b) (c)

Figure 59. SEM images of diamond films deposited by MPCVD at 900oC on
silicon using different compositions of precursor gas; (a) rough and faceted
films (grain size ~ 0.2–1.5 µm, rms roughness ~ 29 nm) using 20% CO+80% H2,
(b) cauliflower-type film (grain size ~ 1-3 µm, rms roughness ~ 21 nm) using
0.5% CH4+N2+97% H2, (c) smooth film (grain size ~ 0.1-0.3 µm, rms roughness
~ 11 nm) using 2% CH4+ 98% H2.[115]
Figure 60. (a) Coefficient of friction of diamond films whose surface morphologies
are shown in Fig. 59, slid against a diamond coated silicon nitride, silicon nitride,
alumina, zirconia, and steel at 4.2N load, 35 mm s-1 sliding speed, and 55–85%
relative humidity and (b) average wear rate of slider materials slid for 1 h against
the three diamond films.[115]
Gupta et al.[129] deposited 20 µm thick diamond films on

silicon using HFCVD technique from a mixture of methane (typi-
cally 1%) diluted in hydrogen at 800°C substrate temperature. Prior
to deposition, the silicon substrates were abraded with 4 to 8 µm and
submicron (0.5 µm and less) diamond powder in order to obtain
coarse-grain (grain size ~ 10 to 15 µm), and fine-grain (grain size
~ 3 to 5 µm) films, respectively. SEM and AFM images of these as-
deposited coarse-grain film, chemomechanically- and laser-polished
coarse-grain film, and as deposited fine grain film are shown in Fig.
61. The friction profiles of as deposited coarse-grain (rms ~650 nm)
film, as-deposited fine-grain (rms ~180 nm) film, laser polished (rms
~ 97 nm) coarse-grain film, and natural diamond (rms ~ 1.72 nm),
slid against an alumina ball are compared in Fig. 62. The polishing of
diamond films resulted in a significant reduction of the coeffi-
cient of friction from about 0.4 to 0.09. Diamond films with fine-
grains (~ 3–5 µm) exhibit considerably lower coefficients of friction
(0.20) as compared to the films having larger grains (~ 10–15 µm).
Wear of the mating alumina ball sliding against diamond surfaces
was considerably reduced after polishing. The wear tracks on the
mating alumina ball surface, and diamond films were examined by
optical microscopy, and the degree of wear was estimated in terms of
the diameter of the round wear scar on the ball, and the amount of the
wear debris collected on the film. Figure 63 shows the optical
micrographs of as-deposited, and chemomechanically-polished
coarse-grain diamond films, as-deposited fine-grain diamond films
and of natural diamond after a wear test. We note that the wear scar
on the alumina ball slid against the polished diamond film is compa-
rable to that slid against natural diamond after a wear test. Further,
the amount of alumina transferred on to the polished diamond film
during sliding is considerably lower than that for the as-deposited
coarse-grain diamond film (Fig. 63). The debris seen on the natural
diamond surface comes from the fracture of the asperities on the
alumina ball surface. It was observed that most of the debris in
natural diamond is generated during the early stage of sliding. The
role of surface roughness on the friction of diamond coatings is
shown in Fig. 64. The coefficient of friction decreased from 0.4 to

0.09 as the rms roughnesses of polished films were reduced from 650
to 170 nm, suggesting the important role of surface roughness. This
study demonstrated that the friction and wear characteristics of CVD
diamond films can be controlled by their grain size and surface
roughness; and the friction and wear characteristics of polished CVD
diamond films are comparable to natural diamond.
Figure 61. SEM and AFM images of (a) as-deposited coarse-grained films,
(b) as-deposited fine-grained films polished diamond films, (c) chemo-
mechanically-polished coarse-grained films, and (d) laser-polished coarse-grained
films, deposited by HFCVD.[129]
Figure 62. The coefficient of friction profile as a function of sliding distance of

as-deposited and polished HFCVD diamond films slid against an alumina ball at
1N in ambient conditions in reciprocating mode.[129]
Since a-C and a-C:H coatings can be deposited at lower tem-

peratures (< 200–300°C) on substrates ranging from soft polymers to
hard ceramic substrates without significantly affecting their surface
topography, numerous studies have been performed to correlate the
friction and wear properties of carbon films to their properties, and
subsequently to their deposition conditions.[9][106][107][116][120][122]–
[125][131]–[133][143][293][294] Like hardness and elastic modulus, friction
and wear properties of a-C and a-C:H coatings vary dramatically

with the deposition conditions. The environment plays a crucial role
in the friction and wear properties of a-C:H coatings because of their
lower thermal stabilities, the a-C:H coatings start to graphitize by
hydrogen loss above 400°C.[102][197][202] In this section, we present
some representative data to illustrate the effect of deposition condi-
tions, and the environment on the friction and wear properties of a-C
and a-C:H coatings. For more detailed information reader can refer
to a review by Grill.[120]
Figure 63. The optical images of the diamond surface and the alumina ball after
the sliding tests on (a) as-deposited coarse-grained films, (b) chemomechanically-
polished coarse-grained films, (c) as-deposited fine-grained films polished films,
and (d) natural diamond.[129]
Figure 64. Effect of surface roughness on the coefficient of friction of HFCVD

diamond films slid against an alumina ball at 1N normal load in ambient
conditions in reciprocating mode.[129]
The effect of ion beam energy and substrate temperature on the

friction and wear properties of a-C coatings deposited by direct ion
beam on aluminum substrate using a ring-cusp ion source operating
on methane gas was studied by Wei et al.[293] The variation in the
coefficients of friction as a function of number of reciprocating
cycles, of a-C coatings deposited at different ion beam energies, and
substrate temperatures sliding against an alumina ball at 0.2 N
normal load under 50 ±10% relative humidity and 21 ±2 oC tempera-
ture are shown in Fig. 65. The coatings deposited at 450 to 550 eV
ion beam energy and 50–100oC substrate temperatures exhibit the
lowest coefficient of friction of about 0.08–0.12 at 0.2 N normal load
and the highest number of cycles required to initiate damage at 5N
normal load.
The effect of pulsed bias voltage on the coefficient of friction
of hydrogen free amorphous carbon films deposited by cathodic arc
deposition technique, sliding against a diamond sphere of 2 mm
radius at 0.3 N normal load in ambient environment is shown in Fig.
66.[9] We note that the friction profile as a function of pulsed bias
voltage attains a minima of 0.09 at about -100 V. The authors believe
that the deposition conditions at -100 V biasing resulting in higher
hardness and mass density probably formed increased fraction of
sp3-bonded carbon which is responsible for lower friction. Anders et

al.[9] also observed that for a given bias voltage, increasing the
pulsed bias duty cycle results in lower coefficient of friction.
Figure 65. Coefficient of friction of a-C coatings deposited at different ion beam
energies and substrate temperatures sliding against an alumina ball at 0.2 N
normal load under 50 ±10% relative humidity and 21 ±2°C temperature.[293]
Figure 66. The coefficient of friction of a-C coatings deposited by cathodic arc
deposition on silicon slid against diamond at 0.3 N as a function of pulsed bias
voltage for pulse duty cycles of 10 and 33%.[9]
Gangopadhyay et al.[116] deposited a-C:H films by RF capaci-

tively coupled plasma-assisted CVD using methane-based precursor
gas at substrate biasing ranging from -100 to -1250 V corresponding
to ion energies from 30 to 400 eV. The effect of substrate biasing on
the coefficient of friction of a-C:H films slid against steel ball at 4.2
N in ambient humidity of 70% RH are compared in Fig. 67. We note
that films deposited at the lowest (-100 V) and the highest (-1250 V)
biasing exhibit high coefficient of friction in the range of 0.2 to 0.3,
while films deposited at the biasing from -200 to -300 V, exhibit the
lowest coefficient of friction of about 0.08. The lowest coefficient of
friction in these films can related to their high hardness of about 16
GPa, Fig. 52. Gangopadhyay et al. [116] have also studied the effect of
humidity on the friction and wear properties of the films deposited at
a substrate biasing of -350 V, and found that both friction and wear
increases with the relative humidity, Fig. 68. They observed a more
uniform transfer film on a steel ball at lower humidity (4%) than the
film at higher humidity (67%). Enke et al.[104] reported a low coeffi-
cient of friction of about 0.01–0.02 for RF PECVD deposited a-C:H
films using acetylene and argon mixture precursor gas at a relative
humidity of 1% and an increase in the coefficient of friction to 0.19
as relative humidity was increased to 100%.
Figure 67. The coefficient of friction of a-C:H films slid against steel at 4.2 N in
ambient conditions. These films were deposited voltage by rf capacitively coupled
plasma-assisted CVD at varying substrate biasing from -100 to -1250 V.[116]
Figure 68. Effect of relative humidity on (a) the coefficient of friction and (b)
wear volume of a-C:H films deposited at -350 V bias voltage by rf capacitively
coupled plasma-assisted CVD properties of a-C:H films.[116]
Memming et al.[197] reported an increase in the coefficient of

friction from about 0.04 to 0.25 for RF-PECVD a-C:H films, depos-
ited from acetylene based precursor gas, with an increase in relative
humidity from 1 to 100%. They measured an extremely low coeffi-
cient of friction of 0.02 in 10-8 torr vacuum environment; this is
contrary to the high coefficient of friction of about 1.6 of polycrystal-
line diamond films. Furthermore, Memming et al.[197] observed an
increase in the coefficient of friction from about 0.02 to 0.60 when
dry nitrogen was replaced with dry oxygen. Based on AES analyses
of worn coatings, the authors concluded that in case of films that
exhibit a low coefficient of friction, a carbonaceous material was
transferred from the carbon coating to the steel ball, while in case of
films that exhibit a high coefficient of friction (in dry oxygen) iron
was transferred to the coating. Miyoshi et al.[202] deposited 0.25-µm
thick for 30 kHz AC-PECVD a-C:H films, using methane and butane
precursor gases. They found low coefficients of friction of 0.02 and
0.06 between these films and diamond in dry nitrogen and ambient
air, respectively. The a-C:H coatings exhibit coefficients of friction
of 0.15 and 0.25 when slid against silicon nitride in dry nitrogen and
humid air, respectively. This increase in friction is due to the silicon
oxide films produced on the surface of silicon nitride pins in humid air.
Erdemir et al.[108] studied the nature of transfer layers formed
on M50 steel balls during sliding against 1.5 µm thick ion-beam
deposited carbon coatings on M50 steel disks. These coatings exhib-
ited a low coefficient of friction of about 0.12 at initial stage of
sliding against M50 steel ball in nitrogen at 5N normal load under
ambient temperature. The coefficient of friction decreased with
continuing sliding, and reached to a value of about 0.07 toward the
end of 1 km sliding test, and to about 0.03 toward the end of a long
test. Ultra low friction of these carbon films was attributed to the
formation of transfer films that were rich in carbon and had a
disordered graphitic structure.
Memming et al.[197] annealed a-C:H films at 550°C to study the
effect of hydrogen depletion during annealing on the friction and
wear properties. They found the annealing resulted in a high coeffi-
cient of friction of 0.68 in vacuum, as well as in dry nitrogen.
Miyoshi et al.[202] also observed a similar increase in the coefficient
of friction on annealing; Miyoshi and co-workers suggested the
formation of a graphite layer through a two stage process: a carbur-
ization stage resulting in hydrogen depletion and a polymerization
stage resulting in the formation of graphite crystallites.
Grill et al.[123] reported the effects of substrate temperature and
substrate biasing on the friction and wear properties of RF PECVD
a-C:H coatings deposited on silicon using acetylene. The coefficient
of friction of a-C:H films (hydrogen content-40%) sliding against
a steel ball in 40–70%, relative humidity was decreased from
0.35 ±0.04 to 0.20 ±0.04 as the substrate temperature increased from
100oC to 250°C. The substrate biasing did not show any significant
variation in friction. The insensitivity of the coefficient of friction to
the annealing temperature was similar to what was reported by
Memming et al.[197] and Miyoshi[201] in humid air.
Kim et al.[156] reported the effects of oxygen and humidity on
friction and wear of RF PECVD a-C coatings deposited on silicon
from methane and hydrogen mixture. The coefficient of friction of
these films against silicon nitride ball varied from 0.2 in 50% humid
argon and 100% humid air, to 0.06 in dry argon, Fig. 69. The
variations in friction were attributed to a transfer layer, produced by
friction, that covered the contact surface of the ball in all cases.[156]
Low friction in dry argon was resulted from the material (unoxidized
DLC debris) transferred from the carbon coating to the silicon nitride
ball. The oxidized DLC debris (carbonyl compounds) formed on the
a-C coating, as well as on the silicon nitride ball in air, and humid
argon increases the friction.[156] The lowest wear rates were observed
in dry environments. The humidity increases the wear rates of a-C
coatings in 50 and 100% argon, and humidity saturated air by five times.
Wear of these coatings was dominated by the tribochemistry in air
and humid argon, but in dry argon it is controlled by adhesive wear.
Figure 69. Coefficient of friction and wear rate of a-C coatings sliding against
silicon nitride ball in air and argon as a function of relative humidity. The test
conditions were as follows: silicon nitride ball diameter-7.94 mm, sliding speed-
18.7 ±0.8 mm s-1, load- 9.8 N, environment-ambient temperature.[156]
Donnet et al.[102] studied the tribochemistry of RF PECVD

deposited carbon coatings on silicon sliding against 52,100 steel
balls in a vacuum range from 10-7 to 50 Pa, dry nitrogen, and ambient
air. In all cases, a transfer layer was observed on the steel ball during
the first 100 cycles, associated with relatively a high coefficient of
friction of about 0.2–0.3. Friction decreased to 0.006–0.008 beyond
100 cycles in a vacuum below 10-1 Pa, and to 0.01–0.07 in a vacuum
of 10–50 Pa and in dry nitrogen. A high vacuum was found to be
associated with the ultralow friction and low wear. On the contrary,
a poor vacuum, and inert atmosphere was found to be associated with
the low friction, and moderate wear. In vacuum and dry nitrogen
atmospheres, the wear particles were consisted of amorphous hydro-
genated carbon. Ambient air was found to be associated with the
relatively high friction and severe wear, coupled with the formation
of roll-shaped debris of amorphous carbon containing iron oxide
precipitates.[102]
Based on the friction and wear data just presented, it is appar-
ent that the friction and wear properties of amorphous carbon coat-
ings are very sensitive to the sp3-bonded carbon fraction and the
amount of hydrogen present in the coatings as well as the operating
environment.
5.0 CLOSURE
The micromechanical and microtribological properties mea-

surements can be made on ultrathin coatings, and on the skin of bulk
materials by simulating single asperity contact scratching and slid-
ing situations by using a nanoindenter, an atomic force microscope,
and a friction force microscope. Depth profiles of various mechani-
cal properties such as elastic-plastic deformation behavior, hardness,
Young’s modulus of elasticity, fracture toughness, fatigue, and coat-
ing-substrate adhesion can be obtained by making nanoindentations
at different loads by using a nanoindenter. The evolution of wear on
a typical surface can be observed by using an atomic force micro-

scope or a friction force microscope.
Diamond films deposited by plasma enhanced CVD exhibit
hardness, elastic modulus, friction, and wear properties comparable
to those of natural diamond. High friction (0.4) of rough CVD
diamond films having grain size of diamond crystallites in the range
of 10–20 µm and rms roughness of about 700 nm can be reduced to
0.08 by polishing by using chemomechanical, and laser polishing
techniques. Properties of amorphous and hydrogenated amorphous
carbon coatings can be tailored by optimizing the deposition param-
eters. Amorphous carbon coatings with a hardness of more than 40
GPa can be deposited by cathodic arc deposition. The friction and
wear properties and wear mechanisms of amorphous and hydroge-
nated amorphous carbon coatings are found be very sensitive to the
test environment.
REFERENCES
1. Ageev, V. P., Armeyev, V., Yu, Chapliev, N. I., Kuzmichov, A. V.,

Pimenov, S. M., and Ralchenko, V. G., “Laser Processing of Diamond
and Diamond-Like Films,” Materials and Manufacturing Processes,
8:1–8 (1993)
2. Ahn, J., Mittal, K. L., and Macqueen, R. H., “Hardness and Adhesion
of Filmed Structures as Determined by the Scratch Technique,” in:
Adhesion Measurement of Thin Films, Thick Films, and Bulk
Coatings, (K. L. Mittal, ed.), STP 640, pp. 134–157, ASTM,
Philadelphia (1978)
3. Aisenberg, S., “Properties and Applications of Diamond Like Carbon
Films,” J. Vac. Sci. Technol. A, 2:369–371 (1984)
4. Akatsuka, F., Hirose, Y., and Komaki, K., “Rapid Growth of Diamond
Films by Arc Discharge Plasma CVD,” Jpn. J. Appl. Phys.,
27:L1600–L1602 (1988)
5. Alba, S., Loubet, J. L., and Vovelle, L., “Evaluation of Mechanical
Properties and Adhesion of Polymer Coatings by Continuous
Hardness Measurements,” J. Adhesion Sci. Technol., 7:131–140
(1993)
6. Alekhin, V. P., Berlin, G. S., Isaev, A. V., Kalei, G. N., Merkulov,

V. A., Skvortsov, V. N., Ternovskii, A. P., Krushchov, M. M.,
Shnyrev, G. D., and Shorshorov, M. K., Micromechanical Testing
by Micromechanical Testing of Materials by Microcompression,
Zavod. Lab., 38:619–621 (1972)
7. Anders, A., Anders, S., and Brown, I. G., “Macroparticle-Free Thin
Films Produced by an Efficient Vacuum Arc Deposition Technique,”
J. Appl. Phys., 74:4239–4241 (1993)
8. Anders, S., Anders, A., Brown, I. G., Wei, B., Komvopoulos, K.,
Ager, J. W., III, and Yu, K. M., “Effect of Vacuum Arc Deposition
Parameters on the Properties of Amorphous Carbon Thin Films,”
Surface Coat. Technol., 68/69:388–393 (1994)
9. Anders, A., Anders, S., Brown, I. G., Dickinson, M. R., and MacGill,
R. A., “Metal Plasma Immersion Ion Implantation and Deposition
Using Arc Plasma Sources,” J. Vac. Sci. Technol. B, 12:815–820
(1994)
10. Angus, J. C., “Diamond and Diamond-Like Films,” Thin Solid
Films 216:126–133 (1992)
11. Angus, J. C., Koidl, P., and Domitz, S., Carbon Thin Films, in
Plasma Deposited Thin Films, (J. Mort and F. Jensen, eds.), pp. 89–
127, CRC Press, Boca Raton, FL (1986)
12. Angus, J. C., and Hayman, C. C., “Low Pressure Metastable Growth
of Diamond and Diamondlike Phase,” Science, 241:913–921 (1988)
13. Angus, J. C., Wang, Y., and Sunkara, M., Metastable Growth of
Diamond and Diamond-Like Phases, Annu. Rev. Mater. Sci., 21:221–
248 (1991)
14. Angus, J. C., Argoitia, A., Gat, R., Li, Z., Sunkara, M., Wang, L.,
and Wang, Y., “Chemical Vapor Deposition of Diamond,” Philos.
Trans. R. Soc. London, Ser. A, 342:195–208 (1993)
15. Anonymous, Standard Test Method for Microhardness of Materials,
ASME Designation: E384–73:359–379 (1979)
16. Anonymous, NanoIndenterTMII Operating Instructions, Nano
Instruments, Inc., 1001 Larson Drive, Oak Ridge, TN 37830 (1991)
17. Anthony, T. R., “Methods of Diamond Making,” in Diamond and
Diamond-Like Films, (R. E. Clausing, L. L. Horton, J. C. Angus, and
P. Koidl, eds.), pp. 555–575, Plenum Press, New York, NY (1991)
18. Antis, G. R., Chantikul, P., Lawn, B. R., and Marshall, D. B., “A
Critical Evaluation of Indentation Techniques for Measuring Fracture
Toughness: I Direct Crack Measurements,” J. Am. Ceram. Soc.,
64:533–538 (1981)
19. Atkins, A. G., Silverio, A., and Tabor, D., “Indentation Hardness
and the Creep of Solids,” J. Inst. Metals, 94:369–378 (1966)
20. Bachmann, P. K., and Lydtin, H., “High Rate Versus Low Rate
Diamond CVD Methods,” in: Diamond and Diamond-Like Films
and Coatings, (R. E. Clausing, L. L. Horton, J. C. Angus, and P.
Koidl eds.), pp. 829–853, Plenum, New York (1991)
21. Bachmann, P. K., Lade, H., Leers, D., and Wiechert, D. U., “Wear
Testing of CVD Diamond Films,” Diamond and Related Materials,
3:799–804 (1994)
22. Bangert, H., Wagendristel, A., and Aschinger, H.,
“Ultramicrohardness Tester for Use in a Scanning Electron
Microscope,” Colliod and Polymer Sci., 259:238–240 (1981)
23. Bangert, H., and Wagendristel, A., “Ultralow Low Hardness Testing
of Coatings in a Scanning Electron Microscope,” J. Vac. Sci. Technol.
A, 4:2956–2958 (1986)
24. Bayer, R. G. (ed.), Selection and Use of Wear tests for Metals, STP
615, ASTM, Philadelphia (1976)
25. Bayer, R. G. (ed.), Wear tests for Plastics: Selection and Use, STP
701, ASTM, Philadelphia (1979)
26. Bayer, R. G. (ed.), Selection and Use of Wear tests for Coatings,
STP 769, ASTM, Philadelphia (1982)
27. Bayer, R. G., Mechanical Wear Prediction and Prevention, Marcel
Dekker Inc., New York, NY (1994)
28. Beerschwinger, U., Albrecht, T., Mathieson, D., Reuben, R. L.,
Yang, S. J., Taghizadeh, M., “Wear at Microscopic Scales and Light
Loads for MEMS Applications,” Wear, 181–183:426–435 (1995)
29. Beetz, C. P. Jr., Cooper, C. V., and Perry, T. A., “Ultralow-Load
Indentation Hardness and Modulus of Diamond Films Deposited by
Hot-Filament-Assisted CVD,” J. Mater. Res., 5:2555–2561 (1990)
30. Benjamin, P., and Weaver, C., “Measurement of Adhesion of Thin
Films,” Proc. R. Soc. London, Ser. A, 254:163–176 (1960)
31. Benzing, R. J., Goldblatt, I., Hopkins, V., Jamison, W., Mecklenburg,
K., and Peterson, M. B., Friction and Wear Devices, 2d Ed., ASLE,
Park Ridge, IL (1976)
32. Berg, S., and Anderson, L. P., “Diamond-Like Carbon Films
Produced in Butane Plasma,” Thin Solid Films, 58:117–120 (1979)
33. Berkovich, E. S., “Three-Faceted Diamond Pyramid for Micro-
Hardness Testing,” Indus. Diamond Rev., 11:129–132 (1951)
34. Bhattacharya, A. K., and Nix, W. E., “Finite Element Simulation of
Indentation Experiments,” Int. J. Solids Struct., 24:881–891(1988)
35. Bhattacharya, A. K., and Nix, W. D., “Analysis of Elastic and
Plastic Deformation Associated with Indentation Testing of the
Thin Films on Substrates,” Int. J. Solids Struct., 24:1287–1298
(1988)
36. Bhushan, B., “Stick-Slip Induced Noise Generation in Wear-
Lubricated Complaint Rubber Bearing,” ASME Journal of Tribology,
102:201–212 (1980)
37. Bhushan, B., “Overview of Coating Materials, Surface Treatments,
and Screening Techniques for Tribological Applications-Part 2:
Screening Techniques,” Testing of Metallic and Inorganic Coatings,
(W. B. Harding and G. A. DiBari, eds.), STP 947, pp. 310–319,
ASTM, Philadelphia (1987)
38. Bhushan, B., Tribology and Mechanics of Magnetic Storage Devices,
Second Edition, Springer-Verlag, New York (1996)
39. Bhushan, B., “Nanomechanical Properties of Solid Surfaces and
Thin Films,” in Handbook of Micro/Nanotribology, (B. Bhushan,
ed.), Second Edition, CRC Press, Boca Raton, FL (1999)
40. Bhushan, B., “Micro/Nanotribology and Micro/Nanomechanics of
MEMS,” Handbook of Micro/Nanotribology, (B. Bhushan, ed.),
Second Edition, CRC Press, Boca Raton, FL (1995)
41. Bhushan, B., (ed.) Handbook of Micro/Nanotribology, Second
Edition, CRC Press, Boca Raton, FL (1999)
42. Bhushan, B., and Sibley, L. B., “Silicon Nitride Rolling Bearing for
Extreme Operating Conditions,” ASLE Trans., 25: 417–428 (1982)
43. Bhushan, B., Williams, V. S., and Shack, R. V., “In-Situ

Nanoindentation Hardness Apparatus for Mechanical
Characterization of Extremely Thin Films,” ASME J. Tribology,
110:563–571(1988)
44. Bhushan, B., and Doerner, M. F., “Role of Mechanical Properties
and Surface Texture in the Real Area of Contact of Magnetic Rigid
Disks,” ASME J. Tribology, 111:452–458 (1989)
45. Bhushan, B., and Blackman, G. S., “Atomic Force Microscopy of
Magnetic Rigid Disks and Sliders and Its Applications to Tribology,”
ASME J. Tribology, 113:453–458 (1991)
46. Bhushan, B., and Gupta, B. K., Handbook of Tribology: Materials,
Coatings and Surface Treatments, McGraw Hill, New York, NY
(1991)
47. Bhushan, B., Kellock, A. J., Cho, N. H., and Ager, J. W.,
“Characterization of Chemical Bonding and Physical Characteristics
of Diamond-like Amorphous Carbon and Diamond Films,” J. Mater.
Res., 7:404–410 (1992)
48. Bhushan, B., and Venkatesan, S., “Mechanical and Tribological
Properties of Silicon for Micromechanical Applications: A Review,”
Adv. Info. Storage Syst., 5:211–239 (1993)
49. Bhushan, B., Subramaniam, V. V., Malshe, A., Gupta, B. K., and
Ruan, J., “Tribological Properties of Polished Diamond Films,” J.
Appl. Phys., 74:4174–4180 (1993)
50. Bhushan, B., and Koinkar, V. N., “Nanoindentation Hardness
Measurements Using Atomic Force Microscopy,” Appl. Phys. Lett.,
64:1653–1655 (1994)
51. Bhushan, B., and Koinkar, V. N., “Tribological Studies of Silicon
for Magnetic Recording Applications,” J. Appl. Phys.,
75:5741–5746 (1994)
52. Bhushan, B., and Ruan, J., “Atomic-Scale Friction Measurements
Using Friction Force Microscopy Part II-Application to Magnetic
Media,” ASME J. Tribology, 116:389–396 (1994)
53. Bhushan, B., Koinkar, V. N., and Ruan, J. A., “Microtribology of

Magnetic Media,” I MechE: Proc. Instn. Mech. Engrs., 208:17–29
(1994)
54. Bhushan, B., Subramaniam, V. V., and Gupta, B. K., “Polishing of
Diamond Films,” Diamond Films and Technology, 4:71–97 (1994)
55. Bhushan, B., and Gupta, B. K., “Micromechanical Characterization
of Ni-P Coated Aluminum-Magnesium, Glass, and Glass-Ceramic
Substrates, and Finished Magnetic Thin-Film Rigid Disks,” Adv.
Info. Storage Syst., 6:208–216 (1995)
56. Bhushan, B., Gupta, B. K., and Azarian, M. H., “Nanoindentation,
Microscratch, Friction and Wear Studies of Coatings for Contact
recording Applications,” Wear, 181–183:743–758 (1995)
57. Bhushan, B., Gupta, B. K., Sundaram, R., Dey, S., Anders, S.,
Anders, A., Brown, I. G., and Reader, P. D., “Development of Hard
Carbon Coatings for Thin-Film Tape Heads,” IEEE Trans. Mag.,
31:2976–2978 (1995)
58. Bhushan, B., Israelachvili, J. N., Landman, U., “Nanotribology:
Friction, Wear, and Lubrication on the Atomic Scale,” Nature,
374:607–616 (1995)
59. Binnig, G., Quate, C. F., and Gerber, C., “Atomic Force Microscope,”
Phys. Rev. Lett., 56:930–933 (1986)
60. Blau, P. J., and Lawn, B. R. (eds.), Microindentation Techniques in
Materials Science and Engineering, STP 889, ASTM, Philadelphia
(1986)
61. Bogli, U., Blatter, A., Pimenov, S. M., Smolin, A. A., and Konov, V.
I., ´“Smoothening of Diamond Films with an ArF Laser,” Diamond
and Related Materials, 1:782–788 (1992)
62. Bouilov, L. L., Chapliev, N. I., Konov, V. I., Pimenov, S. M.,
Smolin, A. A., and Spitsyn, B. V., “Excimer Laser Etching and
Polishing of Diamond Films,” J. Electrochem. Soc., 91:357 (1991)
63. Bovard, B. G., Zhao, T., and Macleod, H. A., “Oxygen-ion Beam
Polishing of a 5-cm-diameter Diamond Film,” Appl. Opt.,
31:2366–2369 (1992)
64. Bowden, F. P., and Tabor, D., The Friction and Lubrication of
Solids, pp. 172–175, Clarendon Press, Oxford Part I (1950), and
Part II (1964)
65. Bravman, J. C., Nix, W. D., Barnett, D. M., and Smith, D. A. (eds.),
“Thin Films: Stresses and Mechanical Properties, Symp. Proc.,”
Mat. Res. Soc., Vol. 130, Pittsburgh, PA (1989)
66. Brookes, C. A., and Brookes, E. J., “Diamond in Perspective: a
Review of Mechanical Properties of Natural Diamond,” Diamond
67. Brown, I. G., Anders, A., Anders, S., Dickinson, M. R., Ivanov, I.
C., McGill, R. A., Yao, X. Y., Yu, K. M., “Plasma Synthesis of
Metallic and Composite Thin Films with Atomically Mixed Substrate
Bonding,” Nucl. Instrum. Method Phys. Res., B 80/81:1281–1287
(1993)
68. Bubenzer, A., Dischler, B., Brandt, G., and Koidl, P., “RF Plasma
Deposited Amorphous Hydrogenated Hard Carbon Thin Films:
Preparation, Properties, and Applications,” J. Appl. Phys.,
54:4590–4595 (1983)
69. Buckle, H., The Science of Hardness Testing and its Research
Applications, (J. W. Westbrook and H. Conrad, eds.), pp. 453–491,
Am. Soc. Metals, Metals Park, OH (1973)
70. Bull, S. J., and Rickerby, D. S., “New Developments in the Modeling
of the Hardness and Scratch Adhesion of Thin Films,” Surf. Coatings
Technol., 42:149–164 (1990)
71. Bull, S. J., Chalker, P. R., Johnston, C., and Moore, V., “The Effect
of Roughness on the Friction and Wear of Diamond Thin Films,”
Surf. Coat. Technol., 68/69:603–610 (1994)
72. Bulychev, S. I., Alekhin, V. P., and Shorshorov, M. K., “Studies of
Physico-Mechanical Properties in Surface Layers and Microvolumes
of Materials by the Method of Continuous Application of an
Indenter,” Fizika Khim. Obrab. Materialov., No. 5 (1979)
73. Burnett, P. J., and Rickerby, D. S., “The Relationship Between
Hardness and Scratch Adhesion,” Thin Solid Films, 154:403–416
(1987)
74. Burnett, P. J., and Rickerby, D. S., “The Mechanical Properties of

Wear Resistant Coatings I: Modelling of Hardness Behavior,” Thin
Solid Films, 148:41–50 (1987)
75. Burnett, P. J., and Rickerby, D. S., “The Mechanical Properties of
Wear-Resistant Coatings II: Experimental Studies and Interpretation
of Hardness,” Thin Solid Films, 148:51–65 (1987)
76. Campbell, D. S., “Mechanical Properties of Thin Films,” Handbook
of Thin Film Technology, (L. I. Maissel and R. Glang, eds.), Ch. 12,
McGraw-Hill, NY (1970)
77. Catherine, Y., “Preparation Techniques for Diamond-Like Carbon,”
Diamond and Diamond-Like Films and Coatings, (R. E. Clausing,
L. L. Horton, J. C. Angus, and P. Koidl, eds. ), pp. 193–227, Plenum
Press, New York (1991)
78. Catherine, Y., and Couderc, P., “Electrical Characteristics and
Growth Kinetics in Discharges used for Plasma Deposition of
Amorphous Carbon,” Thin Solid Films, 144:265–280 (1986)
79. Chantikul, P., Anstis, G. R., Lawn, B. R., and Marshall, D. B., “A
Critical Evaluation of Indentation Techniques for Measuring Fracture
Toughness: II, Strength Method,” J. Am. Ceram. Soc., 64:539–543
(1981)
80. Cheng, W., Ling, E., and Finnie, I., “Median Cracking of Brittle
Solids Due to Scribing with Sharp Indenters,” J. Am. Ceram. Soc.,
73:580–586 (1990)
81. Chiang, S. S., Marshall, D. B., and Evans, A. G., “Simple Method
for Adhesion Measurement,” Surfaces and Interfaces in Ceramics
and Ceramic-Metal Systems, (J. Pask and A. G. Evans, eds.), pp.
603–612, Plenum, New York (1981)
82. Chiang, S. S., Marshall, D. B., and Evans, A. G., “The Response of
Solids to Elastic/Plastic Indentation: I. Stresses and Residual
Stresses,” J. Appl. Phys., 53:298–311 (1982)
83. Cho, N. H., Krishnan, K. M., Veirs, D. K., Rubin, M. D., Hopper, C.
B., Bhushan, B., and Bogy, D. B., “Chemical Structure and Physical
Properties of Diamond-like Amorphous Carbon Films Prepared by
Magnetron Sputtering,” J. Mater. Res., 5:2543–2554 (1990)
84. Chu, S. N. G., and Li, J. C. M., “Impression Creep: A New Creep
Test,” J. Mater. Sci., 12:2200–2208 (1977)
85. Chu, S. N. G., and Li, J. C. M., “Localized Stress Relaxation by
Impression Testing,” Mater. Sci. Eng., 45:167–171 (1980)
86. Clausing, R. E., Horton, L. L., Angus, J. C., Koidl, P., Diamond and
Diamond-Like Films and Coatings, NATO ASI Series, Vol. 266,
Plenum, New York, NY (1991)
87. Clauss, F. J., Solid Lubrication and Self-Lubricated Solids, Academic,
New York, NY (1972)
88. Cook, R. F., and Pharr, G. M., “Direct Observation and Analysis of
Indentation Cracking in Glasses and Ceramics,” J. Am. Ceram. Soc.,
73:787–817 (1990)
89. Cuomo, J. J., Pappas, D. L., Bruley, J., Doyle, J. P., Saenger, K. L.,
“Vapor Deposition Processes for Amorphous Carbon Films With
sp 3 Fractions Approaching Diamond,” J. Appl. Phys.,
70:1706–1711 (1991)
90. Cuomo, J. J., Doyle, J. P., Bruley, J., and Liu, J. C., “Ion Beam
Sputtered Diamond-Like Carbon with Densities of 2.9 g/cc,” J. Vac.
Sci. Technol. A, 9:2210–2215 (1991)
91. Cuomo, J. J., Pappas, D. L., Lossy, R., Doyle, J. P., Bruley, J.,
DiBello, G. W., and Krakow, W., “Energetic Carbon Deposition at
Oblique Angles,” J. Vac. Sci. Technol. A, 10:3414–3418 (1992)
92. Dekempeneer, E. H. A., Jacobs, R., Smeets, J., Meneve, J., Eersels,
L., Blanpain, B., Roos, J., Oostra, D. J., “RF Plasma-Assisted
Chemical Vapor Deposition of Diamond-Like Carbon: Physical and
Mechanical Properties,” Thin Solid Films, 217:56–61 (1992)
93. Deng, K., Ko, W. H., and Michal, G. M., “A Preliminary Study on
Friction Measurements in MEMS,” 6th Intl. Conf. Solid-State Sensors
and Actuators (Transducers’ 91), San Francisco, CA, pp. 213–216,
IEEE, New York, NY (1991)
94. Deng, K., and Ko, W. H., “Static Friction of Diamond-Like Carbon
Film in MEMS,” Sensors and Actuators, A35:45–50 (1992)
95. Deng, K., and Ko, W. H., “A Study of Static Friction Between
Silicon and Silicon Compounds,” J. Micromech. Microeng.,
2:14–20 (1992)
96. Deryaguin, B. V., and Fedoseev, D. V., “Growth of Diamond and

Graphite from the Gas phase,” Nauka, Moscow (1977)
97. Deshpandey, C. V., and Bunshah, R. F., “Diamond and Diamondlike
Films: Deposition Processes and Properties,” J. Vac. Sci. Technol.
A, 7:2294–2302 (1989)
98. DeVries, R. C., “Synthesis of Diamond Under Metastable
Conditions,” Annu. Rev. Mater. Sci., 17:161–187 (1987)
99. Doerner, M. F., Gardner, D. S., and Nix, W. D., “Plastic Properties
of Thin Films on Substrates as Measured by Submicron Indentation
Hardness and Substrate Curvature Techniques,” J. Mater. Res.,
1:845–851 (1986)
100. Doerner, M. F. and Nix, W. D., “A Method for Interpreting Data
from Depth-Sensing Indentation Instruments,” J. Mater. Res.,
1:601–609 (1986)
101. Doerner, M. F., Oliver, W. C., Pharr, G. M., and Brotzen, F. R.,
(eds.), “Thin Films: Stresses and Mechanical Properties II, MRS
Proceedings,” Vol. 188, Mat. Res. Soc., Pittsburgh, PA (1990)
102. Donnet, C., Belin, M., Auge, J. C., Martin, J. M., Grill, A., Patel, V.
V., “Tribochemistry of Diamond-Like Carbon Coatings in Various
Environments,” Surf. Coat. Technol., 68/69:626–631 (1994)
103. Eddy, C. R., Youchison, D. L., Sartwell, B. D., and Grabowski, K.
S., “Deposition of Diamond onto Aluminum by Electron Cyclotron
Resonance Microwave Plasma-Assisted CVD,” J. Mater. Res.,
7:3255–3259 (1992)
104. Enke, K., Dimigen, H., and Hubsch, H., “Frictional Properties of
Diamondlike Carbon Layers,” Appl. Phys. Lett., 36:291–292 (1980)
105. Erck, R. A., Nichols, F. A., Dierks, J. F., “Pull-Test Adhesion
Measurements of Diamondlike Films on Silicon carbide, Silicon
Nitride, Aluminum Oxide, and Zirconium Oxide,” J. Vac. Sci.
Technol. A, 12:1583–1586 (1994)
106. Erdemir, A., Switala, M., Wei, R., and Wilbur, P., “A Tribological
Investigation of the Graphite-to-Diamond-Like Behavior of
Amorphous Carbon Films Ion Beam Deposited on Ceramic
Substrates,” Surface Coat. Technol., 50:17–23 (1991)
107. Erdemir, A., Nichols, F. A., Pan, X. Z., Wei, R., and Wilbur, P. J.,
“Friction and Wear Performance of Ion-Beam-Deposited Diamond-
Like Carbon Films on Steel Substrates,” Diamond and Related
Materials, 3:119–125 (1993)
108. Erdemir, A., Bindal, C., Pagan, J., Wilbur, P., “Characterization of
Transfer Layers on Steel Surfaces Sliding Against Diamondlike
Carbon in Dry Nitrogen,” Thin Solid Films, (1995)
109. Erlandsson, R., McClelland, G. M., Mate, C. M., and Chiang, S.,
“Atomic Force Microscopy Using Optical Interferometry,” J. Vac.
Sci. Technol. A, 6:266–270 (1988)
110. Evans, T., and James, P. F., “A Study of the Transformation of
Diamond to Graphite,” Proc. R.. Soc. London, Ser. A, 277:260–269
(1964)
111. Fabes, B. D., Oliver, W. C., McKee, R. A., and Walker, F. J., “The
Determination of Film Hardness From the Composite Response of
Film and Substrate to Nanometer Scale Indentation,” J. Mater. Res.,
7:3056–3064 (1992)
112. Fujimori, S., Kasai, T., and Inamura, T., “Carbon Film Formation
by Laser Evaporation and Ion Beam Sputtering,” Thin Solid Films,
92:71–80 (1982)
113. Furakawa, R., Uyama, H., and Matsumoto, O., “Diamond Deposition
With Plasma Jet at Reduced Pressure,” IEEE Trans. Plasma Sci.,
18:930–933 (1990)
114. Gabriel, K. J., Behi, F., Mahadevan, R., Mehregany, M., “In Situ
Friction and Wear Measurement in Integrated Polysilicon
Mechanisms,” Sensors and Actuators, A21–A23:184–188 (1990)
115. Gangopadhyay, A. K., and Tamor, M. A., “Friction and Wear
Behavior of Diamond Films Against Steel and Ceramics,” Wear,
169:221–229 (1993)
116. Gangopadhyay, A. K., Vassell, W. C., Tamor, M. A., Willermet, P.
A., “Tribological Behavior of Amorphous Hydrogenated Carbon
Films on Silicon,” ASME J. Tribology, 116:454–462 (1994)
117. Gardos, M. N., “Tribology and Wear Behavior of Diamond,” in
Synthetic Diamond: Emerging CVD Science and Technology, (K. E.
Spear and J. P. Dismukes, eds.), pp. 419–504, John Wiley, New
York, NY (1994)
118. Gardos, M. N., and Soriano, B. L., “The Effect of Environment on

The Tribological Properties on Polycrystalline Diamond Films,” J.
Mater. Res., 5:2599–2609 (1990)
119. Greene, J. E., Woodhouse, J., and Pestes, M., A “Technique for
Detecting Critical Loads in the Scratch Test for Thin-Film Adhesion,”
Rev. Sci. Instrum., 45:747–749 (1974)
120. Grill, A., “Review of the Tribology of Diamond-Like Carbon,”
Wear, 168:143–153 (1993)
121. Grill, A., Meyerson, B. S., and Patel, V. V., “Interface Modifications
for Improving the Adhesion of a-C:H Films to Metals,” J. Mater.
Res., 3:214–217 (1988)
122. Grill, A., Meyerson, B. S., and Patel, V. V., “Diamond-like Carbon
Films by RF Plasma-Assisted Chemical Vapor Deposition from
Acetylene,” IBM J. Res. Develop., 34:849–857 (1990)
123. Grill, A., Patel, V. V., and Meyerson, B. S., “Optical and Tribological
Properties of Heat-Treated Diamond-like Carbon,” J. Mater. Res.,
5:2531–2537 (1990)
124. Grill, A., Horng, C. T., Meyerson, B. S., Patel, V. V., and Russak,
M. A., “Magnetic Head Slider Having a Protective Coating Thereon,”
U S Patent No. 5,159,508, (Oct. 27, 1992)
125. Grill, A., and Meyerson, B. S., “Development and Status of
Diamondlike Carbon,” Synthetic Diamond: Emerging CVD Science
and Technology, (K. E. Spear and J. P. Dismukes, eds.), pp. 91–141,
John Wiley, New York, NY (1994)
126. Grogan, D. F., Zhao, T., Bovard, B. G., and Macleod, H. A.,
“Planarizing Technique for Ion-Beam Polishing of Diamond Films,”
Appl. Opt., 31:1483–1487 (1992)
127. Gupta, B. K., Chevallier, J., and Bhushan, B., “Tribology of Ion
Bombarded Silicon for Micromechanical Applications,” ASME J.
Tribology, 115:392–399 (1993)
128. Gupta, B. K., Bhushan, B., and Chevallier, J., “Modification of
Tribological Properties of Silicon by Boron Ion Implantation,”
Tribol. Trans., 37:601–607 (1994)
129. Gupta, B. K., Malshe, A., Bhushan, B., Subramaniam, V. V., “Friction
and Wear Properties of Chemomechanically Polished Diamond
Films,” ASME J. Tribology, 116:445–453 (1994)
130. Gupta, B. K., and Bhushan, B., “The Nanoindentation Studies of Ion
Implanted Silicon,” Surface Coat. Technol., 68/69:564–570 (1994)
131. Gupta, B. K., and Bhushan, B., “Wear Resistant Coatings for Contact
Recording Applications,” IEEE Mag. Trans., 31:3012–3014 (1995)
132. Gupta, B. K., and Bhushan, B., “Micromechanical Properties of
Amorphous Carbon Coatings Deposited by Different Deposition
Techniques,” Thin Solid Films, 270:391–398 (1995)
133. Gupta, B. K., and Bhushan, B., “Mechanical and Tribological
Properties of Hard Carbon Coatings for Magnetic Recording Heads,”
Wear, 190:110–122 (1995)
134. Harding, D. S., Oliver, W. C., and Pharr, G. M., “Cracking During
Nanoindentation and its use in the Measurement of Fracture
Toughness,” Vol. 356, Mat. Res. Soc., Pittsburgh, PA (1985)
135. Hannula, S. P., Stone, D., and Li, C. Y., “Determination of Time-
Dependent Plastic Properties of Metals by Indentation Load
Relaxation Techniques,” Electronic Packaging Materials Science,
(E. A. Giess, K. N. Tu, and D. R. Uhlmann, eds.), 40:217–224, Mat.
Res. Soc., Pittsburgh, PA (1985)
136. Hart, E. W., and Solomon, H. D., “Load Relaxation Studies of
Polycrystalline High Purity Aluminium,” Acta Metall., 21:295–307
(1973)
137. Hayward, I. P., “Friction and Wear Properties of Diamond and
Diamond Coatings,” Surface Coat. Technol., 49:554–559 (1991)
138. Hayward, I. P., Singer, I. L., and Seitzman, L. E., “Effect of
Roughness on the Friction of Diamond on CVD Diamond Coatings,”
Wear, 157:215–227 (1992)
139. Heavens, O. S., “Some Factors Influencing the Adhesion of Films
Produced by Vacuum Evaporation,” J. Phys. Rad., 11:355–360
(1950)
140. Hirata, A., Tokura, H., and Yoshikawa, M., “Smoothing of Diamond
Films by Ion Beam Irradiation,” in: Application of Diamond Films
and Related Materials, (Y. Tzeng, M. Yoshikawa, M. Murakawa,
and A. Feldman, eds.), pp. 227–232, Elsevier Sci. Pub., Amsterdam,
Netherlands (1991)
141. Hirose, Y., Ananuma, S., Okada, N., Komaki, K., Proc. 1st Intl.
Sypm. on Diamond and Diamond-Like Films, The Electrochemical
Society, 89(12):80, Pennington, NJ (1989)
142. Holmberg, K., and Matthews, A., Coating Tribology: Properties,
Techniques and Applications in Surface Engineering, Elsevier,
London, UK (1994)
143. Holmberg, K., Koskinen, J., Ronkainen, H., Vihersalo, J., Hirvonen,
J. P., and Likonen, J., “Tribological Characteristics of Hydrogenated
and Hydrogen-Free Diamond-like Carbon Coatings,” Diamond Films
and Technology, 4:113–129 (1994)
144. Hooper, R. M., and Brookes, C. A., “Incubation Periods and
Indentation Creep in Lead,” J. Mater. Sci., 19:4057–4060 (1984)
145. Howe, R. T., “Surface Micromachining for Microsensors and
Microactuators,” J. Vac. Sci. Technol. B, 6:1809–1813 (1988)
146. Ishikawa, J., Takeiri, Y., Ogawa, K., and Takagi, T., “Transparent
Carbon Film Prepared by Mass-Separated Negative-Carbon-Ion-
Beam Deposition,” J. Appl. Phys., 61:2509–2515 (1987)
147. Jacobson, S., Jonsson, B., and Sundquist, B., “The Use of Fast
Heavy Ions to Improve Thin Film Adhesion,” Thin Solid Films,
107:89–98 (1983)
148. Jansen, F., Machonkin, M., and Kuhman, D. E., “The Deposition of
Diamond Films by Filament Techniques,” J. Vac. Sci. Technol. A,
8:3785–3790 (1990)
149. Je, J. H., Gyarmati, E., and Naoumidis, A., “Scratch Adhesion Test
of Reactively Sputtered TiN Coatings on a Soft Substrate,” Thin
Solid Films, 136:57–67 (1986)
150. Jin, S., and Moustakas, T. D., “Growth of Diamond Films by ECR
Plasma-Assisted at Low Pressures and Temperatures,” Diamond
151. Johnson, K. L., “Contact Mechanics,” Cambridge University Press,
Cambridge, UK (1985)
152. Jonsson, B., and Hogmark, S., “Hardness Measurements of Thin
Films,” Thin Solid Films, 114:257–269 (1984)
153. Julia-Schmutz, C., and Hintermann, H. E., “Microscratch Testing to

Characterize the Adhesion of Thin Layers,” Surface Coat. Technol.,
48:1 (1991)
154. Kamo, M., Sato, Y., Matsumoto, S., and Setaka, N., “Diamond
Synthesis From Gas Phase in Microwave Plasma,” J. Cryst. Growth,
62:642–644 (1983)
155. Kawarada, M., Mar, K., and Kiraki, A., “Large Area Chemical
Vapor Deposition of Diamond Particles and Films using Magneto-
Microwave Plasma,” Jpn. J. Appl. Phys., 26:L1032–L1034 (1987)
156. Kim, D. S., Fischer, T. E., and Gallois, B., “The Effects of Oxygen
and Humidity on Friction and Wear of Diamond-Like Carbon Films,”
Surf. Coat. Technol., 49:537–542 (1991)
157. King, R. B., “Elastic Analysis of Some Punch Problems for Layered
Medium,” Int. J. Solids Struc., 23:1657–1664 (1987)
158. Kitahama, K., Hirata, K., Nakamatsu, H., Kawai, S., Fujimori, N.,
Imai, T., Yoshino, H., and Doi, A., “Synthesis of Diamond by
Laser-Induced Chemical Vapor Deposition,” Appl. Phys. Lett.,
49:634–635 (1986)
159. Klages, C. P., “Chemical Vapor Deposition of Diamond,” Appl.
Phys. A, 56:513–526 (1993)
160. Kobayashi, K., Mutsukura, N, and Machi, Y., “Deposition of Hard
Carbon Films by the RF Glow Discharge Method,” Thin Solid
Films, 158:233–238 (1988)
161. Kohzaki, M., Higuchi, K., Noda, S., and Uchida, K., “Tribological
Characteristics of Polycrystalline Diamond Films Produced by
Chemical Vapor Deposition,” J. Mater. Res., 7:1769–1777 (1992)
162. Kohzaki, M., Uchida, K., Higuchi, K., and Noda, S., “Large-Area
High Speed Diamond Deposition by RF Induction Thermal Plasma
Chemical Vapor Deposition Method,” Jpn. J. Appl. Phys.,
32:L438–L440 (1993)
163. Koidl, P., Wild, C., Dischler, B., Wagner, J., and Ramsteiner, M.,
“Plasma Deposition, Properties and Structure of Amorphous
Hydrogenated Carbon Films,” Materials Science Forum,
52&53:41–70 (1989)
164. Kuo, C. T., Yen, T. Y., Huang, T. H., and Hsu, S. E., “Adhesion and
Tribological Properties of Diamond Films on Various Substrates,”
J. Mater. Res., 5:2515–2523 (1990)
165. Kurihara, K., Sasaki, K., Kawarada, M., and Koshino, N., “High
Rate Synthesis of Diamond of DC Plasma Jet Chemical Vapor
Deposition,” Appl. Phys. Lett., 52:437–438 (1988)
166. LaFontaine, W. R., Yost, B., Black, R. D., and Li, C., “Indentation
Load Relaxation Experiments on Al/Si Metallizations,” Symp. Proc.,
188:165–170, Mat. Res. Soc., Pittsburgh, PA (1990)
167. LaFontaine, W. R., Yost, B., Black, R. D., and Li, C. Y., “Indentation
Load Relaxation Experiments with Indentation Depth in the
Submicron Range,” J. Mater. Res., 5:2100–2106 (1990)
168. LaFontaine, W. R., Paszkiet, C. A., Korhonen, M. A., and Li, C. Y.,
“Residual Stress Measurements of Thin Aluminum Metallizations
by Continuous Indentation and X-ray Stress Measurement
Techniques,” J. Mater. Res., 6:2084–2090 (1991)
169. Lankford, J., “Threshold-Microfracture During Elastic/Plastic
Indentation of Ceramics,” J. Mater. Sci., 16:1177–1182 (1981)
170. Laugier, M., “The Development of Scratch Test Technique for the
Determination of the Adhesion of Coating,” Thin Solid Films,
76:289–294 (1981)
171. Laursen, T. A., and Simo, J. C., “A Study of The Mechanics of
Microindentation Using Finite Elements,” J. Mater. Res.,
7:618–626 (1992)
172. Lawn, B., Fracture of Brittle Solids, 2nd Ed., Cambridge Univ.
Press (1993)
173. Lawn, B., and Wilshaw, R., “Review Indentation Fracture: Principles
and Applications,” J. Mater. Sci., 10:1049–1081 (1975)
174. Lawn, B. R., Evans, A. G., and Marshall, D. B., “Elastic/Plastic
Indentation Damage in Ceramics: The Median/Radial Crack System,”
J. Am. Ceram. Soc., 63:574–581 (1980)
175. Lee, Y. H., Richard, P. D., Bachmann, K. J., and Glass, J. T., “Bias-
Controlled Chemical Vapor Deposition of Diamond Thin Films,”
Appl. Phys. Lett., 56:620–622 (1990)
176. Lettington, A. H., “Applications of Diamond-like Carbon Thin

Films,” Philos. Trans. R.. Soc. London, Ser. A, 342:287–296 (1993)
177. Li, J. C. M., and Chu, S. N. G., “Impression Fatigue,” Scr. Metall.,
13:1021–1026 (1979)
178. Li, W. B., Henshall, J. L., Hooper, R. M., and Easterling, K. E., “The
Mechanism of Indentation Creep,” Acta Metall. Mater.,
39:3099–3110 (1991)
179. Lim, M. G., Chang, J. C., Schultz, D. P., Howe, R. T., and White, R.
M., “Polysilicon Microstructures to Characterize Static Friction,”
IEEE Proc. MEMS’90, (J. E. Wood and R. T. Howe, eds.), Catalogue
# 90CH 2832–4, IEEE, pp. 82–88, New York (1990)
180. Lin, M. R., Ritter, J. E., Rosenfeld, L., and Lardner, T. J., “Measuring
the Interfacial Shear Strength of Thin Polymer Coatings on Glass,”
J. Mater. Res., 5:1110–1117 (1990)
181. Lossy, R., Pappas, D. L., Roy, R. A., and Cuomo, J. J., “Filtered Arc
Deposition of Amorphous Diamond,” Appl. Phys. Lett., 61:171–173
(1992)
182. Lucas, B. N., and Oliver, W. C., “The Elastic, Plastic and Time
Dependent Properties of Thin Films as Determined by Ultra Low
Load Indentation,” Thin Films: Stresses and Mechanical Properties
III, (W. E. Nix, J. C. Bravman, E. Arzt, and L. B. Freund, eds.),
183. Marshall, D. B., and Lawn, B. R., “Residual Stress Effects in Sharp
Contact Cracking Part 1 Indentation Fracture Mechanics,” J. Mat.
Sci., 14:2001–2012 (1979)
184. Marshall, D. B., and Oliver, W. C., “Measurement of Interfacial
Mechanical Properties in Fiber-Reinforced Ceramic Composites,”
J. Am. Ceramic Soc., 70:542–548 (1987)
185. Mate, C. M., McClelland, G. M., Erlandsson, R., and Chiang, S.,
“Atomic-Scale Friction of a Tungsten Tip on a Graphite Surface,”
Phys. Rev. Lett., 59:1942–1945 (1987)
186. Matsumoto, O., Toshima, H., and Kanzaki, Y., “Effect of Dilution
Gases in Methane on the Deposition of Diamond-Like Carbon in a
Microwave Discharge,” Thin Solid Films, 128:341–351 (1985)
187. Matsumoto, S., Sato, Y., Kamo, M., and Setaka, N., “Vapor
Deposition of Diamond Particles From Methane,” Jpn. J. Appl.
Phys., 21:L183–L185 (1982)
188. Matsumoto, S., Hino, M., and Kobayashi, T., “Synthesis of Diamond
Films in a RF Induction Thermal Plasma,” Appl. Phys. Lett.,
51:737–739 (1987)
189. Matthews, A., and Bachmann, P. K., Diamond and Diamond-Like
Carbon Coatings, Elsevier, Laussane, Switzerland (1990)
190. Mayo, M. J., and Nix, W. D., “A Micro-Indentation Study of
Superplasticity in Pb, Sn, and Sn-38Wt% Pb,” Acta Metall.,
36:2138–2192 (1988)
191. Mayo, M. J., Siegel, R. W., Narayanasamy, A., and Nix, W. D.,
“Mechanical Properties of Nanophase TiO 2 as Determined by
Nanoindentation,” J. Mater. Res., 5:1073–1082 (1990)
192. McClelland, G. M., Erlandsson, R., and Chiang, S., “Atomic Force
Microscopy: General Principles and a New Implementation,” In:
Review of Progress in Quantitative Nondestructive Evaluation, (D.
O. Thompson and D. E. Chimenti, eds.), 6B:1307–1314 Plenum
Press, New York (1987)
193. McKenzie, D. R., Muller, D., and Pailthorpe, B. A., “Compressive-
Stress-Induced Formation of Thin-Film Tetrahedral Amorphous
Carbon,” Phys. Rev. Lett., 67:773–776 (1991)
194. Mecholsky, J. J., Tsai, Y. L., and Drawl, W. R., “Fracture Studies of
Diamond on Silicon,” J. Appl. Phys., 71:4875–4881 (1992)
195. Mehregany, M., Senturia, S. D., and Lang, G. H., “Measurement of
Wear in Polysilicon Micromotors,” IEEE Trans. Electron Devices,
39:1136–1143 (1992)
196. Mehrotra, P. K., and Qunito, D. T., “Techniques for Evaluating
Mechanical Properties of Hard Coatings,” J. Vac. Sci. Technol.,
A3:2401–2405 (1985)
197. Memming, R., Tolle, H. J., and Wierenga, P. E., “Properties of
Polymeric Layers of Hydrogenated Amorphous Carbon Produced
by a Plasma-Activated Chemical Vapor Deposition Process II:
Tribological and Mechanical Properties,” Thin Solid Films,
143:31–41 (1986)
198. Mitsuda, Y., Yoshida T., and Akashi K., “Development of a New
Microwave Plasma Torch and Its Application to Diamond Synthesis,”
Rev. Sci. Instrum., 60:249–252 (1989)
199. Mittal, K. L., (ed.), Adhesion Measurements on Thin Coatings,
Thick Coatings and Bulk Coatings, STP 640, ASTM, Philadelphia
(1978)
200. Miyasato, T., Kawakami, Y., Kawano, T., Hiraki, A., “Preparation
of sp3-Rich Amorphous Carbon Film by Hydrogen Gas Reactive
RF-Sputtering of Graphite, and Its Properties,” Jpn. J. Appl. Phys.,
23:L234–L237 (1984)
201. Miyoshi, K., “Studies of Mechanochemical Interactions in the
Tribological Behavior of Materials,” Surf. Coat. Technol.,
43/44:799–812 (1990)
202. Miyoshi, K., Wu, R. L. C., and Garscadden, A., “Friction and Wear
of Diamond and Diamondlike Carbon Coatings,” Surf. Coat.
Technol., 54/55:428–434 (1992)
203. Miyoshi, K., Wu, R. L. C., Garscadden, A., Barnes, P. N., and
Jackson, H. E., “Friction and Wear of Plasma-Deposited Diamond
Films,” J. Appl. Phys., 74:4446–4454 (1993)
204. Mott, B. W., Microindentation Hardness Testing, Butterworths,
London (1957)
205. Moustakas, T. D., “Growth of Diamond CVD Method and Effect of
Process Parameters,” in: Synthetic Diamond: Emerging CVD
Science and Technology, (K. E. Spear and J. P. Dismukes, eds.),
pp. 145–192, John Wiley, New York, (1994)
206. Mulhearn, T. O., and Tabor, D., “Creep and Hardness of Metals: A
Physical Study,” J. Inst. of Metals, 89:7–12 (1960–61)
207. Nastasi, M., Parkin, D. M., and Gleiter, H., (eds.), Mechanical
Properties and Deformation Behavior of Materials Having Ultra-
Fine Microstructures, Kluwer Academic Pub., Dordrecht,
Netherlands (1993)
208. Nemanich, R. J., “Growth and Characterization of Diamond Thin
Films,” Annu. Rev. Mater. Sci., 21:535–558 (1991)
209. Nicoll, A. R., “A Survey of Methods Used for the Performance

Evaluation of High Temperature Coatings,” in: Coatings for High
Temperature Applications, (E. Lang, ed.), pp. 269–339, Applied
Science Publishers, London (1983)
210. Nix, W. D., “Mechanical Properties of Thin Films,” Metall. Trans.,
20A:2217–2245 (1989)
211. Nix, W. D., Bravman, J. C., Arzt, E., and Freund, L. B., (eds.), Thin
Films: Stresses and Mechanical Properties III, Symp Proc., Vol
239, Mat. Res. Soc., Pittsburgh, PA (1992)
212. Noguchi, K., Fujita, H., Suzuki, M., Yoshimura, N., “The
Measurements of Friction on Micromechatronics Elements,” Proc.
IEEE Micro Electro Mechanical Systems, (M. Esashi and H. Fujita,
eds.), pp. 148–153, IEEE, New York (1991)
213. O’Hern, M. E., and McHargue, C. J., “Mechanical Properties Testing
of Diamond and Diamond-Like Films by Ultra-Low Load
Indentation,” in: Diamond and Diamond-Like Films and Coatings,
(R. E. Clausing, L. L. Horton, J. C. Angus, P. Koidl, eds.), pp.
715–721 Plenum Press, New York (1991)
214. Ohtake, N., and Yoshikawa, M., “Diamond Film Preparation by Arc
Discharge Plasma Jet Chemical Vapor Deposition in the Methane
Atmosphere,” J. Electrochemical Soc., 137:717–722 (1990)
215. Ojha, S. M., Norstorm, M., and McCulluch, D., “The Growth
Kinetics and Properties of Hard and Insulating Carbonaceous Films
Grown in an RF Discharge,” Thin Solid Films, 60:213–225 (1980)
216. Oliver, W. C., Hutchings, R., and Pethica, J. B., “Measurement of
Hardness at Indentation Depths as Small as 20 nm,” Microindentation
Techniques in Materials Science and Engineering, (P. J. Blau and B.
R. Lawn, eds.), STP 889, pp. 90–108, ASTM, Philadelphia, PA
(1986)
217. Oliver, W. C., and Pethica, J. B., “Methods for Continuous
Determination of the Elastic Stiffness of Contact Between Two
Bodies,” U.S. Patent No. 4,848,141 (July 18, 1989)
218. Oliver, W. C., and Pharr, G. M., “An Improved Technique for
Determining Hardness and Elastic Modulus Using Load and
Displacement Sensing Indentation Experiments,” J. Mater. Res.,
7:1564–1583 (1992)
219. O’Neill, H., Hardness Measurement of Metals and Alloys, Chapman

and Hall, London (1967)
220. Palmquist, S., Jernkontorets Ann., 141:300 (1957)
221. Pappas, D. L., Saenger, K. L., Bruley, J., Krakow, W., and Cuomo,
J. J., “Pulsed Laser Deposition of Diamond-Like Carbon Films,” J.
Appl. Phys., 71:5675–5684 (1992)
222. Pappas, D. L., and Hopwood, “Deposition of Diamond-Like Carbon
Using a Planar Radio Frequency Induction Plasma,” J. Vac. Sci.
Technol. A, 12:1576–1582 (1994)
223. Perry, A. J., “The Adhesion of Chemically Vapor-Deposited Hard
Coatings on Steel-The Scratch Test,” Thin Solid Films, 78:77–93
(1981)
224. Perry, A. J., “Scratch Adhesion Testing of Hard Coating,” Thin
Solid Films, 197:167–180 (1983)
225. Perry, A. J., “Scratch Adhesion Testing: A Critique,” Surface
Enginnering, 2:183–189 (1986)
226. Perry, S. S., Ager, J. W., III., Somorjai, G. A., “Combined Surface
Characterization and Tribological (Friction and Wear) Studies of
CVD Diamond Films,” J. Mater. Res., 8:2577–2586 (1993)
227. Peterson, M. B., and Winer, W. O., (eds.), Wear Control Handbook,
ASME, New York (1980)
228. Pethica, J. B., Hutchings, R., and Oliver, W. C., “Hardness
Measurements at Penetration Depths as Small as 20 nm,” Phil.
Mag., A48:593–606 (1983)
229. Pethica, J. B., and Oliver, W. C., “Mechanical Properties of
Nanometer Volumes of Material: Use of the Elastic Response of
Small Area Indentations,” Thin Films: Stresses and Mechanical
Properties, (J. C. Bravman, W. D. Nix, D. M. Barnett, and D. A.
Smith, eds.), 130:13–23, Mat. Res. Soc., Pittsburgh, PA (1989)
230. Pharr, G. M., and Oliver, W. C., “Measurement of Thin Film
Mechanical Properties Using Nanoindentation,” MRS Bulletin, pp.
28–33 (July 1992)
231. Pharr, G. M., Oliver, W. C., and Brotzen, F. R., “On the Generality
of the Relationship Among Contact Stiffness, Contact Area, and
Elastic Modulus During Indentation,” J. Mater. Res., 7:613–617
(1992)
232. Pharr, G. M., Harding, D. S., and Oliver, W. C., “Measurement of

Fracture Toughness in Thin Films and Small Volumes Using
Nanoindentation Methods,” Mechanical Properties and Deformation
Behavior of Materials Having Ultra-Fine Microstructures, (M.
Nastasi, D. M. Parkin, and H. Gleiter, eds.), pp. 449–461, Kluwer
Academic Pub., Dordrecht, Netherlands (1993)
233. Pulker, H. K., and Salzmann, K., “Micro-/Ultramicro Hardness
Measurements with Insulating Films,” SPIE Thin Film Technologies,
652:139–144 (1986)
234. Raman, V., and Berriche, R., “Creep Behavior of Sputtered TiN
Films Using Indentation Testing,” Thin Films: Stresses and
Mechanical Properties II, (M. F. Doerner, W. C. Oliver, G. M.
Pharr, and F. R. Brotzen, eds.), pp. 171–176, Mat. Res. Soc.,
Pittsburgh, PA (1990)
235. Raman, V., and Berriche, R., “An Investigation of the Creep Processes
in Tin and Aluminum Using a Depth-Sensing Indentation
Technique,” J. Mater. Res., 7:627–638 (1992)
236. Raveh, A., Martinu, L., Hawthorne, H. M., and Wertheimer, M. R.,
“Mechanical and Tribological Properties of Dual-Frequency Plasma-
Deposited Diamond-Like Carbon,” Surf. Coat. Technol., 58:45–55
(1993)
237. Rickerby, D. S., and Matthews, A., Advanced Surface Coatings: A
Handbook of Surface Engineering, Chapman and Hall, New York
(1991)
238. Robertson, J., “Amorphous Carbon,” Advances in Phys.
35:317–374 (1986)
239. Robertson, J., “Structure and Electronic Properties of Diamond-
Like Carbon,” in: Diamond and Diamond-Like Carbon Films, (R. E.
Clausing, et al., eds.), pp. 331–356, Plenum Press, New York (1991)
240. Robertson, J., “Properties of Diamond-Like Carbon,” Surf. Coat.
Technol., 50:185–203 (1992)
241. Robertson, J., “Deposition of Diamond-Like Carbon,” Philos. Trans.
R. Soc. London, Ser. A, 342:277–286 (1993)
242. Ruan, J., and Bhushan, B., “Atomic-Scale Friction Measurements
Using Friction Force Microscopy: Part I-General Principles
and New Measurement Techniques,” ASME J. Tribology,
116:378–388 (1994)
243. Rubin, M., Hopper, C. B., Cho, N. H., and Bhushan, B., “Optical and
Mechanical Properties of DC Sputtered Carbon Films,” J. Mater.
Res., 5:2538– 2542 (1990)
244. Sargent, P. M., “Use of the Indentation Size Effect on Microhardness
of Materials Characterization,” Microindentation Techniques in
Materials Science and Engineering, (P. J. Blau, and B. R. Lawn,
eds.), STP 889, pp. 160–174, ASTM, Philadelphia (1986)
245. Savvides, N., “Diamondlike Thin Films and Their Properties,”
Materials Science Forum, 52 /53:407–426 (1989)
246. Savvides N., and Bell, T. J., “Microhardness and Young’s Modulus
of Diamond and Diamondlike Carbon Films,” J. Appl. Phys.,
72:2791–2796 (1992)
247. Sawabe, A., and Inuzuka, T., “Growth of Diamond Thin Films by
Electron Assisted Chemical Vapor Deposition,” Appl. Phys. Lett.,
46:146–147 (1985)
248. Scheibe, H. J., and Schultrich, B., “DLC Film Deposition by Laser-
Arc and Study of Properties,” Thin Solid Films, 246:92–102 (1994)
249. Scruby, C. B., “An Introduction to Acoustic Emission,” J. Phys. E:
Sci. Instrum., 20:946–953 (1987)
250. Seal, M., “The Effect of Surface Orientation on the Graphitization
of Diamond,” Phys. Stat. Sol., 3:658–664 (1963)
251. Seino, Y., Hida, N., and Nagai S., “Mechanical Properties of Diamond
Thin Film Prepared by Chemical Vapor Deposition,” J. Mater. Sci.
Lett., 11:515–517 (1992)
252. Seino, Y., and Nagai S., “Temperature Dependence of Young’s
Modulus of Diamond Thin Film Prepared by Microwave Plasma
Chemical Vapor Deposition,” J. Mater. Sci. Lett., 12:324–325 (1993)
253. Sekler, J., Steinmann, P. A., and Hintermann, H. E., “The Scratch
Test: Different Critical Load Determination Techniques,” Surf. Coat.
Technol., 36:519–529 (1988)
254. Shih, C. W., Yang, M., and Li, J. C. M., “Effect of Tip Radius on
Nanoindentation,” J. Mater. Res., 6:2623–2628 (1991)
255. Sneddon, I. N., “The Relation Between Load and Penetration in the
Axisymmetric Boussinesq Problem for a Punch of Arbitrary Profile,”
Int. J. Eng. Sci., 3:47–57 (1965)
256. Spear, K. E., “Diamond-Ceramic Coating of the Future,” J. Am.

Ceram. Soc., 72:171–191 (1989)
257. Spear, K. E., and Dismukes, J. P., Synethtic Daimond: Emerging
CVD Science and Technology John Wiley, New York, NY (1994)
258. Spitsyn, B. V., “Chemical Crystallization of Diamond from the
Activated Vapor Phase,” J. Crystal Growth, 99:1162–1167 (1990)
259. Spitsyn, B. V., and Deryaguin, B. V., “A Technique of Diamond
Growth on a Diamond Face,” USSR Patent No. 339,134 (May 5,
1980)
260. Spitsyn, B. V., Bouilov, L. L., and Deryaguin, B. V., “Diamond and
Diamond-Like Coatings: Deposition from the Vapor Phase, Structure
and Properties,” Prog. Crystal Growth and Character., 17:79–170
(1988)
261. Steinmann, P. A., Tardy, Y., and Hintermann, H. E., “Adhesion
Testing by the Scratch Test Method: The Influence of Intrinsic and
Extrinsic Parameters on the Critical Load,” Thin Solid Films,
154:333–349 (1987)
262. Stilwell, N. A., and Tabor, D., “Elastic Recovery of Conical
Indentation,” Proc. Phys. Soc., 78: 169–179 (1961)
263. Stone, D., LaFontaine, W. R., Alexopoulos, P. S., Wu, T. W., and
Li, C. Y. , “An Investigation of Hardness and Adhesion of Sputter-
Deposited Aluminum on Silicon by Utilizing a Continuous
Indentation Test,” J. Mater. Res., 3:141–147 (1988)
264. Stoner, B. R., Ma, G. H. M., Wolter, S. D., and Glass, J. T.,
“Characterization of Bias-Enhanced Nucleation of Diamond on
Silicon by in vacuo Surface Analysis and Transmission Electron
Microscopy,” Phys. Rev. B, 45:11067–11084 (1992)
265. Sundgren, J. E., and Hentzell, H. T. G., “A Review of the Present
Status of the Art in Hard Coatings Grown from the Vapor Phase,” J.
Vac. Sci. Technol. A, 4:2259–2279 (1986)
266. Suzuki, K., Sawabe, A., and Yasuda, H., Inuzuka, T., “Growth of
Diamond Thin Films by DC Plasma Chemical Vapor Deposition,”
Appl. Phys. Lett., 50:728–729 (1987)
267. Suzuki, J. I., Kawarada, H., Mar, K. S., Wei, J., Yokota, Y., and
Hiraki, A., “The Synthesis of Diamond Films at Lower Pressure and
Lower Temperature Using Magneto-Microwave Plasma CVD,” Jpn.
J. Appl. Phys., 28 L281–L283 (1989)
268. Suzuki, S., Matsuura, T., Uchizawa, M., Yura, S., and Shibata, H.,
“Friction and Wear Studies on Lubricants and Materials Applicable
to MEMS,” Proc. IEEE Micro Electro Mechanical Systems, (M.
Esashi, and H. Fujita, eds.), pp. 143–147, IEEE, New York (1991)
269. Suzuki, S., Matsuura, T., Karino, I., Sakai, Y., and Shibata, H.,
“Quality and Wear Durability of Diamond-Like Carbon Protective
Films for Thin Film Magnetic Media, Tribology and Mechanics of
Magnetic Storage Systems,” STLE SP-29, pp. 144–150 ,STLE, Park
Ridge, IL (1990)
270. Swain, M. V., Hagan, J. T., Field, J. E., “Determination of the
Surface Residual Stresses in Tempered Glasses by Indentation
Fracture Mechanics,” J. Mater. Sci., 12:1914–1917 (1977)
271. Swec, D. M., Mirtich, M. J., and Banks, B. A., “Ion Beam and
Plasma Methods of Producing Diamondlike Carbon Films,” 32nd
Annual International Technical Symposium on Optical and
Optoelectronic Applied Science and Engineering, Proc. SPIE,
Bellingham, WA (1989)
272. Tabor, D., The Hardness of Metals, Clarendon Press, Oxford, UK
(1951)
273. Tabor, D., “The Hardness of Solids,” Rev. Phys. Technol.,
1:145–179 (1970)
274. Tai, Y. C., and Muller, R. S., “Frictional Study of IC Processed
Micromotors,” Sensors and Actuators, A21–A23:180–183 (1990)
275. Tai, Y. C., and Muller, R. S., “IC Processed Electrostatic Synchronous
Micromotors,” Sensors and Actuators, A20:49–55 (1989)
276. Takagi, T., “Surface Interactions during Thin Film Deposition,” J.
Vac. Sci. Technol. A, 2:382–388 (1984)
277. Tamor, M. A., Vassell, W. C., and Carduner, K. R., “Atomic
Constraint in Hydrogenated Diamondlike Carbon,” Appl. Phys.
Lett., 58:592–594 (1991)
278. Tangena, A. G., and Hurkx, G. A. M., “The Determination of Stress-
Strain Curves of Thin Layers Using Indentation Tests,” ASME J.
Eng. Mat. Technol., 108:230–232 (1986)
279. Ternovskii, A. P., Alekhin, V. P., Shorshorov, M. K., Khrushchov,
M. M., and Skvortsov, V. N., “Micromechanical Testing of Materials
by Depression,” Zavod. Lab., 39:1620–1624 (1973)
280. Thronton, A. G., and Wilks, J., “Diamond Research 1974,” Suppl.
Ind. Diamond Rev., 39–42 (1974)
281. Tokura, H., Yang, C. F., and Yoshikawa, M., “Study on the Polishing
of Chemically Vapor Deposited Diamond Films,” Thin Solid Films,
212:49–55 (1992)
282. Townsend, P. H., Weihs, T. P., Sanchez, J. E., and Borgesen, P.
(eds.), “Thin Films: Stresses and Mechanical Properties IV,” Vol.
308, Mat. Res. Soc., Pittsburgh, PA (1993)
283. Tsai, H., “Structure and Physical Properties of Amorphous
Hydrogenated Carbon (a-C:H) Films,” Materials Science Forum,
52/53:71–102 (1989)
284. Tsai, H., and Bogy D. B., “Characterization of Diamond-like Carbon
Films and Their Application as Overcoats on Thin-Film Media for
Magnetic Recording,” J. Vac. Sci. Technol. A, 5:3287–3312 (1987)
285. Tsukamoto, Y., Kuroda, H., Sato, A., and Yamaguchi, H.,
“Microindentation Adhesion Tester and its Applications to Thin
286. Tzeng, Y., Yoshikawa, M., Murakawa, M., and Feldman, A., (eds.),
Applications of Diamond Films and Related Materials, Elsevier
Science Pub., Amsterdam, The Netherlands (1991)
287. Valli, J., “A Review of Adhesion Test Method for Thin Hard
Coatings,” J. Vac. Sci. Technol. A, 4:3007–3014 (1986)
288. Valli, J., Makela, U., and Matthews, A., “Assessment of Coating
Adhesion,” Surface Engineering, 2:49–53 (1986)
289. Vandentop, G. J., Kawasaki, M., Nix, R. M., Brown, I. G., Salmeron,
M., and Somorjai, G. A., “Formation of Hydrogenated Amorphous
Carbon Films of Controlled Hardness From a Methane Plasma,”
Phys. Rev., B41:3200–3210 (1990)
290. Walker, W. W., The Science of Hardness Testing and its Research
Applications, (J. H. Westbrook, and H. Conrad, eds.), pp. 258–273,
Am. Soc. Metals, Metals Park, OH (1973)
291. Wang, M., Schmidt, K., and Reichelt, K., “Characterization of
Metal-Containing Amorphous Hydrogenated Carbon Films,” J.
Mater. Res., 7:667–676 (1992)
292. Wei, J., Kawarada, H., Suzuki, J. I., and Hiraki, A., “Low-
Temperature Synthesis of Diamond Films using Magneto-Microwave
Plasma CVD,” Jpn. J. Appl. Phys., 29:L1483–L1485 (1990)
293. Wei, R., Wilbur, P. J., Erdemir, A., and Kustas, F. M., “The Effects
of Beam Energy and Substrate Temperature on the Tribological
Properties of Hard-Carbon Films on Aluminum,” Surf. Coat.
Technol., 51:139–145 (1992)
294. Wei, R., Wilbur, P. J., and Liston, M. J., “Effects of Diamond-Like
Hydrocarbon Films on Rolling Contact Fatigue of Bearing Steels,”
Diamond and Related Materials, 2:898–903 (1993)
295. Weihs, T. P., Lawrence, C. W., Derby, C. B., and Pethica, J. B.,
“Acoustic Emissions During Indentation Tests,” Thin Films: Stresses
and Mechanical Properties III, Symp. Proc., (W. D. Nix, J. C.
Bravman, E. Arzt, and L. B. Freund, eds.), 239:361–370, Mat. Res.
Soc., Pittsburgh, PA (1992)
296. Weissmantel, C., Bewilogua, K., Breuer, K., Dietrich, D., Ebersbach,
U., Erler, H. J., Rau, B., and Reisse, G., “Preparation and Properties
of Hard i-C and i-BN Coatings,” Thin Solid Films, 96:31–44 (1982)
297. Westbrook, J. H., Proc. Am. Soc. Test. Mater., 57:873 (1957)
298. Westbrook, J. H., and Conrad, H., (eds.), “The Science of Hardness
and its Research Applications,” Am. Soc. Metals, Metals Park, Ohio
(1973)
299. White, R. L., Nelson, J., and Gerberich, W. W. , “Residual Stress
Effects in the Scratch Adhesion Testing of Tantalum Thin Films,”
Thin Films: Stresses and Mechanical Properties IV, (P. H. Townsend,
T. P. Weihs, J. E. Sanches, and P. Borgesen, eds.), 308:141–146,
Mat. Res. Soc., Pittsburgh, PA (1993)
300. Wood, P., Wydeven, T., and Tsuji, O., “Effect of Temperature on
the Deposition Rate and Properties of Hydrogenated Amorphous
Carbon Films,” Surf. Coat. Technol., 49:399–405 (1991)
301. Wu, T. W., “Microscratch Test for Ultra-Thin Films,” Thin Films:
Stresses and Mechanical Properties II, (M. F. Doerner, W. C.
Oliver, G. M. Pharr, and F. R. Brotzen, eds.), Symp. Proc.,
302. Wu, T. W., “Microscratch and Load Relaxation Tests for Ultra-Thin
Films,” J. Mater. Res., 6:407–426 (1991)
303. Wu, T. W., “The AC Indentation Technique And Its Applications,”

Mater. Chem. Phys., 33:15–30 (1993)
304. Wu, T. W., Hwang, C., Lo. J., and Alexopoulos, P., “Microhardness
and Microstructure of Ion-Beam-Sputtered, Nitrogen Doped NiFe
305. Wu, T. W., Burn, R. A., Chen, M. M., and Alexopoulos, P. S.,
“Micro-Indentation and Micro-Scratch Tests on Sub-Micron Carbon
Films,” Symp. Proc., 130:117–121, Mat. Res. Soc., Pittsburgh, PA
(1989)
306. Wu, T. W., Shull, A. L., and Lin, J., “Microscratch Test on Carbon
Films as Thin as 20 nm,” Symp. Proc., 188:207–212, Mat. Res. Soc.,
Pittsburgh, PA (1990)
307. Wu, T. W., Shull, A. L., and Berriche, R. , “Microindentation
Fatigue Tests on Submicron Carbon Films,” Surf. Coat. Technol.,
47:696–709 (1991)
308. Wu., T. W., and Lee, C. K., “The Micro-wear Technique and its
Application to Ultrathin Film Systems,” J. Mater. Res., 9:805–811
(1994)
309. Wyon, C., Gillet, R., and Lombard, L., “Properties of Amorphous
Carbon Films Produced by Magnetron Sputtering,” Thin Solid Films,
122:203–216 (1984)
310. Yarbrough, W. A., Inspector, A., and Messier, R., “The Chemical
Vapor Deposition of Diamond,” Materials Science Forum, 52/
53:151–174 (1989)
311. Yarbrough, W. A., and Messier, R., “Current Issues and Problems in
the Chemical Vapor Deposition of Diamond,” Science,
247:688–696 (1990)
312. Yeak-Scranton, C. E., “Novel Piezoelectric Transducer to Monitor
Head-Disk Interactions,” IEEE Trans. Magn., Mag. 22:1011–1016
(1986)
313. Yust, C., and Bayer, R. G. (eds.), Selection and Use of Wear Tests
for Ceramics, STP 1010, ASTM, Philadelphia (1988)
314. Zhao, T., Grogan, D. F., Bovard, B. G., and Macleod, H. A.,
“Diamond Film Polishing With Argon and Oxygen Ion Beams,” in:
SPIE Diamond Optics III, 1325:142–151(1990)
Applications to Cutting
Tools
Otto Knotek, F. Löffler, and G. Krämer
1.0 INTRODUCTION
Aspects of quality and economy of industrial production pro-

cesses are to a high degree determined by the selection and the
design of appropriate manufacturing operations. For many manufac-
turing operations, especially for the technologically relevant pro-
cessing of metallic materials, machining with geometrically speci-
fied cutting edges are applicable. Enhancing the performance of
machining operations is therefore an economically important goal;
for achieving that, the coating technology may contribute in varying
ways, as shown below.
The cutting tool is the component most stressed, and therefore
limits the performance in machining operations. Among manifold
tribological stresses, thermal and mechanical loads affect the cutting
tool edges in a continuous or intermitting way. As a result, in
addition to good wear resistance, high thermal stability and high
mechanical strength are properties required for cutting materials.
Opposing this objective of an ideal cutting material is the
fundamental contradiction of properties like hardness, strength at
elevated temperature and wear resistance on one hand, and bending
strength and bending elasticity on the other hand, as indicated
370
Applications to Cutting Tools 371
qualitatively in Fig. 1. Cutting materials for extreme requirements

(for example, interrupted cuts or machining of high strength materi-
als) can consequently not be made from one single material, but may
be realized by composite materials. Surface coatings may improve
the tribological properties of cutting tools in an ideal way (cf., Fig. 1)
and therefore allow the application of tough or ductile substrate
materials, respectively.
Figure 1. Characteristics of different cutting materials.
This possibility for optimizing properties of cutting tools is

increasingly applied in machining technology. In 1991 about 80% of
the indexible tips used for turning were coated; the percentage of
coated inserts used for milling, however, reached only about
25–30%.[1] More recently, in the field of coated cemented carbides
used for milling, high rates of increase were recorded, so that the
percentage of coated inserts for that operation is probably up to about
40% by now.
Suitable coating processes for tools are CVD (chemical vapor
deposition) and PVD (physical vapor deposition) technologies. Both,
partly competing deposition techniques, permit the deposition of
thin hard coatings (about 3–12 µm) on tool materials, but differ
regarding process temperatures and process flexibility.
The advantages and disadvantages of these technologies con-
cerning coatings of tools are explained in detail in the following parts
of this study. Furthermore, the performance characteristics of CVD
and PVD coated tools will be described. The performance character-
istics for a special tool is determined by the combination of substrate
material, coating material, and coating process (Fig. 2). For each
application, specified through parameters such as cutting speed,
feed, and cutting material, an optimal set of parameters has to be
evaluated.
Figure 2. Parameters for the development of new coatings for machining

applications.
In order to understand the dependencies between properties of

coated tools, and their performance characteristics, the knowledge of
the evident wear and failure mechanisms taking place at the tool
during application is necessary. Therefore, a survey of these mecha-
nisms is given prior to the descriptions of cutting performance
characteristics of coated tools.
1.1 Wear Mechanisms in Machining
The tribological, thermal, and mechanical stresses due to ma-

chining processes lead to wear and mechanical overloading of the
cutting edges of machining tools. Owing to the failure mechanisms
present, typical signs of wear, e.g., cratering, top face wear, chipping
of various dimensions, or plastic deformations may result (Fig. 3).
These signs of deterioration are, however, rarely due to one indi-
vidual wear mechanism, but are rather a result of a complex interac-
tion of various factors.[2][3]
Figure 3. Failures of cutting tools.[3]
The entire interactions at the cutting edge and the relevant

parameters may be described using a tribomechanical system (Fig.
4) which considers not only the wear loads in areas close to the
surface, but also the loads of the substrate material. This is especially
important for interrupted machining processes, e.g., milling, since
thermal and mechanical cyclic stressing in these applications may
lead quickly to cracks and cutting edge fractures.[3][4]
In Fig. 5, the differences of the stresses impacting upon cutting
tools due to continuous and interrupted cutting are characterized by
the resulting wear patterns. While in continuous cutting, cratering
and top face wear is caused by abrasion and tribo-oxidation; the wear
pattern in interrupted cutting is determined by thermal and mechani-
cal fatigue cracks and cutting edge chipping.
Figure 4. Tribomechanical system of the machining process.

Figure 5. Typical tool wear patterns[5] (a) in continuous cutting, showing crater
and flank wear caused by high temperature chemical interaction and abrasion; (b)
in interrupted cutting, showing thermal and mechanical fatigue cracks and edge
fractures.
In addition to continuity and discontinuity of the cutting pro-

cess, material properties and cutting parameters play an important
part for the intensity of the resulting stresses, and therefore, the
presence of certain failure mechanisms. Adhesion between tool and
workpiece is determined by type and similarity of binding structures
of both partners, and abrasion is determined by microstructure and
hardness of contacting partners. Soft, ductile workpiece materials
tend, generally speaking, to adhesive wear while hard, brittle materi-
als promote abrasion.[6]
The cutting temperature at the location of the chip formation
has the largest influence on the effects of the various wear mecha-
nisms. This temperature is an indirect parameter, depending on
material properties and direct machining parameters. Among the
latter, the cutting speed can be regarded as the dominant influence, so
that the existence of different failure mechanisms can be correlated
to it. In Fig. 6, this is shown for continuous and interrupted cutting,
and it is obvious that for low and medium ranges of cutting speeds,
mainly adhesion and abrasion contribute to tool failure. Adhesion
leads to microweldings and to the formation of built-up cutting edges,
through which the cutting edge is locally exposed to extreme stresses
for a short time. This will lead quickly, especially in interrupted
cutting, to edge break-out and chip-off. Furthermore, in interaction
with abrasive mechanisms, wear at the top face will occur. Abrasive
and tribochemical mechanisms of wear become more important with

increasing cutting speed for continuous cutting, showing obvious
signs of cratering. In addition, the formation of cracks due to ther-
mally induced cyclic thermal shock stresses will lead, for interrupted
cutting, to quick failures by cutting edge break-outs (cf. Fig. 5),
which is the dominant failure mode for high strength materials.
(a)
(b)
Figure 6. Failure mechanisms in (a) continuous cutting (in Ref. 7) (b) interrupted
cutting[3]
From these so far described correlations in the tribomechanical

system of cutting processes, the requirements for tool coatings may
be derived. In order to achieve an improved cutting performance, or
an increased tool life, the coatings have to prevent the appearance of
tribological and mechanical failure mechanisms. This may take
effect, basically, by increasing the loadability of the tool or by

decreasing the stresses acting upon the tool.
Increasing the loadability of the tool may result directly, e.g.
through a coating induced improvement of the resistance against
abrasion and tribochemical wear of the tool surface. It is also pos-
sible indirectly, by applying more ductile, less wear resistant sub-
strate tool materials, when the coating takes over the surface related
wear protection functions.
As examples for the coating induced decrease of the stresses
impacting on the tool, avoiding the adhesion induced stress peaks in
the cutting wedge volume through coatings, with favorable friction
behavior, or reducing of thermal loads for the tool body through
thermal barrier effects of the coating, can be mentioned. The follow-
ing chapters describe the possibilities that can be achieved by CVD
and PVD coatings.
1.2 CVD-Coated Cutting Tools

Owing to the opposing behavior of important material proper-
ties for cutting (as indicated in Sec. 1), i.e., wear resistance and
ductility, it was attempted early to improve wear resistance of tools
through surface coatings. The deposition of hard materials in the
systems TiC and TiN with CVD processes was described in the
German patent literature in 1953 for the first time.[8][9] In 1957 the
first practical results of these coatings deposited on steel substrates
were presented.[10] Even then it was observed that the natural brittle-
ness of the hard materials could be encountered by reducing the
coating thickness; but it was not realized that the advantages of thin
hard coatings can only take effect on a very tough substrate made
from hardened high speed steel, or from cemented carbide.[11] This
discovery was first made in the years 1967–69 with TiC-coated
cemented carbide tools,[12][13] and lead to a push in developments for
machining of steel, and cast iron materials by suppressing the
tribochemical reactions of the WC-based cemented carbides with the
material of the workpiece. In addition, the resistance of the tools
against abrasive wear was improved by the high hardness of TiC.
The obvious inhibiting effects on the wear progress during application
was surprisingly evident even after local wear through the mono-
layer coatings, which were only about 3–5 µm thick.[11]
With these TiC monolayer coatings, which are often called wear
protective coatings of the “first generation,” problems arose through
the formation of brittle η-phases at the interface between coating and
substrate. This resulted in a distinct sensitiveness against impact and
strain of the cutting edges,[14] leading to a strong restriction concern-
ing the range of applications of coated cemented carbides.[11]
The η-phase formation at the interface could be largely sup-
pressed through improvements in controlling the carbon balance in
the CVD process around 1970. The resulting increase of the ductility
did not only improve the machining reliability for application condi-
tions with cutting force fluctuations, but also permitted the increase
of realizable film thicknesses up to ~ 6–7 µm (see Fig. 7a) and
therefore improving tool life time. In this “2nd generation” of CVD
coatings, in addition to TiC, coatings of TiN, Ti(C,N), and Al2O3
monolayers were deposited. Each of these coating materials has
special advantages and disadvantages, and developed a specific
spectrum of application.[11]
Especially for abrasive applications like the machining of grey
cast iron, TiC achieves good results and is distinguished by a low
wear of the top face owing to its high hardness, high strength at
elevated temperatures, and outstanding coating adhesion to the sub-
strate material.
In contrast, TiN shows a better chemical inertness, thus inhib-
iting diffusion, and has a lower tendency of adhesion to steel materi-
als, leading to a better cratering wear resistance at the top face. A
compromise regarding chemical stability and hardness of both bi-
nary hard materials TiN and TiC, which crystallize in the same
structure and are therefore totally miscible, is given by the ternary
Ti(C,N). Ceramic Al2O3 shows the lowest coating adhesion on
cemented carbide, but offers highest oxidation resistance and is
therefore best suited for machining at high cutting speeds.[11][15]
The next development stage of CVD hard coatings, the “3rd
generation,” began between 1972 and 1974, and was characterized
by the deposition of multilayer structures and gradient structures,
permitting the achievement of various aims.
Figure 7. Structure of CVD coatings of different development stages in

fractographs and schematics,[15] and in cross sections.[5] (a) TiC-monolayer
coatings (about 1970). (b) TiC/Ti(C,N)/TiN-multilayer coatings (since about
1975). (c) Ceramic-metallic multilayer coatings (since about 1983) left: TiN/
Ti(C,N)/Al2O3, right: TiC/Al2O3 /TiN. (d) Ceramic-metallic multilayer coatings
(since about 1983) left: TiC/Ti(C,N)/TiN/4x(Al-O-N/TiN) multilayer coating,
right: Ti(C,N)/4x(Al2O3/TiN) multilayer coating.
The combination of different coating materials, each especially

suitable for different applications as pointed out earlier, enlarges the
universal applicability of these coatings on one hand, and improves
the resistance against complex wear loads for suitable design of the
coating on the other hand. Therefore, in many cases, TiC is used as a

bond layer to improve adhesion, which is supposed to take abrasive
loads that softer layers could not withstand (Fig. 7b,c). In contrast
to that, coating materials showing diffusion inhibiting and oxidation
resistant properties resulting in better resistance against cratering
wear, such as TiN, ZrN, or Al2O3, are applied in outer layers of the
coatings (Fig. 7c,d). The transition between TiC and TiN is often
realized by one or more Ti(C,N)-layers arranged in steps or with
graded composition (Fig. 7b,d), providing a high cohesive strength
of the multilayered coating structure.[11][15][16]
Owing to the thinness of these individual layers, two effects
also add to the performance of the multilayer coatings under load.
One is the limitation of the grain growth through the thinness of the
coatings. This is an important means to improve coating qualities
produced by CVD processes which run at high temperatures
(>1000°C) and therefore favor the growth of diffusion-controlled
coarse-grain structures. Hardness, strength, and wear resistance of
bulk and coating materials are correlated by the Hall-Petch relation-
ship to the grain size of their microstructure. Figure 8 illustrates that
by comparing CVD and PVD TiN coatings. PVD coatings are
typically deposited at temperatures of about 350–500°C and show a
substantially finer grain size than CVD coatings, as becomes evident
in Fig. 9, and therefore they have higher hardness.[5][17]
The other effect results from the fact that small cross sections
of materials, e.g., in thin fibers or very thin films, show a signifi-
cantly higher strength, and an improved ductility is observed. Coat-
ings composed of various layers of thin films can consequently
follow the elastic and plastic deformations taking place at the cutting
edge of a machining tool better, i.e., without danger of cracking or
chipping off. Compared to monolayer coatings, the better ductility of
multilayer coatings, consisting of many thin films, permits higher
total coating thickness (8–12 µm) for many applications and there-
fore greater cutting wear volumes.[11] Typical exponents of these
CVD multilayer coatings of the “4th generation” developed in the
80s are presented in Fig. 7d.
Figure 8. Hall-Petch relationship in TiN coatings with varying grain size d

processed by CVD and PVD methods and their corresponding microhardness. [5]
Figure 9. TEM bright field and selected area diffraction patterns[5] (a) CVD TiN.
(b) PVD TiN.
Besides the possibility of producing multilayered structures by

depositing different coating materials, the thin film effects, i.e., limiting
the grain size and improving the strength, may also be applied by
deposition of coating layers with specifically optimized alternating
crystal orientations. It was found for the CVD deposition of TiC that
coatings produced with benzene show a (111)-preferred orientation,
while the use of heptane promotes a (200) texture. An alternating
change of the applied reactive gases produces the formation of a
multilayer coating having the same material composition in all
layers, but possessing alternating crystal orientations (Fig. 10). Such
coatings are called alternating coatings, and they show significantly
improved wear performance compared to monolayer coatings.[18][19]
Figure 10. TiC-heptane/benzene alternating coatings (chemically etched with

HF/HNO3/H2O).[18]
With complex structured CVD multilayer coatings in continu-

ous-cut machining, e.g., turning, a tremendous improvement in cut-
ting performance is achieved compared to uncoated tools or tools
coated with monolayers, shown exemplarily in Figs. 11, 12, and 13.
For high cutting speeds, these coatings with tribochemically inert in-
between or top layers (Al2O3, ZrN), are often combined with P- or
M-grade cemented carbides as substrates with high temperature
strength. For more difficult machining conditions with high or dy-
namically alternating cutting forces, higher requirements concerning
ductility, thermal shock, and abrasive wear resistance have to be met
by the coating-substrate composite. An attempt to meet these de-
mands combines thin (ca. 3–5 µm), less complicated structures but
high thermal conductivity coatings based on TiC/Ti(C,N)/TiN with

ductile M- and K-grade cemented carbides.[16]
It became evident that good results for tool life and satisfying
cutting reliability could not be achieved for CVD-coated cemented
carbides under machining conditions which require high tool ductili-
ties, e.g., milling of high strength steels (Fig. 14). Even by using very
ductile substrate materials, cutting edge chip-off took place after
short times of application, due to an embrittlement of the coating-
substrate composite resulting from the coating process. Therefore,
CVD-coated cemented carbides were unable to become generally
accepted for use in milling processes for quite some time.[1][5][17]
Figure 11. Cutting performance of uncoated and CVD-coated cemented carbides

for turning of steel.
Figure 12. Comparison of different CVD coatings for turning of steel.[16]

Figure 13. Comparison of different CVD multilayer coatings for machining of

grey cast iron.[20]
Figure 14. Performance of uncoated, CVD-coated and PACVD coated inserts in

disc type milling.[1]
The brittleness of the coating materials given by nature, the

partially coarse grained microstructure of the coatings, existing of
internal tensile stresses, and also the formation of brittle η-phases in
the interfacial zone, have been discussed as reasons for the occurring
embrittling effects of CVD-coated cemented carbides. Recent re-
search of this matter has shown that besides the inevitable formation
of η-phases at high coating temperatures, a temperature-induced
embrittlement within the whole cemented carbide body probably takes
place. This has led to different approaches to lower the temperatures
in CVD processes which are normally higher than 1000°C.[3]
Ti(C,N) coatings having a specified C:N-ratio can be deposited
at about 900°C by using C- and N-containing reactive gas. This leads
to a certain improvement in ductility and cutting performance in
interrupted cut machining compared to standard CVD processes.
Even further improvements are possible through plasma activation
of CVD processes, since that lowers the deposition temperatures to
about 700°C. Therefore, plasma activated (PA) CVD-coated ce-
mented carbides are significantly superior to standard CVD-coated
grades in interrupted cutting (Fig. 14).[21]–[25] Cutting material duc-
tility and reliability against breaking, as necessary for extreme stresses
and achievable by PVD coatings, however, can not be reached by
medium temperature (MT) CVD or PA-CVD coating processes.[26][27]
A further decrease of process temperatures down to PVD level
is not possible, since that will cause the chemical reaction to stop,
which leads to the deposition of hard materials in a MT-CVD
process. In PA-CVD processes this will result in a decrease of
coating adhesion due to less diffusion activities. Further, owing to
lower process temperatures (below 700°C), an increasing incorpora-
tion of chlorine coming from the metallic chlorides into PA-CVD
coatings, will result, leading to a significant drop in wear properties.
Although recent studies concentrate on improving coating adhesion
by suitably optimized surface activations, and avoiding the incorpo-
ration of chlorine by using metal-organic compounds, there are, so
far, no satisfying solutions to these questions, so CVD processes will
have only a limited applicability for coatings of cemented carbides
used in interrupted-cut machining.
For coatings of other substrate materials, the process tempera-

tures in CVD processes are also a problem. Ni-based cermets, like
cemented carbides, show significant embrittlement at high coating
temperatures.[28] Furthermore, in CVD processes, nickel diffuses at
these high coating temperatures from the substrate into the hard
coatings, which reduces their adhesion and wear resistance. Practi-
cally, high speed steels cannot be coated with CVD, MTCVD, and
PACVD processes due to their softening temperature of about 550°C.
The disadvantages of chemical vapor deposition, as described
above, lead to the fact that recently PVD processes are gaining more
and more competitiveness against CVD technologies. PVD process-
es take place at lower temperatures, typically around 350–500°C,
thus avoiding the above mentioned difficulties. PVD coatings for cut
machining tools are therefore described in detail in the following
section.
Besides the so far described classic CVD coating materials,
mainly applied for cut machining of iron materials, there are recently
also chemically vapor deposited crystalline diamond-coated cutting
tools in use. These coatings are characterized by unmatched values
of microhardness, and their correspondingly outstanding abrasive
wear behavior is better than that of polycrystalline diamond, because
the crystallites from CVD coatings are directly grown to one another
(Fig. 15), while the grains within powder metallurgically produced
bulk material are held together by a less tough binder material.
Furthermore, the diamond-coated tools are far less expensive than
tools with inlays of polycrystalline diamond or diamond single
crystal material.
CVD diamond coatings are especially suited for machining
processes with extreme abrasive wear loads, e.g., machining of green
compacts of cemented carbides, raw Al2O3 or graphite, and cut
machining of plastics or wood.[44] Also for machining aluminum
with high cutting speeds, diamond-coated tools have shown a very
good potential. Diamond is, however, not suitable for machining of
iron materials due to the affinity of carbon to iron.
Figure 15. Cutting edge of a diamond-coated cemented carbide indexable tip.

(CemeCon GmbH, Germany)
1.3 PVD-Coated Cutting Tools
Tools, especially cut machining tools, have been coated using

PVD processes since the early 80s. Here, especially, the magnetron
sputter ion plating (MSIP) process, the low voltage discharge process of
Balzers Co. and the cathodic arc evaporation, and their reactive
variants, have proven to be successful. Each of these processes has
specific advantages and disadvantages which will be briefly ex-
plained below.
The sputter process offers the possibility of depositing any
complex composition coatings onto virtually any type of substrate
material. Also, electrically nonconductive coatings can be deposited
and ceramic substrates may be coated when applying an RF dis-
charge. For the deposition of metallic hard materials onto metallic
substrates, however, the DC-sputter process is preferred for techni-
cal and economical reasons. For DC-sputtering the low degree of
ionization of the process and the strong dependency of the spatial
plasma distribution on the magnetic field are a problem. The low
energy densities permit the deposition of coatings with very little
effect on the substrate, but to some extent they are not sufficient for
the realization of high coating adhesion and high strength coating
structures. This shortcoming led to the fact that the sputter process
was not industrially accepted for a long time for tool coatings which
required a high degree of mechanical loadability of the coating-
substrate composite under any circumstances. The necessary quality
and reproducibility of coating adhesion for industrial applications
could first be guaranteed with plants that apply an intensive RF glow
discharge during the deposition initiating and adhesion-ensuring ion
etching stage, and that realize a high spatial distribution of the
plasma through an optimized arrangement of the magnetic fields.
In the Balzers process, the adhesion problem was solved through
multistage etching and heating periods. During the course of these,
the electron and ion bombardments of the substrates are specifically
controlled, allowing the coating of parts with different dimensions in
one coating batch. Regarding the selection of coating materials, the
flexibility of the low voltage arc discharge process is limited by the
fact that due to the evaporation from the liquid state, only pure metals
can be evaporated. Therefore, complex hard materials can not be
produced based on alloys, but with reactive gas mixtures. Further-
more, the deposition of electrically nonconductive coatings or the
coating of ceramic substrates is impossible so far.
The latter mentioned restrictions are also valid for the arc-PVD
process, however, by application of suitable cathodes, alloyed mate-
rials can be evaporated. Due to the high degree of ionization, usually
very good coating adhesion, properties of the coatings, and a highly
productive process performance are the special advantages of the
arc-PVD process. Even very large components, e.g., broaches up to
3 m long, can be economically coated applying the arc-PVD process.
Questions concerning droplets are so far unsolved. These liquefied
or solid macroparticles are emitted from the vaporizer and have
negative effects on the homogeneity of the coating. However, for
cutting tool applications they have proven to be without harm.
As stated repeatedly, a very significant characteristic common
to all PVD processes is the low deposition temperatures compared to
CVD processes. They range typically between 350 and 500°C and
permit the coating of HSS-tools without loss of hardness in this

substrate material. The coatings of high speed steel tools, especially
drills and end mills, but also plain milling cutters, broaches, and
other cutting tools, represent a wide range of applications in which
the PVD technology has quickly established itself during the 80s
(Fig. 16).
(a)
(b)
Figure 16. PVD-coated tools. (a) Drills and end mills with TiAlN-coating
(CemeCon GmbH, Germany). (b) Plain milling cutters and broaches with TiN-
coating (Arthur Klink Metallurgica LtdA, Brazil).
With PVD-coated tools, remarkable improvements of perfor-

mance were possible compared to uncoated tools. Important reasons
for that are the reduction of adhesion effects between tool steel and
workpiece, an improvement of abrasive wear resistance of the rela-
tively soft tool surface due to the metallic hard coatings, and an
improvement of the inertness against tribo-oxidation. In addition, an
improvement of chip flow results in relief of the cutting edges and to
some degree provides a thermal barrier effect. This counteracts the
softening of the heat treated microstructure. The sum of the protec-
tive effects not only permits a tool life which is several times longer
than that of uncoated tools, but also permits, in many cases, a drastic
rise of machining parameters which is several times longer than that
of uncoated tools.
Although research of PVD processes started with coating depo-
sitions on cemented carbide substrates, the significance of coating
temperature for this application has not been fully recognized for a
long time. On the contrary, when comparing PVD and CVD pro-
cesses, the better efficiency of CVD coating processes, owing to
their larger batch size and their outstanding wear performance in
smooth cutting, based on larger coating thickness and multilayered
coating structures, was placed into the foreground. Consequently,
PVD coatings were practically not applied for coating cemented
carbides until into the 90s. This situation changed when different
studies clearly indicated that embrittlements occurring with CVD-
coated cemented carbides, apparent under the extreme stresses of
interrupted, cutting can be avoided with PVD coatings.[5][16][17]
Figure 17 exemplarily indicates that for face milling. PVD-coated
cemented carbides are significantly superior to CVD-coated ce-
mented carbides for machining high strength steels.
For PVD-coated cemented carbides, the coating as well as the
interface zone and the substrate, possess a higher resistance against
crack formation. Within the coating a fine grained microstructure
(cf. Fig. 9b) and a favorable residual compressive stress state, due to
high growth rates, are responsible for that. More important for the
loadability of the whole composite, however, is that with PVD-
typical coating temperatures, at the interface and within the substrate
volume, no η-phases are formed or other embrittling material changes
take place.[3]
Figure 17. Performance of CVD- and PVD-coated milling inserts.[17]
Therefore, PVD-coated qualities show a totally different per-

formance behavior than CVD-coated grades in interrupted-cut ma-
chining. With PVD coatings, the mechanical loadability of the ce-
mented carbide cutting edges is not affected, quite opposite to CVD
coatings, as Fig. 18 shows, for cutting of heat treatable steel as an
example. Furthermore, the outstanding mechanical properties of
cemented carbide can only be effectively utilized through PVD
coatings. The adhesion-reducing effects of hard coatings, e.g., TiN,
prevent, in ranges of lower cutting speed, microweldings and the
formation of built-up cutting edges and thus avoid stress peaks and
mechanical overloading of the cutting edges, which will lead quickly
to break-out and chip-off compared to uncoated cemented carbides.
Furthermore the coating reduces the abrasive wear favored by
tribochemical reactions between WC-Co based cemented carbides
and workpieces containing iron, thus improving the performance at
higher cutting speeds.
The sum of the stated protective effects leads to a significant
expansion of the realizable range of cutting speeds in machining
technology and tool lives several times longer compared to uncoated
or CVD-coated cemented carbide qualities in interrupted-cut
machining (Fig. 18). Furthermore, and this fact is especially impor-

tant for automated production lines, PVD coatings of cemented
carbide give a high degree of cutting reliability in low ranges of
cutting speeds. These advantages of PVD coatings are utilized in an
optimal way when applying very ductile cemented carbide grades
that are not wear resistant if uncoated.[29]
Figure 18. Machining performance of uncoated, PVD- and CVD-coated ce-

mented carbides for interrupted cutting of heat treatable steel.
For the machining performance of PVD-coated tools, besides

the properties of the substrate cutting material and the coating adhe-
sion, the mechanical, thermophysical and tribological properties of
the coating material are also of great importance. These are deter-
mined by structure and composition of the coatings and may, in PVD
processes, be specifically influenced by settings of process param-
eters and the selection of evaporated materials and reactive gases.
Therefore a great diversity of metallic or ceramic, stable or meta-
stable crystalline or amorphous coating materials may be deposited.
The resulting great variability offers a wide spectrum of possible
applications for PVD technologies, within which the coating properties
can be specifically adjusted for particularly specified applications.[30]
The application of PVD coatings to machining tools was restricted

for a long time to a single coating material: TiN. This binary metallic
hard material (HV 0.05 ≈ 2400) is easily produced in PVD processes
and is characterized by a well balanced property profile, which makes
TiN rather universal for cut machining applications. Other binary,
nitridic hard materials like ZrN, HfN, or NbN showed no significant
advantages in cut machining applications[31][32] and their acceptance
in practical use was low due to the much higher prices of the cathode
materials. Recently (since about 1992) only chromium nitride has
proved successful for cut machining of nonferrous metals. Binary
carbidic hard materials, like TiC or ZrC, cannot be deposited easily in
PVD processes and show no favorable properties for practical use.
However, different research studies have already shown in the
mid-80s that with ternary hard materials, e.g., Ti-C-N or the meta-
stable Ti-Al-N, PVD coatings for specific applications with a supe-
rior cutting performance can be produced.[33]–[35]
Ti-C-N is characterized by a higher hardness (~ 3000 HV 0.05)
and a better abrasion resistance compared to TiN.[36] This has positive
effects for machining heat treatable or hardened materials. The optimal
machining conditions for Ti-C-N are in the area of low application
temperatures, e.g., low cutting speeds or interrupted cutting. For
higher cutting temperatures Ti-C-N has proven less successful. The
outstanding qualification of Ti-C-N for interrupted cutting is based
on a very favorable friction behavior at low temperatures and a high
heat conductivity, resulting in a chip flow which is smooth and easy
on the cutting edge.[37] Figures 18 to 23 show the results of machin-
ing tests, proving the superior performance of titanium-carbonitride.
Ti-Al-N also possesses a higher hardness compared with TiN
and is therefore especially qualified for machining of abrasive mate-
rials, e.g., grey cast iron or cast Al-Si-alloys. Ti-Al-N has, however,
in opposition to Ti-C-N, a lower heat conductivity and a higher
friction coefficient.[21][37] The resulting chip flow, characterized by a
small crater mean distance, leads in interrupted cutting to high
stresses upon the cutting edge, so that edge break-outs and edge
chipping may occur more easily. On the other hand, Ti-Al-N shows
a higher strength at elevated temperatures, higher oxidation resis-
tance, and better thermal barrier properties compared to TiN and is,
therefore, ideally qualified for abrasive cuts with high cutting tem-
peratures. Figures 24 to 27 show this with examples of turning,
drilling, and milling at high cutting speeds applying cemented car-
bide and HSS tools.
Figure 19. Performance of PVD-coated HSS inserts in face milling. [37]
Figure 20. Milling distance extension of HSS groove milling cutters through
PVD coating.[38]
Figure 21. Tool life of uncoated and PVD-coated solid cemented carbide shank
end cutters.[21]
Figure 22. Wear behavior of uncoated and PVD-coated solid cemented carbide
broaches.[21]
Figure 23. Performance of uncoated and PVD-coated solid cemented carbide

plain milling cutters.[21]
Figure 24. Performance of PVD-coated HSS inserts in turning of quenched

tempered steel.[37]
Figure 25. Superior performance of Ti-Al-N as compared to TiN in turning and

in milling with higher speed.[30]
Figure 26. Comparison of PVD coating materials in drilling of grey cast iron
with solid cemented carbide drills.[21]
Figure 27. Drilling of grey cast iron and of cast Al-Si-alloy with uncoated, TiN-
and Ti-Al-N-coated HSS.[30]
The increased hardness and stability of complex hard material

composites are due to the formation of mixed crystals by substitu-
tion, and this phenomenon also has positive effects on wear proper-
ties in other coating material systems. In various investigations,
ternary hard materials within the systems Ti-Zr-N, Ti-Hf-N, Ti-Nb-
N, Ti-V-N and Ti-Cr-N show performance advantages compared to
simple TiN (Figs. 28 and 29).[38]–[40] The same has to be said about
quaternary hard materials, e.g., Ti-Al-V-N (Fig. 30).[41] To what
extent these complex hard materials will find access in practical cut
machining operations is, so far, uncertain since the determination of
the performance characteristics of these different coatings becomes
extremely expensive and difficult.
Figure 28. Wear behavior of Ti-Nb-N- and Ti-Hf-N-coated indexable tips as

compared to TiN-coated and uncoated milling inserts.[39]
Figure 29. Performance of various PVD coating materials in continuous

cutting. [30]
Figure 30. Wear behavior of Ti-Al-V-N in dependency of the coating

composition.[41]
Questions arise because the type of PVD process and the

deposition conditions have influences on the wear performance as
well as the chemical composition. For nitridic hard materials, it was
shown that a high pressure of the reactive gas and resulting high
levels of stoichiometric saturation or even supersaturation improves
the wear performance. Figure 31 shows this exemplarily for Ti-Zr-N
coatings. The same can be said about Ti-C-N coatings, where the carbon
incorporation is not only determined by the reactive gas pressure but
also by the type of the carbon containing gas. Gases with a high
reactivity favor the incorporation of C, resulting in longer tool lives.
Figure 31. Cutting performance of arc-PVD Ti-Zr-N coatings in dependence of

reactive gas pressure.[42]
Also the residual stress state and the formation of microstruc-

tures in PVD coatings can be controlled by coating process param-
eters and have influence on the wear behavior. Compressive stresses
in coatings have proven to be positive for many wear and cutting
applications. Figure 32 illustrates that for the example of arc-PVD
Ti-Zr-N coatings, whose state of compressive stress is essentially
influenced by the process parameter bias voltage. The superior
cutting performance of the Ti-Zr-N coatings deposited with -100 V
correspond to maxima compressive stresses of about -4 GPa.
Recent developments in cut-machining orientated coating re-
search tend to optimize the cutting performance of PVD coatings by
multilayer structures. Here the same objectives are pursued that are
already state-of-engineering in CVD coatings, especially an increase
of the film thickness from now 3–5 µm to about 10 µm and the
development of multilayer coatings containing Al2O3 films. The
most important objective is to become competitive in smooth cut
operations which are now dominated by CVD-coated cemented

carbides. It is possible that the outstanding wear performance of
complex hard materials, that can not be produced by CVD methods,
will be helpful. Especially Ti-Al-N seems to have a promising
potential.
Figure 32. Performance characteristics of arc-PVD-Ti-Zr-N-coated cemented

carbide depending on substrate potential.[42]
Whether multilayer coating designs are really advantageous

regarding the cutting performance of PVD-coated tools will have to
be evaluated for each application. For machining of heat treatable
steel in interrupted cutting it was found that multilayer structures and
gradient structures had negative effects on the cutting performance
of PVD-coated cemented carbides. Reasons for the decrease of
performance of TiN/Ti-C-N multilayer coatings compared to Ti-C-
N monolayer coating, illustrated in Fig. 33, are probably the reduc-
tion of the thermal conductivity and a decrease of the resistance
against abrasive wear. Under different circumstances, however, the
complex structures can also show advantages. So it seems that for

cutting operations with a high load of tribo-oxidation the PVD
multilayer coatings with oxide top layers have a high potential.
Figure 33. Influence of the coating design on the performance characteristic of

arc-PVD-coated cemented carbide in interrupted cutting.[43]
1.4 Conclusions
The performance properties of cutting tools can be signifi-

cantly improved by CVD or PVD coating of cutting materials. The
metallic or ceramic hard materials, which are only a few micrometers
thick, increase the resistance against abrasive and tribochemical
wear of the cutting materials and decrease the mechanical and

thermal loads of the tools by reducing adhesive failure mechanisms,
by lowering friction between tool and workpiece/chip, and by influ-
encing the heat fluxes in the area of contact. Chip formation and chip
flow are optimized, so that in many applications higher machining
efficiencies and tool lives, which are several times longer than
without coatings, can be realized.
The state of engineering is set for CVD coatings mainly by
multilayer coating structures, based on the coating material sys-
tems Ti-C-N and Al-O-N, where the design of the coating is
determined for each specified application. CVD-coated cemented
carbides are very qualified for smooth cut operations with re-
quirements, which are not too high regarding ductility of the tool,
especially in turning. For cutting operations with higher fluctua-
tions of cutting force, as take place in milling, for example, CVD-
coated cemented carbides are less appropriate owing to their
reduced ductility. In Fig. 34, a survey of CVD coatings and their
applications is given.
The embrittlement of cemented carbide taking place during
high temperature CVD processes can be avoided by medium
temperature CVD or plasma activated CVD, so that cemented
carbide coated with these process variants can, to some extend, be
applied in milling operations. However, in strongly interrupted
cutting, for example, milling of heat treated or hardened steel,
even these do not achieve the outstanding performance of PVD-
coated cemented carbides.
PVD processes run at coating temperatures below 500°C
and permit not only the coating of cemented carbide or cermets
without loss of ductility, but also the coating of high speed steels.
The ranges of applications of the so-far industrially established
coatings (mostly deposited as monolayers or graded, binary or
ternary PVD metallic hard coatings) are summarized and pre-
sented in Fig. 35.
Application Thickness Substrate Coating composition
Turning of steel, high cutting speed, 12 µm CC-P15 Ti(C,N)-Al2O3 -TiN

(milling of low strength steel)
Turning of steel, high feed, 12 µm CC-P35 TiC-Al2O3 -TiN
Turning of steel, general, 12 µm CC-P20- TiC-[Ti(C,N)-Al2 O3]

(milling of low strength steel) P-40 Multilayer
Turning of cast iron 10 µm CC-K15 TiC-Al2O3
Turning of high strength steel, 8 µm CC-P25 TiC-TiN

turning of stainless steel
Turning of high strength steel, 8 µm CC-M15 Ti(C,N)-TiN

turning of stainless steel
Threading operation 3 µm CC-P25 TiN-TiC-TiN
Interrupted cutting 3 µm CC-P45 TiN-TiC-TiN
Machining graphite, plastics, wood, 5...10 µm CC or Diamond

aluminum, raw Al2O3 , cemented Si 3N4
carbide green compacts
Figure 34. Industrial CVD coatings for cutting applications.
Applications Coating material

Cutting performance
Outstanding very good good
High alloyed steels, low alloyed steels, Ti-Al-N Ti-C-N TiN

austenitic steels at medium and high
cutting speeds
High alloyed steels, low alloyed steels, Ti-C-N Ti-Al-N TiN

and austenitic steels at low cutting speeds
and for interrupted-cut machining
Aluminum alloys, cast aluminum, grey Ti-Al-N –––– ––––

cast iron, Al-Si-alloys
Copper alloys, brass alloys, bronze, CrN –––– ––––

aluminum bronze, titanium and
titanium alloys,
Nickel, Ni-alloys, hard alloys, Ti-Al-N –––– ––––

super alloys
Figure 35. Performance characteristic of PVD coating materials in cutting appli-

cations. (After CemeCon GmbH, Germany.)
REFERENCES
1. König, W., Fritsch, R., Zäher durch PVD, Industrieanzeiger 43, pp.
66–78 (1991)
2. König, W., Fertigungsverfahren, Bd. 1 Reihe Studium und Praxis,
VDI-Verlag, Düsseldorf (1981)
3. Krämer, G., Arc-PVD-Beschichtung von Hartmetallen für den
unterbrochenen Schnitt, Fortschritt-Berichte VDI, Reihe 5, Nr. 290,
VDI-Verlag, Düsseldorf (1993)
4. Müller, M., Zerspankraft, Werkzeugbeanspruchung und Verschleiß
beim Fräsen mit Hartmetall, Springer, Verlag, Berlin, Heidelberg,
New York (1987)
5. Quinto, D. T., “Overview on CVD and PVD Coated Carbide
Metalcutting Tools,” in: Plasma Surface Engineering, (Broszeit, E.,
Münz, W. D., Oechsner, H., Rie, K. T., Wolf, G. K., eds.), Vol. 1.,
DGM Verlag, Oberursel (1989)
6. Zum Gahr, K. H., Grundlagen des Verschleißes; VDI-Berichte Nr.
600.3, VDI Verlag, Düsseldorf, pp. 29–56 (1987)
7. Vieregge, G., Zerspanung der Eisenwerkstoffe, 2. Auflage, Verlag
Stahleisen, Düsseldorf (1970)
8. N. N., D. B. P., 954301 (1953)
9. N. N., D. B. P., 954564 (1953)
10. Münster, A., Z. angewandte Chemie 69 pp. 281–290 (1957)
11. Schedler, W., Hartmetall für den Praktiker, VDI-Verlag, Düsseldorf,
p. 205 (1988)
12. Rix, W. F. G., Dix, G. E., DOS 151166 (1966)
13. Ö. P., 295953 (1969)
14. Sproul, W. D., Richmann, M. H., J. Vac. Sci. Technol., 12:842–844
(1975)
15. König, U., “CVD-Beschichtung von Hartmetallen,” in:
Pulvermetallurgie der Hartmetalle, 7. Vorlesung, Verlag
Fachverband Pulvermetallurgie, Hagen (1992)
16. Reiter, N., Van den Berg, H., König, U., “Beschichtete Hartmetalle
zur Produktivitätssteigerung spanender Bearbeitung,” in:
Beschichtungen für Hochleistungsbauteile; VDI Berichte 624, VDI-
Verlag, Düsseldorf, pp. 167–184 (1986)
17. Reiter, N., Kolaska, H., Schneidstoffe, “Stand der Technik und
Entwicklungstendenzen,” in: 11th Int. Plansee Seminar’85,
Metallwerk Plansee, (Bildstein, H., Ortner, H. M., eds.), pp.
335–376, Reutte, (1985)
18. Leonhardt, A., Wolf, E., Schönherr, M., Seidler, M., Bartsch, U.,
Selbmann, D., Ein neues Prinzip der Herstellung
verschleißhemmender Vielfachschichten, Neue Hütte, 31. Jahrg.,
Heft 11, pp. 423–425 (Nov. 1986)
19. Leonhardt, A., Wolf, E., Schönherr, M., Seidler, M., Bartsch, U.,
Selbmann, D., Westphal, H., Voigt, K., “Zusammenhang von Gefüge
und Struktur CVD-abgeschiedener Hartstoffschichten und ihrer
Verschleißfestigkeit,” in: VIII. Int. Pulvermetallurgische Tagung,
Bd.3, Beitrag 36, Dresden, p. 197 (1985)
20. Schintlmeister, W., Wallgram, W., Kunz, J.; Gigl, K., “Application
and Manufacture of Tool Coated by Chemical Vapour Deposition,”
in: Proceedings of the 8th ICVM, Linz, pp. 293–309 (Sept./Oct.
1985)
21. König, W., Gerschwieler, K., Fritsch, K., “Leistung und Verschleiß
neuer beschichteter Harmetalle,” in: Beschichten und Verbinden in
Pulvermetallurgie und Keramik, Reihe Pulvermetallurgie in
Wissenschaft und Praxis, (Kolaska, H., ed.), VDI-Verlag,
Düsseldorf, pp. 161–203 (1992)
22. Küberl, E., Verbesserung der Zerspanleistungen von beschichteten
Hartmetall-Wendeschneidplatten durch ein neuartiges CVD-
Beschichtungsverfahren, VDI-Berichte, Nr. 670, p. 625 (1988)
23. Icks, G., Naßfräsen mit beschichtetem Hartmetall, VDI-Berichte
Nr. 762, p. 221 (1989)
24. König, U., Van den Berg, H., Tabersky, R., Sottke, V.,
“Niedrigtemperaturbeschichtungen für Hartmetalle,” in: Plansee
Proceedings’89, 12th Int. Plansee Seminar, (Bildstein, H.,
Ortner, H., eds.), Reutte, 3:13 (1989)
25. König, W., Fritsch, R., “PVD- und CVD-beschichtete Hartmetalle
im Leistungsvergleich,” in: Plansee Proceedings’93, 13th Int.
Plansee Seminar, Reutte, (Bildstein, H., Eck, R., eds.), 3:1–15 (1993)
26. König, W., Fritsch, R., “Physically Vapour Deposited Coatings on
Tools-Performance and Wear Phenomena,” Surface and Coatings
Technology, pp. 316–324 (1991)
27. Santhanam, A. T., Quinto, D. T., Gras, G. P., “Comparison of Steel

Milling Performance of Carbide Inserts with MTCVD- and PVD-
TiN Coatings,” in: Plansee Proceedings’93, (Bildstein, H., Eck, R.,
eds.), 13th Int. Plansee Seminar, Reutte, 3:31–44 (1993)
28. Knotek, O., Löffler, F., Krämer, G., “Performance Behaviour of
PVD-Coated Cermets in Interrupted Cut Machining,” paper presented
at the ICMCTF’93, San Diego (April, 1993), to be published in
Surface and Coatings Technology
29. Knotek, O., Löffler, F., Krämer, G., “Substrate and Interface Related
Influences on the Performance of Arc-PVD-Coated Cemented
Carbides in Interrupted Cut Machining,” Surface and Coatings
Technology, 54/55:476–481 (1992)
30. Knotek, O., Löffler, F., Krämer, G., “Multicomponent and Multilayer
Physically Vapor Deposited Coatings for Cutting Tools,” Surface
and Coatings Technology, 54/55:241–248 (1992)
31. Sproul, W. D., Bunshah, R. F., Kramer, B. M., Fenske, G. R.,
Kaufherr, N., Lee, R. H., “Characterization of Coating Wear
Phenomena in Nitride- and Carbide-Coated Tool Inserts,” Surface
and Coatings Technology, pp. 791–800 (1988)
32. Johnson, P. C., Randhawa, A., “Zirconium Nitride Films presented
by Cathodic Arc Plasma Deposition Process,” Surface and Coatings
Technology, pp. 53–62 (1987)
33. Leyendecker, T., Über neuartige Schneidwerkzeugbeschichtungen
auf Titan- und Aluminiumbasis; Dissertation, RWTH Aachen (1985)
34. Bosch, W., Über den Zusammenhang zwischen
Wachstumsbedingungen und Eigenschaften verschleißbeständiger,
kathodenzerstäubter Titanhartstoffe auf Sintermetall, Dissertation,
RWTH Aachen (1986)
35. Knotek, O., Leyendecker, T., Bosch, W., “Verschleißmindernde
Hartstoffschichten auf der Basis Ti-Al-C-N-O mittels
Kathodenzerstäubung,” in: Tribologie, Reibung, Verschleiß,
Schmierung, Bd. 9, Springer Verlag, Berlin, Heidelberg, New York,
Tokyo, pp. 229–274 (1985)
36. Bergmann, E., Kaufmann, H., Schmid, R., Vogel, J., “Ion Plated
Titanium Carbonitride Films,” Surface and Coatings Technology,
pp. 237–251 (1990)
37. König, W., Kammermeier, D., “Performance of TiN, Ti(C,N) and

(Ti,Al)N Coated Cutting Tools,” Materials Science Forum, 102–
104:623–632 (1990)
38. Böhmer, M., Kathodenzerstäubte Hartstoffschichten auf Basis Ti-
Zr-N-C-O und deren Stabilität auf Hartmetall und Gebrauchsstählen;
Dissertation, RWTH Aachen (1986)
39. König, U., “Untersuchung von kathodenzerstäubten binären und
ternären Hartstoffschichten zur Verschleißminderung von
Hartmetallen,” in: Tribologie, Reibung, Verschleiß, Schmierung,
Bd. 9, Springer Verlag, Berlin, Heidelberg, New York, Tokyo, pp.
275–334 (1985)
40. Randhawa, H., Gabriel, H. M., “Alternative Nitride Coatings by
Cathodic Arc Evaporation,” in: Plasma Surface Engineering,
(Broszeit, E., Münz, W. D., Oechsner, H., Rie, K. T., Wolf, G. K.,
eds.), 1:547, DGM Verlag, Oberursel (1989)
41. Knotek, O., Atzor, M., Jungblut, F., Prengel, H. G., “Wear Resistance
of Arc Evaporated and Magnetron Sputtered Coatings on Cemented
Carbides,” Surface and Coatings Technology, 39/40:445–453 (1989)
42. Knotek, O., Löffler, F., Krämer, G., “Arc-Deposited Ti-Zr-N
Coatings on Cemented Carbides for Use in Interrupted Cutting,”
43. Knotek, O., Löffler, F., Krämer, G., “Leistungsverhalten von
Mehrkomponenten und Mehrlagenbeschichtungen auf Hartmetallen
und Cermets im unterbrochenen Schnitt,” in: Plansee
Proceedings’93, (Bildstein, H., Eck, R., eds.), 3:236–250, 13th Int.
Plansee Seminar, Reutte (1993)
44. Leyendecker, T., Lemmer, O., Jürgens, A., Ebberink, J., “Einsatz
von kristallinen Diamantschichten auf Werkzeugen und
Verschleißteilen,” in:, pp. 215–225, Reibung und Verschleiß,
(Grewe, H., ed.), DGM Verlag, Oberursel (1992)
Coatings for Non-Cutting Tool Applications 411
Wear and Corrosion

Resistant Hard Coatings
For Non-Cutting Tool
Applications
J. Wesley Cox[a]
1.0 INTRODUCTION
The need to withstand wear and corrosion is an old and well

recognized problem that limits the useful life of engineered compo-
nents. Today, the needs for specialty materials and components are
escalating as their property requirements are becoming more de-
manding. It is one of the causes of high manufacturing costs, and
consumer prices. Indications of national impact are that annual
losses due to the sum effects of both wear and corrosion phenomena
in the United States exceed $300 billion and continues to grow.[1] In
1990 this figure represented approximately 7% of the Gross National
Product (GNP).[2]
a Currently Vice President Agile ® Manufacturing, Inc., Milwaukee, WI. For-

merly Assoc. Director, Basic Industry Research Laboratory, Northwestern
University, Evanston, IL.
411
Undoubtedly, wear and corrosion contribute to large economic

losses, and they often pose significant technical barriers to produc-
tion. They affect almost every aspect of our lives. We read daily
news accounts about some incident, often tragic, that has been
caused by wear and corrosion. They cause bridges to collapse, ships
to sink, airplanes to crash, automobiles to fail, and the list goes on
and on.
There are, however, several tactics to combat wear and corro-
sion. One direct approach is to construct components entirely from
specialty wear- and corrosion-resistant materials, but this can be
very expensive. Since wear and corrosion are surface phenomena, a
coating approach can be effective for minimizing costs and maxi-
mizing the performance life of components. With the performance
needs and operating conditions of today’s engineered components,
especially those conditions associated with high temperatures, hard
coating materials with inert and refractory properties are required.
Engineered materials like advanced ceramics, metal alloys, and
composite mixtures, including those with whisker and fiber rein-
forcements, are being used to reduce the influence of wear and
corrosion. Hard materials are usually taken to have Vickers (HV) or
Knoop (HK) hardness values ≥1000 kg/mm2. In this chapter, some
liberty has been taken with this loose definition.
There are numerous coating and surface modification methods
capable of applying coatings, such as those listed and classified by
Bunshah.[3] The four generic methods listed below have accounted
for the majority of hard coating applications for wear and corrosion
protection.
(1) Thermal Spray Coating (TSC)
(2) Chemical Vapor Deposition (CVD)
(3) Physical Vapor Deposition (PVD)
(4) Electro/Electroless Plating (EP/ElsP)
Components of almost every shape and size can be coated by these
techniques, from thread guides to bearing surfaces, from sporting
goods to surgical implants, using coating materials from simple
metals to complex ceramics. Choosing the coating method and

coating material for both performance and economy can be complex
and confusing.
In this chapter, combinations of hard coating materials, and
coating methods for non-cutting tool applications are presented as
examples of methods and materials to combat wear and corrosion.
Hopefully, the information presented here will contribute to reduc-
ing the complexity and confusion of selecting coating methods and
materials. Hard coatings for electronic and optical applications have
not been included.
2.0 WEAR AND CORROSION PHENOMENA
Although this chapter is concerned with preventing the harm-

ful effects of wear and corrosion, it is appropriate that mention be
made of the practical uses of these phenomena. Sanding, grinding,
polishing, and etching are all useful industrial processes that take
advantage of these phenomena. Even certain mechanical property
tests use wear and corrosion as the basis of measurement. The
scratch test, which is used to quantify coating adherence of thin
films, is a type of a wear phenomenon.
2.1 Wear
Wear involves the physical removal of material from a solid

object. Wear rates are less affected by temperature than is corrosion.
However, as the wear surface temperature approaches the softening
temperature of the substrate, wear rates increase dramatically. Wear
can be classified into three general categories: abrasive, adhesive,
and fatigue wear.
Abrasive wear can result from two surfaces rubbing together,
with the harder surface grinding away the softer surface. It can be
characterized by a rough appearance, due to the scoring or grooving
of the softer surface or by a relatively well-polished appearance, if
the two interface materials are similar in hardness. Often, work

hardening of the surface can take place. Thick, hard, well-adhered
coatings are effective in relieving this problem. Erosive wear, a
subset of abrasive wear, is caused by hard particles striking the
surface of a component with sufficient force that the surface material
is removed by a “microcutting” or gouging action in the areas of
impingement. It is characterized by a pitting appearance in the
impinged area. Hard and tough coatings can be effective deterrents.
Adhesive wear, like abrasive wear, is caused by loaded sur-
faces rubbing together. With adhesive wear, high localized tempera-
tures are created by friction at the tips of opposing asperities on
rubbing surfaces. These tips can deform and “weld” together due to
localized temperatures, only to break and fall away as debris. Be-
cause it is a welding process, adhesive wear can be considered a type
of chemical wear. As the interfacial load increases, the adhesive
wear rate increases. Often this is called thermoelastic instability
(TEI).[4][5] Adhesive wear is characterized by a non-oxidized, torn
surface littered with coarse metallic debris. Adherent, inert, hard
coatings with high thermal conductance can reduce adhesive wear.
Fatigue wear occurs whenever a surface is subjected to re-
peated high stress loads. With high loads on surfaces, as on gears and
bearings, cracks often originate at the surface and progress inward.
This is called contact fatigue wear. Eventually, fatigue cracks form
on or below the surface, and the surface material breaks away or
“pits.” Once initiated, pitting proceeds progressively until retarded
by the increased strength of a work hardened surface or by the
application of an adherent hard coating.
2.2 Corrosion
Many corrosion resistant coatings that are not considered to be

“hard” are quite effective when wear is not a concern. Such examples
are the sacrificial coatings of zinc and the self limiting coatings of
aluminum, not to mention polymers and paints. Corrosion is in-

cluded here, not because it requires a hard coating remedy, but
because wear is often accompanied by corrosion and for high-
temperature oxidation (corrosion) protection, the refractory and inert
coating materials used often display high hardness.
Corrosion can be regarded as a chemical degradation of a
component by its environment. It is a “chemical” material-removal
process that consumes and damages the surface of a component. The
severity of corrosion is controlled by chemical kinetics with higher
temperatures increasing the effect. However, “apparent” corrosion
rates may decrease or even fall to zero as a protective scale forms on
the surface of the substrate. In the latter case, corrosion can be self
limiting.
Corrosion may be classified as dry corrosion, wet corrosion,
and stress enhanced corrosion. Dry corrosion is a chemical process
that usually involves a gas environment and a solid surface. High-
temperature air oxidation is a common example. Gaseous halides
and sulfides also are notorious for their corroding effect on metal
surfaces.
Wet corrosion may be galvanic or nongalvanic. Wet galvanic
corrosion occurs in an aqueous medium with dissolved salts and is an
electrochemical redox process. In cases where no film deposit is
formed on the corroding surface, corrosion may proceed unchecked
until the entire component is consumed. Wet, non galvanic corrosion
may be a film-forming, non-redox, chemical-displacement reaction
occurring on solid surfaces within a liquid medium, or a dissolution
process where the solid surface dissolves into a liquid medium
without the formation of a protective scale.
Surface damage is intensified whenever corrosion is concen-
trated in localized areas. Applied stress and fatigue can enhance the
localized effect to promote corrosion in otherwise noncorroding
conditions. Some solid materials are susceptible to cracking, i.e.,
intergranular and/or intragranular cracking, whenever the surface of
the solid is exposed simultaneously to a corrosive environment and
an applied stress. This phenomenon is called stress-corrosion crack-

ing. In cases where variable loads or stresses are involved, the
phenomenon is called corrosion fatigue. These phenomena can cause
catastrophic failure of components operating well within the normal
load safety limits in low to mildly corrosive environments.
Continuous, chemically inert, well adherent coatings, free of
porosity and pinholes are effective for preventing pure corrosion
problems. Adherent coatings 1-µm thick can be effective, if the
coating material is continuous, compatible with the substrate, and
inert to the application environment. When variable stresses are
involved, a flexible coating maybe needed. Flexibility can be achieved
by depositing ductile metals or thin hard coatings, < 10 µm, of such
materials as metaloxide ceramics. A more rigorous description of
corrosion is given elsewhere.[8]
3.0 SELECTION OF COATING MATERIALS AND

COATING PROCESSES
Selection of a coating-material and coating-process combina-

tion for a specific substrate can be complex. There are a great number
of possible combinations, not all of which lead to satisfactory solu-
tions. Coating material selection is key to finding acceptable solu-
tions to wear and corrosion problems. It begins with the examination
of the damaged surface to reveal the nature of the degrading mecha-
nism, which can indicate the coated surface properties required to
provide a satisfactory remedy. With this information, both candidate
coating materials and coating processes can be selected.
Several excellent review articles of the materials selection
process for hard coatings have been published.[7][9][10] Confusion
can occur in selecting coating materials due to combinations of
properties that are difficult to optimize simultaneously, such as
hardness and toughness. Or perhaps a coating material has excellent

chemical inertness for corrosion protection, and therefore, may resist
bonding to the substrate; or a material may have high hardness and
inertness for high-temperature wear and oxidation protection, but it
also has a coefficient of thermal expansion (CTE) so different from
that of the substrate, that the coating will not remain bonded at high
service temperatures. Composite, composition graded or multi lay-
ered coatings can be utilized to optimize these properties.
Figure 1 serves as a material/process coating selection model.
The upper left box is a list of typical coating-material properties and
the upper right box lists coating-process properties. The combination
of the coating material and coating-process properties determines the
properties of the coated surface. It is important to keep in mind that
certain of the coated-surface properties can be affected more by the
deposition process than the coating material. These include adher-
ence, thickness, uniformity, residual stress, porosity, density, surface
roughness, microstructure, and to some extent composition.
Table 1 provides a general summary of coated surface needs
for several categories of wear and corrosion, along with candidate
coating processes, and materials. Coated surface property needs for
wear phenomena can be quite different from that needed for corro-
sion phenomena. Even different degradation mechanisms within the
wear category or the corrosion category, can require different coated
surface properties. It is important to keep in mind that corrosion
often occurs without wear, but wear is often accompanied by corro-
sion. In these instances, where wear and corrosion occur together,
the best coated surface solution is often a weighted compromise
considering both the coating material and the coating process.
Finally, and most important on a production scale, are the
economic considerations of both the coating process and the coating
materials.
Figure 1. Coatings selection model.

Table 1. General Summary of Surface Coating Needs for Wear and

Corrosion Applicationsa
Candidate
Categories Coating Requirements Coating Processes & Materials
WEAR
1. Abrasive High hardness, TSC alloys, cermets & ceramics;
med thick to thick, PVD & CVD ceramics; CVD/
low coefficient of TSC self-lubricating ceramic
friction. composites.
ElsP metal/ceramic composites.
(Erosive) High hardness, TSC alloys, cermets & ceramics;
adequate toughness CVD & PVD ceramics; ElsP
med thick to thick. metal/ceramic composites.
2. Adhesive Inert, high thermal ElsP metals/ceramic composites;
conductivity, hard, CVD & PVD alloys, cermets &
dense & lubricous. ceramics; CVD/TSC self-
lubricating composites.
3. Fatigue High yield strength, TSC metals, alloys, cermets;
(Contact Fatigue) hard, thin to thick PVD & CVD alloys & ceramics;
adequate toughness. ElsP metal/ceramic composites.
CORROSION
1. Dry Inert to environment, Paints & polymers; EP/ElsP
thin to thick, dense, metals & alloys; PVD, CVD,
continuous, and TSCc metals, ceramics & alloys.
nonporous.
2. Wet Inert to environment, Paints & polymers; EP/ElsP
(General) thin to thick, dense, metals & alloys; PVD, CVD,
nonporous.
(Electrochemical) Inert to environment, Paints & polymers; EP/ElsP;
thin to thick, dense, metals & alloys; PVD, CVD,
nonporous.
3. Stress Enhanced Inert to environment, ElsP metal/ceramic composites;
(Stress Corrosion hard, thin to thick, PVD, CVD, TSCc metals, alloys,
Cracking and continuous, dense, ceramics, cermets, & composites.
Corrosion Fatigue) nonporous, and
adequate toughness.
a Taken in part from Wear Resistant Surfaces In Engineering, Ref. 7.

b Well adhered coatings are assumed for all categories.
c Thermal spray coatings for corrosion applications may require sealants.
4.0 APPLICATIONS OF HARD COATINGS FOR WEAR

AND CORROSION
There are many examples of hard coatings on a variety of

materials, and components to combat wear and corrosion. The ef-
fects of wear and corrosion for cutting tools are well documented,
and are described in Ch. 7. There is an even larger number and
variety of applications needs associated with non cutting tool com-
ponents. Table 2 is a representative listing of some “traditional” non-
cutting tool components that often require hard coatings’ protection
against wear and corrosion.
Table 2. Traditional Non Cutting Tool Coating Candidates
Medical Implants Dies Wood Shredders Thread Guides

Surgical Tools Molds Propellers Slitter Blades
Prosthesis Gauges Landing Gear Spinnerettes
Dental Tools Boiler Components Turbine Blades Bearings
Dobbin Assemblies Motor Bushings Exhaust Components Bearing Races
Automotive Valves Ball Valves Fishing Reels Precision Gears
Combustion Liners Gate Valves Fishing Rod Eyelets Meat Cutters
Piston Rings Tomato Slicers Punches Eating Utensils
Needle Valves Onion Slicers Watch Castings Marine Trim
Pump Components Fasteners Steering Components Guide Vanes
In this section, traditional non-cutting tool application ex-

amples for the reduction of various types of wear and corrosion are
presented for the four coating methods listed in the introduction.
Within each one of the coating method/example groups, a descrip-
tive critique is provided for each of the generic coating methods. The
examples contained herein are not to be taken necessarily as the
“best-solution” recommendations for theapplications described, but as
examples illustrating the use of a coating material and coating

process combination. It is suggested that the reader refer back to
Table 1 as the various examples are encountered.
4.1 Applications with Thermal Spray Coating (TSC)

Methods
Thermal spray coating, a line of sight coating technology,

offers a number of method variations to deposit relatively thick
coatings, 50 microns to 0.25 inch or thicker, of hard materials to
combat wear and corrosion. Almost any material, from plastics, to
metals, to ceramics, that can withstand melting without decompos-
ing, can be thermally sprayed to form a coating. With the exception of
the transferred-arc methods, substrate temperatures rarely exceed 150°C.
Figure 2 shows that TSC utilizes flame combustion or electric
(arc or plasma) energy sources to form a region of high enthalpy. A
solid coating material, usually in powder, wire, or rod form, is
inserted into the high enthalpy region where it is converted into
molten or plastic state droplets. The droplets are accelerated toward
the substrate surface in the gas supply stream where the impacted
melt “splats” and rapidly cools to form the coating. The droplet
plasticity and velocity must be great enough to cause the molten
particles to mold and adhere to the topology of the substrate. Mul-
tiple-layered splats usually contain considerable porosity.
Advantages of TSC are associated with its economy of apply-
ing a wide range of coating materials on component surfaces, from
plastics to metals to ceramics. Size of components is limited only by
the capability of the handling equipment. Disadvantages are associ-
ated with low coating/substrate bond strengths and coating porosity
that can vary from 1 to 20 volume percent. Coating thickness usually
is limited by stresses at the coating/substrate interface arising from
thermal contraction of the coating during rapid cooling, although
thicknesses greater than 0.25 inches can be deposited to produce near
net shape forms on sacrificial mandrels.
Figure 2. Generic thermal spray coating schematic. (Reprinted with permission from TAFA, Inc.)
Hard coating materials, such as Al2O3, even applied by high-

temperature plasma, do not display the high hardness values gener-
ated by CVD or PVD methods. Plasma-sprayed coatings of alumina
rarely exhibit hardness values above 1300 kg/mm2, while dense and
well-adhered alumina coatings deposited by CVD and PVD methods
typically show hardness values above 2000 kg/mm2. Two major
reasons for this hardness reduction are: (1) thermally sprayed coat-
ings contain relatively high levels of porosity, and (2) the rapid
cooling rates (104 to 106 °C per second)[12] of the molten splats can
produce multiple, non-equilibrium, crystal phases (amorphous or
crystalline) of a single coating material. Fully dense coatings gener-
ally exhibit higher hardness values than porous coatings, and not all
crystal phases of the same material possess maximum hardness.
Porosity is often the more dominating factor of these two.
Feedstock particle sizes typically range from 50 to 150 mi-
crons. Fine powders, ≤50 µm, can be very conducive to higher spray
velocities, (i.e., higher densities) with relatively short nozzle to
substrate (standoff) distances. However, due to the higher cooling
rates of the resulting fine molten splats, high residual stresses that are
produced at the coating/substrate interface often result in poor adher-
ence. Coarser molten splats from coarser particulates cool more
slowly and produce lower stress levels, but also produce higher
levels of residual porosity. In one case, coarse feedstock materials
produce lower coating/substrate interface stresses that enhance ad-
herence, but they also produce higher porosity levels, which lower
the hardness. In the other case, fine feedstock materials can result in
higher densities or higher hardness values (if no lower hardness,
non-equilibrium, crystal phases are formed), but due to higher re-
sidual stresses, the coating may spall or peel away.
Adherence of the coating is usually classified as a mechanical
bond, although on pristine-clean surfaces and in inert coating envi-
ronments, chemical bonding may be achieved. Surface roughening
by grit blasting or etching is often required to maximize surface to
coating bonding. Well-adhered plasma sprayed ceramic coatings on
metal substrates usually require an intermediate bond coat. The
bond-coat material, often an MCrAlY, is chosen to compensate for

both the CTE mismatch and chemical incompatibility between the
metal substrate and the ceramic top coat. The effect of the bond coat
is to increase adherence and minimize interfacial stresses.
TSC techniques can be divided into two major categories based
on their thermal energy source: combustion methods and electrical
methods. Figure 3 subdivides these categories into subsets to include
designs that optimize other key parameters, such as particle velocity
and spray environment.
Figure 3. Thermal spray coating methods. (Courtesy of Basic Industry Research

Laboratory, Northwestern University.)
Combustion methods use oxygen/fuel gas flames to melt the

feedstock material. With the oxygen/acetylene flame spray gun,
temperatures of 3000°C can be achieved. The incoming and expand-
ing burning gases produce molten drop velocities from 100 to 1000
ft/sec.[11][12] Flame-spray techniques utilize wire, rod, or powder
feedstocks.
The D-gun (detonation-gun) technique is a higher velocity
combustion method in which both the powder and the fuel-gas
mixture are introduced into a “gun barrel” type of chamber and
ignited. The resultant explosive wave propels the powder toward the
substrate surface at impact velocities of about 2400 ft/sec.[12][18] This
process is repeated several times per second. Because of the higher
velocities, both bond strengths and densities of the coating are often
higher than those with conventional flame guns.
The (HVOF) high-velocity, oxygen fuel, flame-spray tech-
nique is an even higher velocity combustion method. It uses a special
“sonic” nozzle to maximize the flame temperature and to produce
particle velocities from 2000 to 3200 ft/sec.[11] With supersonic
particle velocities, high coating densities and excellent adherence
can be achieved.
Electrical methods are higher-temperature coating processes
than the combustion methods. With electric wire arc spray, consum-
able electrodes that serve as the coating feedstock material are fed
into the gun where a high temperature electric arc is formed to
convert the electrode material into melted droplets. A stream of high-
pressure gas flows through the gun, atomizes, and propels the melted
electrode droplets to the substrate surface. Temperatures can range
from 2200°C to more than 5800°C, with particle velocities from 800
to 1100 ft/sec.[11] The electrode feedstock is usually a metal wire or a
hollow metal-sheathed wire filled with a ceramic or cermet composi-
tion. With electric wire arc processes, excellent adherence can be
achieved, though porosity is often greater than with standard flame
spray.
Air-plasma spraying (APS) is a class of even higher temperature
techniques. With plasma energy sources, the plasma temperature can
exceed 30,000 K, a temperature hot enough to melt virtually any
ceramic.[11][12] The high-temperature plasma forms as a result of a
DC arc generated between an internal cathode and anode within the
gun. Both electrodes are water cooled. As the supply gas is passed
through the arc, it super heats, ionizes, and transfers heat to the
feedstock particles injected into the gas stream. With APS guns,
particle velocities range from 1500 ft/sec to 2300 ft/sec, which are
achieved by the rapidly expanding spray gases.[11] Virtually any
hard, refractory, ceramic material can be deposited by this method.
Figure 3 shows another class of plasma spray that is conducted

in a controlled atmosphere. Controlled-atmosphere plasma spraying in
both low-pressure and inert-gas chambers has extended the versatil-
ity of the method to create high purity hard coatings, especially with
the non-oxide ceramics and cermets. For example, with low-pressure
plasma spray (LPPS), it has been reported[14] that the brittle, low
carbon, and undesirable eta phase, Co3W3C, can be avoided in
sprayed WC/Co compositions, in contrast to APS techniques where
eta phase is formed due to carbon losses by air oxidation.
A lesser known method, liquid-generated plasma spray (LGP),
can provide exceptionally high deposition rates. Coating rates of 100
lbs/hr can be achieved with ceramic materials such as Al2O3 and
yttria stabilized zirconia to produce large, near-net-shape compo-
nents on mandrel substrates.[15] With LGP, a DC plasma arc is
established between an anode and cathode, with the arc residing
within the hollow vortex of a swirling liquid, usually water. Water
serves to supply the gas feed by decomposing into hydrogen and
oxygen, and, at the same time, the liquid maintains the electrode
surfaces and other materials of containment within tolerable tem-
perature ranges. With this arrangement, long, high-temperature pro-
cessing times with continuous high-material-rate throughput, can be
achieved. Large free standing shapes of alumina, zirconia, and titania
have been fabricated by this method.[15] An example is the produc-
tion of high-temperature reactor tubes formed by LGP spraying
alumina 5-mm (0.20 inch) thick directly onto tubular metal mandrels
to produce continuous tubes 6-meters in length and 35-mm in diam-
eter. Differential shrinkage between the metal and the ceramic al-
lows the formed tube to be removed from the mandrel.[15]
Hard ceramic coatings of oxides, carbides, nitrides, and borides
are well suited for abrasive and adhesive wear applications not
involving great impact or cyclic loading. For such applications,
metal-bonded ceramic materials or composites, sometimes called
cermets, such as cobalt bonded tungsten carbide or nickel-bonded
chromium carbide are ideal. The metal matrix provides toughness or
ductility, while the dispersed carbide particles provide excellent
wear resistance. These materials can be deposited directly onto metal

surfaces by combustion, plasma, or low-pressure plasma techniques.
By far, the major use of TSC in the USA has been for applica-
tions in gas-turbine engines for both aerospace and stationary indus-
trial needs. Approximately 15 pounds of ceramic and cermet hard
coatings can be found in the modern jet-aircraft engine.[16] These
applications typically include D-gun and plasma gun coatings of
carbide cermets such as WC/Co and Cr3C2/NiCr onto the midspan
shroud of turbine blades (to minimize impact damage as the shrouds
touch one another during operation) and coatings of nitride, oxide,
and carbide cermets on blade tips of compressor rotors (where high-
temperature oxidation and erosive wear are predominant).[17]
In high-temperature gas turbine and diesel engine applications,
thermal-barrier coatings (TBC) applied by plasma spray methods are
used to extend the thermal application range of base metals. TBC’s
are refractory and stable ceramics, usually stabilized zirconia, that
possess low-thermal conductivity values in order to act as thermal
insulators. Actual diesel engine applications,[18] for plasma sprayed
stabilized zirconia coatings deposited over NiCoCrAlY or NiCrAlY
bond coats onto components such as valve faces, piston crowns, and
cylinder heads have proven to perform well in extended performance
tests. However, TBC materials such as zirconia are not recom-
mended for high-temperature applications that include sliding adhe-
sive wear. Loaded sliding surfaces of hard refractory materials with
low thermal diffusibilities can create high frictional heating on local
asperity tips protruding from the surface. With materials of low
thermal diffusibility, localized heating is intensified with sliding
because the generated heat is not dissipated from the local areas fast
enough. This, in turn, causes the hot asperities to expand and forces
the asperities to carry increasingly higher portions of the applied
load. This progressively unstable situation, called thermoelastic in-
stability (TEI),[4] rapidly deteriorates, causing adhesive wear, crack-
ing, and spalling of the coating. Studies by Dow, et al.,[5] have shown
that the equation below can be used to calculate a material dependent
critical sliding speed at which TEI is expected to begin.
4 Kdc
Vc =
Eq. (1)
(uα E )2 π z
where: Vc = critical sliding speed E = Youngs modulus

K = thermal conductivity z = width of slider
u = coefficient of friction d = density
α = thermal expansion coefficient c = specific heat
This model has been confirmed by Dufrane,[6] where several TBC

materials were ranked in the predicted order of sensitivity to TEI in
high temperature sliding wear experiments. In that work, SiC and
Si3N4 were shown to avoid TEI as predicted. This relationship can be
useful for material selection.
4.2 Applications with Chemical Vapor Deposition (CVD)

Methods
Chemical vapor deposition represents a group of vapor coat-

ing methods that are especially useful for depositing hard, refrac-
tory materials—from ceramics to metal alloys and intermetallics
to carbon and diamond—onto substrate surfaces. Several excel-
lent reviews are provided elsewhere. [19]–[24] CVD processes are
associated with medium-to high-temperature and atmosphere to
soft- vacuum coating conditions. Because most CVD methods do
not require high-vacuum system configurations, equipment costs
are relatively low. Initial capital equipment expenditures for con-
ventional CVD systems range between the cost of manual, ther-
mal-spray systems and the cost of high-vacuum PVD systems.
Interest in CVD stems from the fact that it is an omnidirectional
coating process that produces theoretically dense, uniform, and
well- adhered coatings with controlled microstructures at reason-
ably high coating rates. The primary disadvantage of CVD is
associated with high temperatures that can limit the choice of
substrate materials. Table 3 lists representative CVD methods
with their dominant coating characteristics.
Table 3. Representative CVD Methods And Characteristics For Hard Coatings (Courtesy of
Basic Industry Research Laboratory, Northwestern University)
CVD processes generally involve a thermally activated chemi-

cal reaction or decomposition of gas precursors to produce uniform
and dense coatings onto heated surfaces contained within an en-
closed vessel. Coating rates can be controlled from nanometers to
multiple microns per minute by variations in temperature, pressure,
and gas-flow rates. Gas precursors may range from elemental forms
and metal halides to organics and organometallics. With CVD’s
omnidirectional coating property, there is no restriction on the shape
of the substrate. All accessible surfaces, including surfaces within
blind holes, internal surfaces of porous bodies, and surfaces of
powders and fibers can be coated by CVD. Coating microstructures
can be controlled from amorphous to crystalline forms and within the
crystalline forms, columnar to equiaxed structures. Microstructural
control can be accomplished through variation of the coating condi-
tions and choice of precursors.[12][17][45][46] Microstructural control is
especially important in optimizing the mechanical properties of the
coating. Improved fracture toughness can be achieved by forming a
fine grained and equiaxed microstructure or by codeposition of a
second phase which can act as a crack deflector within the coat-
ing.[26][27]
The high coating temperatures associated with CVD, often a
distinct disadvantage, also can be an advantage. These high coating
temperatures, which can cause structural distortion and lower sub-
strate hardness, also promote diffusion bonding between the coating
and the substrate, not obtained by lower temperature processes.
Coating/substrate interfacial stresses are minimized by diffusion
bonding.
CVD coatings can be deposited at lower substrate temperatures
(200°C–800°C) with the use of reactive, metal-organic, gas precur-
sors. With metal organic precursors, the process is called metal-
organic CVD (MOCVD). MOCVD has been used primarily in
electronic and optical applications, and not so much for hard coat-
ings for pure wear and corrosion prevention. At lower coating tem-
peratures, diffusion bonding usually is sacrificed and coating adher-
ence relies on chemical and mechanical bonding.
Energy sources other than direct thermal heating, can be used

to reduce the high-temperature demands on the coating substrate.
Electric discharge, RF, and microwave plasmas have been used to
reduce substrate/coating temperatures to less than 350°C. Laser
sources, operating in either a photolytic or direct pyrolytic thermal
mode, also have been used as CVD energy sources. Pyrolytic laser
techniques can deposit diffusion bonded coatings with localized,
direct heating of the substrate that can minimize distortion and
subsurface grain growth of the substrate. The high-temperatures
required for diffusion bonding of the coating to the substrate can be
limited to the first few microns of the surface. Several articles
describing laser CVD are referenced at the end of the chapter.[28]
An example of a wear coating applied by medium-tem-
perature CVD is the deposition of mixed carbides of tungsten
(W 3C + W 2C + W) deposited from the vapors of WF6 and dimethyl
ether in a H2/Ar atmosphere at temperatures between 350°C and
550°C, and at pressures between 40 torr and 100 torr.[29] No WC is
formed under these conditions. The substrates, coupons of AM-350
stainless steel and Ti/6Al/4V, were plated with a nickel barrier layer
prior to the CVD step to prevent reaction of the substrate with HF, a
by-product of the deposition reaction. Hardness values above 2000
VHN25 with a coating adherence exceeding 8000 psi by the standard
ASTM pull test have been reported. Comparative ball-on-disk wear
tests and erosive sand blast tests using alumina grit showed that the
coated coupons had 3 to 10 times greater wear life than the uncoated
specimens. It is important to note that a fine grain, outer carbide layer
was produced under the reported conditions with dimethyl ether
serving as the carbon source. Under similar conditions using CH 4,
C2H2, or C3H6 as carbon sources, a coarse, columnar grain structure
has been observed.[30] This variation in grain structure is an example of
the influence of precursor choice on the coating microstructure.
Hard coatings deposited by CVD methods, for the prevention
of wear and corrosion, generally are deposited on substrates that can
withstand the high coating temperatures. However, some interesting
trade offs can be accommodated with “low-temperature substrate
materials.” For example, common metal alloys such as the AISI D,

H, M, and T tool, and air-hardened steels, as well as the martinisitic
stainless steels can be good substrate candidates for CVD coating
methods,[31] if dimensional tolerances of components of these com-
positions are greater than 0.001-in (0.03 mm). Because of low
austenitizing temperatures, lower-alloy steels like the A, W, O, and S
series are not good candidates for CVD. Other more exotic, high-
temperature alloys, such as the inconels, hastelloys, waspalloys,
nimonics, and others make excellent CVD substrate candidates.
Steels hardened for wear life often can accommodate small hardness
decreases, and still maintain an adequate hard support base for the
high-temperature CVD hard coating. The harder, diffusion-bonded
coating can provide improved wear resistance due to its harder
surface, while the substrate can provide increased fracture toughness
due to its hardness reduction. This trade-off can be very useful.
The utility of high temperature CVD with steel substrates has
been described by Hintermann, et al.,[32] in an example that involves
the prevention of sliding, abrasive wear on a bearing steel. For this
application, a 10-micron coating of Cr7C3 was deposited at 1000°C
using CrCl2, H2, and Ar gases, onto the wedge pieces of a reverse lock
clutch assembly. The wedge pieces were constructed from a DIN
100Cr6 bearing steel containing 1% carbon. Note that the carbon
source in this example was supplied by the substrate steel and not by
a gas precursor. While the clutch shaft rotates freely in the forward
direction, normal sliding wear occurs between the wedge pieces, and
the clutch casing. Rotation of the shaft in the reverse direction causes
the wedge pieces to lock firmly together, preventing further shaft
rotation. This action creates high mechanical shear stresses on the
contact surfaces between the wedge pieces. Not only is a hard coating
such as Cr7C3 (HV = 2300 kg/mm2) mandatory to resist sliding wear,
but great adherence between the coating and the steel substrate is
required to withstand the severe shear stresses. Once coated and heat
treated, the wedges have been reported to have an operating life
increase of a factor of 20 over that of uncoated wedges. Typical
uncoated wedges were reported to have an operating life of 200 hours,
while the coated wedges have displayed a use life of 4000 hours.
In several examples of high temperature CVD coatings applied

to high-temperature substrates, Hintermann and others[32]–[34] have
demonstrated the utility of several hard coatings to combat wear in
severe roller bearing applications. Substrates included cemented car-
bide, AISI 440C, and DIN 100Cr6 with coatings of TiC, Ti(C,N),
TiN, Al2O3, Cr7C3, and combinations thereof. In one example, ce-
mented carbide ball bearings, 52-mm diameter, with a 2.5-micron
coating of TiC, provided an operating life of more than 107 revolutions
under a 50-kg load at 100 rpm and 300°C in an inert, helium atmosphere
without lubrication. Uncoated, cemented carbide ball bearings failed
to withstand more than 2 x 106 revolutions in the same environment.
Titanium diboride has been used as a successful deterrent
against erosive and abrasive wear. Fifty micron thick TiB2 coatings
with Knoop hardness values of 33 GPa (3365 kg/mm2) were depos-
ited by CVD at a temperature of 900°C on cemented carbide valve
and pump components for use in coal liquification and gasification
plants. Comparative test results after a 1-hour exposure to a hot
impinging, coal oil slurry showed little erosion (0–3 µm crater depth)
on coated substrates, as opposed to softer coatings, and uncoated
substrates that showed less erosion resistance (10–30 µm crater depth).[35]
Hard and dense CVD coatings have been of significant impor-
tance in resisting erosive wear on blade tips of gas turbine engines
from high velocity impact of particulates. Titanium carbonitride
(TiCxN1-x) has been reported to extend the life of turbine blades
against erosive wear by a factor of 70 to 80.[36] Other erosive wear
life increases have been observed with rocket nozzles and letdown
valves in fossil energy systems.[35]
Coatings used only for corrosion resistance do not require great
hardness unless wear phenomena also are involved. Often how-
ever, inert, hard and dense coatings of refractory, ceramic
materials deposited by CVD are called upon to retard high-tem-
perature corrosion (oxidation) or high-temperature materials-com-
patibility problems. One dramatic materials compatibility example
was associated with the protection of a niobium-sheathed fuel pin
located in the core of a liquid sodium, cooled nuclear reac-
tor. [37] In this case, there was a concern that a momentary loss of the
liquid sodium coolant could lead to a very rapid melting of the steel
housing that held the fuel pin. If this situation developed, the radio-
active fuel pin would be destroyed, thereby releasing fissionable
material, by forming a Fe-Ni-Nb eutectic with the melted steel.
Deposition of a multilayered coating of NbC (1 µm), overcoated
with layers of TiC (1 µm), and TiN (2 µm) was able to suppress the
formation of Fe-Ni-Nb eutectic at temperatures up to 1660°C, while
remaining inert to liquid sodium (below its boiling point 892°C).
This coating has proven to be a solution to prevent possible destruction
of the fuel pin until the excess heat can be moderated by other means.
Other CVD hard coatings have been used as protective coat-
ings on high-temperature reaction vessels to resist oxidation. Oxide
coatings, such as Al2O3 and HfO2, can be deposited on the surfaces
of graphite vessels, for use in high-temperature oxidizing environ-
ments, while CVD SiC has been used for protective surfaces on heat
exchangers.[38]
With CVD methods, it also is possible to create fully dense
freestanding shapes in a one step operation. Free standing, near net
shapes of cylinders, crucibles, etc., have been fabricated directly by
the CVD coating of thick, dense, and continuous layers onto sacrifi-
cial mandrels,[21] which can be removed by grinding or chemical
etching. However, with the appropriate match of CTE values of the
coating and mandrel materials, the mandrel may be removed without
damage and reused. As an example, boron nitride has been deposited
onto cylindrical shaped graphite mandrels, which were removed
easily after cooling and subsequently reused.[39]
Because of the omnidirectional coating property that allows for
the coating of all exposed surfaces accessible to circulating gas
precursors and the wide range of solid-coating materials that can be
formed on these surfaces from vapor phase chemical reactions, CVD
offers several unique fabrication routes to produce composite pow-
ders, continuous fibers, and infiltrated fiber reinforced composites.
Composite powders may be used as feedstock materials for thermal
spray coatings or conventional press and sinter operations, and fiber
reinforced structures can be used in high-temperature oxidizing
environments requiring high strength.
As a general example of composite powders, hard and refractory

ceramic materials may be deposited onto individual particles contained
in a heated, gas-fluidized bed of powders. By injecting reactive, precur-
sor gases into the fluidizing-gas stream, deposition occurs upon all
surfaces exposed to the flowing gases within the fluid bed vessel. In this
manner, all the particles contained within the moving bed can be
individually encapsulated with a coating of uniform thickness following
the contours of each particle. This process is called fluidized bed
CVD (FBCVD), and can be operated in both atmospheric-pressure
or soft-vacuum environments. Furthermore, the burnishing action of
the bed particles on all surfaces contained within the bed continually
interrupts the nucleation process, thereby preventing the occurrence of
exaggerated epitaxial or columnar grain growth so often associated with
conventional CVD processes. With FBCVD, small, equiaxed grain
structures can be formed even at relatively high temperatures. This
method can be used to control the coating microstructure.
FBCVD has been practiced for over 40 years, and first reached
prominence with the deposition of pyrolytic carbon, Nb, W, Al2O3,
and BeO onto fissionable, fuel particles of radioactive uranium and
uranium compounds.[40] Figure 4 shows a generic representation of a
FBCVD vessel. FBCVD offers the possibility of creating encapsu-
lated hard, and refractory composite powders that can be used for a
variety of applications from thermal spray powder feedstock materi-
als, to powders for traditional press and sinter operations. Practical
particle diameters may range from 30-µm to 1/8-in. The lower limit,
which is related to the system’s ability to separate particles from the
gas stream so that the powder is maintained in the coating vessel, is
dependent upon particle density and gas flow rates. The upper limit
is more dependent upon the practical gas velocity required to sus-
pend larger shapes. Components larger than 1/8-in-diameter par-
ticles, such as bolts, fasteners, thread guides, and ball bearings, can
be coated in the fluid bed reactor by immersing them in an inert
powder bed whose bulk density is reasonably close to that of the
component substrates. In this arrangement, batches of nominal-sized
components can circulate through the fluidized bed as if they were
giant particles. Components can be loaded directly into the bed
without the intensive hand labor required for placing components on

staging plates associated with other coating methods. Components
with surfaces sensitive to bumping contact with other components,
such as edges and corners that can be chipped, can be loaded into
restraining baskets that in turn are loaded into the coating chamber.
Larger components can be suspended in the bed by wire hangers. As
before, due to the interruption of the nucleation process by the action
of the bed particles on the component surface, fine grained and
equiaxed microstructures are produced. This action also ensures a
uniform coating as opposed to uninterrupted, vapor-coating pro-
cesses that can produce an exaggerated, epitaxial growth of dendritic
forms on surface sites of abrupt contours such as edges and corners.
Figure 4. Fluidized bed CVD vessel schematic.

CVD has come to play an important role in the fabrication of

lightweight, high-strength, space-age materials that are stable in high
temperature oxidizing environments. The need for materials to per-
form successfully at high operating temperatures (> 1000°C) trans-
lates into high-efficiency, high-productivity, and high-energy saving
operations for rocket engines, gas-turbine engines, and internal-
combustion engines. It also includes high-temperature structural
components for devices such as waste incinerators, hot-gas filters,
and heat exchangers. Though there are several classes of lightweight
and high-strength materials, only the refractory ceramics possess
high strength (though hard and brittle) at very high temperatures and
in oxidizing environments. Ceramic matrix composites (CMC), rein-
forced with high-temperature, continuous ceramic fibers, have the
ability to avoid catastrophic brittle failure by imparting a more
ductile, metal like, graceful failure to structures so constructed. This
graceful, non-brittle, failure behavior is due to crack deflection and
crack bridging effects of the fibers within the brittle ceramic matrix.
It is the interface region between the ceramic matrix and the fiber
surface that is critical to the performance of the composite. A bond
between the fiber and the matrix which is too strong results in a
matrix crack proceeding across the fiber to produce brittle failure. A
bond between fiber and matrix which is too weak does not allow for
proper load transfer from the matrix to the reinforcing fiber, which
results in weak structure. Thin (≈ 200-nm) single-and multiple-
interfacial layers have been deposited by CVD to provide a crack
deflecting “debond” layer to promote the fiber pull out failure mecha-
nism. Debond layers of pyrolytic carbon and boron nitride have been
used for this purpose. Boron nitride generally is preferred due to its
higher resistance to oxidation, as compared to carbon. Carbon often
has been used as a high-strength fiber reinforcement in both carbon/
carbon and CMCs. SiC has been used as an oxidation barrier both for
carbon fibers and for the outer surfaces of carbon/carbon composites.
CVD is one of the methods used for the production of continu-

ous fibers that serve as the load-bearing reinforcements for CMCs.
Starting with small diameter (5–30 µm) continuous core fibers such
as carbon or metal filaments as Mo or W, etc., ceramic coatings are
deposited continuously on a moving fiber core within a tube shaped
vessel so as to surround the core fiber to produce a larger diameter
ceramic fiber. With a single pass CVD reactor, ceramic fibers can be
produced at rates of 5 to 10 ft/min and higher. Commercially,
continuous monofilament fibers have been fabricated by CVD meth-
ods on core filament materials. These include boron fibers fabricated
on tungsten cores and silicon carbide fibers fabricated on carbon and
tungsten cores.
In addition to producing fibers and providing fiber/matrix
interfaces, CVD has been used to deposit hard and refractory ceramic
matrix materials into the porous preforms of fiber filaments.[41] This
process is called chemical vapor infiltration (CVI). CVI is simply the
deposition of coating materials onto the internal surfaces of a struc-
ture with open porosity. For open porosity fiber preforms, coating
material is deposited on all accessible fiber surfaces (both internal
and external surfaces of the preform) as the gas precursors flow
across these surfaces. As the coating thickness increases on the fiber
surfaces, the open porosity within the preform begins to close until
these passages seal over or become so small that the precursor gases
can no longer effectively flow through the body. Low pressures and
low deposition rates tend to promote infiltration and minimize
overcoating of the preform structure. Because CVI is a “gentle” gas
infiltration process, it does not damage brittle preform fibers, as do
pressurized hot melt or viscous polymer infiltration processes.
The CVI methods can be classified into six types. Lackey has
described these various types in a recent publication[42] and private
discussions.[43] These six types can be represented by the 2 × 3
matrix combinations of two heating and three gas-diffusion tech-
niques shown in Fig. 5.
Figure 5. Chemical vapor infiltration types.
The preform can be heated isothermally or in a thermal gradi-

ent fashion. With isothermal processes, the entire preform is heated
uniformly so that the precursor gases may deposit hard coating
material on all fiber surfaces simultaneously. However, the coating
process must be interrupted from time to time, to machine away the
dense exterior surface skin that forms, which inevitably limits the
passage of precursor gases through the preform for further infiltra-
tion. With thermal gradient methods, the preform structure is heated
from one side, so as to create a thermal gradient through the thick-
ness of the preform. Very sharp thermal gradients can be induced by
forced cooling of the cool side. With this thermal geometry, coating
of internal surfaces proceeds as a moving ceramic matrix front grows
from the hot side to the cold side of the preform.
In static pressure coating systems, the passage of precursor
gases into the interstices of a porous body proceeds by slow
chemical diffusion. Thus, infiltration times of a fiber preform
using chemical diffusion CVI can be extremely long, on the order
of several weeks. [26][42] Therefore, forced flow processes have been
devised to increase the rate of exposure of the internal preform
surfaces to the reactive precursor gases. With thermal gradient/

forced flow processes, infiltration times are on the order of 10 to
24 hrs.[26][42] Forced flows are accomplished by imposing a pressure
drop through the preform. The edges of the preform must be sealed in
such a manner that the gas must flow through the structure, and not
around it. The edge sealing requirement limits the geometry of
structures to simple shapes, but there are no severe limitations on
thickness. In contrast, with slow chemical diffusion processes, there
is a practical limitation on the thickness of a structure (1/8–1/4 in),
but no limitation on the complexity of the shape.
Both the isothermal and thermal gradient pulse methods can be
achieved by applying pressure vacuum pulse cycles of a precursor
gas through a porous body. However, pulsed flow diffusion methods
have been used only on a limited experimental basis, due to the very
slow deposition rates per pulse cycle. Apparently the additional cost
of the specialized pulsed flow support equipment does not justify its
usage.[42][43]
4.3 Applications with Physical Vapor Deposition (PVD)

Methods
Physical vapor deposition is probably the most versatile of all

the coating processes. It also requires the most expensive process
equipment of the methods described here. PVD generally refers to
three generic coating methods that involve evaporation, ion plating,
or sputtering. With these coating processes, virtually any solid mate-
rial, with the exception of certain polymers, may be deposited onto
any other solid surface. This versatility makes PVD especially useful
for depositing hard coatings for wear and corrosion applications.
PVD methods are line-of-sight, vacuum-coating processes typi-
cally requiring operating pressures from 10-3 to 10-8 torr. With the
high-vacuum requirements, the initial coating equipment, and
maintenance costs can be expensive. These high costs often account
for the lack of use of PVD methods for some applications. Coating
rates can vary from angstroms to tens of microns per minute, with the
highest coating rates associated with electron beam evaporative
methods. Both amorphous and crystalline microstructures can be
generated in a controlled manner, and, under appropriate conditions,
excellent adherence can be achieved. High stresses often are associ-
ated with PVD coatings due to dislocations and CTE mismatches,
but with dislocation stresses, coating materials can exhibit a higher
hardness as a thin coating than in the bulk form. Of major importance
is that the substrate can be maintained at temperatures low enough to
avoid distortion or microstructural changes during the coating pro-
cess. Like the thermal spray methods, its line-of-sight coating geom-
etry can lead to non uniform coating on complex shapes. The latter
usually can be overcome with moving fixture designs. Line-of-sight
processes conveniently lead to localized coating by masking and
shielding portions of the substrate.
Evaporation and sputtering refer to two different physical pro-
cesses of converting atoms or groups of atoms from a solid coating
target into a “transportable vapor” of neutral or charged atoms or
molecular fragments that are translated from the target in straight-
line paths to condense or nucleate on a substrate surface. Methods
that are involved with the simultaneous etching, or sputtering of the
to-be-coated substrate surface, while the coating material is being
supplied to the substrate surface by evaporation, or sputtering tech-
niques to form the coating are called ion plating methods. Excellent
descriptions of evaporation, sputtering, and ion plating have been
provided by several authors.[44]–[47]
Evaporation of metals, alloys, and ceramic coating materials
can be achieved with several energy sources to create a transportable
vapor from the melt. These include resistance and induction heating,
laser heating, and electron beam heating (electric arc, thermionic
guns, and plasma electron beams) operating at pressures of 10-5 torr
or less. Evaporation methods, including those associated with ion
plating, can exhibit coating rates from 100 Å to 20 µm/min. The

higher rates are usually associated with electron beam methods and
metal coating materials. Electron beam methods generally are pre-
ferred over resistance and induction heating methods, in order to
minimize impurity problems associated with the coating-material
melt and its material of containment. Refractory hard materials, such
as oxides, nitrides, carbides, and borides, have been deposited by
direct evaporation. However, at very high melt temperatures, even
the most stable oxides can decompose, and form molecular frag-
ments that can recombine to form unwanted deposits. To avoid such
complications, reactive-evaporation techniques have been used to
deposit hard refractory ceramics. Reactive-evaporation techniques
may use a metal target, such as titanium, to create a metal vapor by an
electric arc or beam. The metal atom vapor reacts with a gas, such as
nitrogen, introduced into the vacuum chamber, to form the refractory
compound, such as TiN. The much-publicized cathodic-arc method
used to deposit TiN is a reactive evaporation method.
Nonuniform thickness and composition can be a concern with
the evaporative coating methods. On both flat and complex-shaped
surfaces, thickness variations may occur due to shadowing and
target-to-surface deposition angle effects. Thickness non-uniformity
can be minimized by moving the substrate around with respect to the
coating source. Because various elements can possess vastly differ-
ing vapor pressures, composition variations may occur with either
multi-element targets or multiple, single-element targets, and this
difference is carried over into the composition of the deposited
coating. Sometimes, this can be overcome by adjusting the composi-
tion of the target material. Of equal concern, is the formation of
microdroplets that can form during evaporation by a process some-
times called spitting. Spitting can cause the formation of
macroparticles on the surface of the coated substrate. For hard wear-
resistance coatings, deposited macroparticles can act as stress raisers
and cause the coating to crack and break away.
With ion plating, a high negative potential is placed on the “to-

be-coated” substrate, forming a gaseous glow-discharge in close
proximity to the substrate surface. This glow-discharge region serves
to strip electrons from many of the incoming vapor species, and these
positively charged ions will be directed toward the substrate surface.
During this process, the substrate surface is continually sputter
etched by the glow discharge. Some of the surface material, includ-
ing coating material, is sputter etched as the coating is deposited,
and, therefore, the deposition rate must exceed the sputter etch rate
for the coating to form. Because of the simultaneous etching and
coating of the substrate, this process is especially effective in ensur-
ing excellent coating adherence. This is one of the reasons why most
PVD hard coatings for wear applications are deposited by ion plating
methods. Usually ion plating is associated with electron-beam evapo-
rative sources, but sputter sources often are used. Therefore, ion
plating methods can be described as evaporative-ion-plating pro-
cesses or sputtered-ion-plating processes.
Sputtering is a low-pressure, mechanical-momentum exchange
process, in which a solid coating material is converted to a transport-
able vapor by ion bombardment. In a low-pressure environment (10
to 10-2 to 10-3 torr) containing a background of a heavy inert gas,
such as argon, a negative bias is applied to the sputtering target. With
this arrangement, a plasma of positively charged argon ions can be
created in the immediate region of the sputtering target. With DC
methods, the target serves as the cathode and the coating chamber
serves as the anode. The negative potential on the target causes the
heavy argon ions to bombard its surface, thereby dislodging target
atoms to form a transportable vapor. The substrate surface to be
coated is placed in front of the coating target so that the ejected target
atoms can nucleate on its surface to form the coating. This DC
sputtering process is used for electrically conductive, metal-target
materials. With nonconductive, ceramic target materials, RF meth-
ods are used to create the transportable vapor. Thorton provides an
excellent description of the RF process.[45]
Sputtering has some distinct advantages over evaporation. With

multi-element target sources, the composition of sputtered coatings
is essentially the same as that of the target. Though different ele-
ments sputter at different rates, the sputtered elements are ejected
from the surface of the target and those elements lying below the
surface atoms are not sputtered away until exposed, so long as the
temperature of the target remains reasonably low to prevent bulk
diffusion effects within the target. An example of the transfer of the
target composition to the coated surface is the sputter ion-plated
CoCrAlY and NiCrAlY overlay coatings deposited onto gas-turbine
rotor tips[48] for oxidation/corrosion protection. These rotor tips have
been tested successfully in simulated marine turbine engines operat-
ing at 850°C. An example of reactive sputtering involving a mul-
tiple-metal target, TiAl, is the deposition of (TiAl)N coatings, first
introduced by Munz,[49] that have improved wear properties over
TiN on high-speed steel tools, but has not yet been demonstrated on
other wear surfaces such as those of bearings and gears.[50]
Though sputtering techniques exhibit lower coating rates than
evaporation or evaporative ion-plating methods, higher sputtering
rates can be achieved with magnetron sputtering.[45] Magnets placed
in close proximity to the coating target can create intense plasma
discharges at the target surface, providing relatively high coating
rates, on the order of 0.5 µm/min with pure metal targets such as
titanium. However, with sputtering targets constructed from pure
ceramic compounds, deposition rates are much lower, on the order of
hundreds of angstroms per minute. Sproul[51] has shown that by
reacting precision controlled flows of a reactive gas with the metal
atom vapors from a magnetron metal target, ceramic hard coatings
can be deposited at rates approaching 0.5 µm/min, shown in Table 4.
This method is called high rate reactive (magnetron) sputtering
(HRRS). Hard coatings deposited by this technique can exhibit much
higher hardness values (Table 4) than those of the same materials in
the bulk form. The hardness increase is thought to be due to disloca-
tions introduced into the crystal domains of the deposit. The partial
pressure of the reactant gas must be precisely controlled to prevent
the formation of the compound on the surface of the metal target

which would reduce the deposition rate by a factor of four or
more.[50]
Table 4. High Rate Reactive Sputtering Coating Rates and Hardnesses
Hardness Hardness
Coating Coating Rate Target Power (kg/mm2) (kg/mm 2)
Material (Å min-1) (kilowatts) Coating Bulk
TiN 4400 10 3680 ± 88 2000

TiC 2300 10 3720 ± 92 2900
ZrN 4000 8 4510 ± 144 1500
ZrC 3500 8 4840 ± 295 2600
HfN 5800 10 4690 ± 138 1600
HfC 4200 8 3900 ± 40 2700
(Taken from W. D. Sproul article, Ref. 51.)
In an example of utilizing HRRS, 0.5-µm coatings of TiN were

deposited on cast-iron end plates of automotive air conditioner
compressors to prevent a corrosive/adhesive welding problem.[52]
Uncoated, the end plates would weld to the surface of the sintered
powder-metal compressor rotor during the high-heat load start up,
which rendered the system inoperable. In another industrial applica-
tion, die-casting core pins made from H-13 steel became soldered or
wetted by aluminum during the die casting operation, which short-
ened the life of these pins. A 2.5-µm coating of TiN deposited by
HRRS proved effective in preventing the soldering effect.[53]
It is essential that coatings remain attached to the substrate in

order to improve the operating life of components. If a hard coating
breaks away from the substrate, more severe failure can occur than if
the component was uncoated. A prime example in a severe operating
environment was a case in which large-caliber gun tubes had been
electroplated with chromium to provide a longer operating life.[54]
However, when the coating failed, the gun bore was far more dam-
aged than if there had been no coating. Analysis of the failure
mechanism revealed that the electroplated hard chromium formed a
highly stressed and microcracked coating that allowed the high
pressure combustion gasses from the propellant charge to strip the
coating away along with some of the substrate. A long, hollow
cathode tube was designed to utilize evaporative ion plating to
deposit chromium from a flowable powder. A dual layered coating
was formed. A 7-µm, dense and well-adhered coating served as the
base layer with a 4-µm overlayer of chromium oxide. Accelerated
testing of this coating has produced promising results.
In recent work at Northwestern University by Cheng, Chang,
and Sproul, a morphological study of rolling contact fatigue (RCF)
was conducted on TiN coated and uncoated steel roller bearings.[55]
HRRS was used to deposit 0.25-µm to 5.0-µm TiN coatings onto
1.1- inch diameter 4118 carburized steel roller specimens, with the
substrate temperature never exceeding 315°C. Excellent coating
adherence was confirmed by scratch testing. Lubricated RCF was
measured by rolling the coated, and uncoated specimens against a
4.4-inch diameter 52100 hardened steel disc at a specimen speed of
4000 rpm with an applied contact stress of 2.2 GPa. The number of
rolling cycles to failure were determined as the point where the
coating began to spall and peel away. Results of this study are shown
graphically in Fig. 6. It is important to note that the thinner coating of
TiN, 0.25 µm, shows a higher spalling resistance than the thicker
coatings, and thereby is superior in preventing fatigue wear and
failure than the thicker coatings. Cheng surmises that interfacial
microcracks initiate the coating failure, and thicker coatings generate
larger microspallings, leading to rapid degradation. This is certainly
a case where thinner coatings are better than thicker coatings, but is
not to be confused with abrasive grinding wear, where thicker coat-
ings are desired. For thin coatings to be effective, a hard supporting
underlayer is desirable to prevent delamination. Plasma nitriding of
the surface, prior to PVD coating, could be an effective way to
improve further the RCF life of bearings and gears.[50]
Coating Thickness
Figure 6. Roller contact fatigue on 4118 Roller Bearing. (Reprinted with permis-
sion of Elsevier Pub. Co., Amsterdam, Ref. 55.)
4.4 Applications with Electro/Electroless Plating (EP/ElsP)

Methods
Electroplating has been commercially practiced since World

War I, while electroless plating has been in commercial use since it
was disclosed in 1946 at the National Bureau of Standards.[56] Both
plating methods have been used to deposit metal, alloy, and certain
composite hard coatings on a variety of substrates. However, the
largest commercial uses of these coatings have been for decorative
purposes, with lesser usages for wear and corrosion.
According to Schwartz,[57] only ten single-metal elements are

used today in large scale production electroplating. These include
Cr, Ni, Cu, Zn, Rh, Ag, Cd, Sn, Au, and Pb. Of these pure metals,
only chromium displays a hardness above 1000 HV. All other pure
metals fall short of chromium’s hardness. Many binary alloy compo-
sitions have been reported,[58]–[60] but few posses hardness values
above 1000 HV.
However, electroless coatings containing 3 to 15 percent of
phosphorus or boron exhibit hardness values above 1000 kg/mm2
after heat treating. The most common of these hard, electroless
coatings are nickel, Ni-P with the phosphorus additive, and Ni-B
with the boron additive. After deposition, these coatings can be heat
treated at approximately 400°C to increase their hardness by precipi-
tation hardening of the contained metal boride, or phosphide. In
addition, both electroless, and electrolytic hard composite coatings
of nickel, containing small particulate inclusions of hard ceramic
materials such as SiC, Al2O3, WC, BN, and diamond, have been
prepared in attempts to develop hard coatings for improved wear
resistance. Figure 7 shows hardness values of several plated hard-
coating materials.
Electroplating usually produces more ductile, pure metal
coatings than do electroless methods, though electrolytic metal
alloy coatings with fine particle inclusions of ceramic materials
have been reported for high-temperature oxidation protection on
metal turbine blades.[61] Electroless plating, on the other hand, is
generally associated with harder coatings and therefore, the em-
phasis in this section is on electroless plating. Readers interested
in electroplating are directed to other sources for detailed descrip-
tions of the fundamental process and examples.[57][62]–[65]
Hard coatings can be deposited by electroless plating, but not
all metals can be deposited by electroless methods.[65] The far left
column of Table 5 lists the common metals plated by this method.
Figure 7. Hardness of plated hard coatings.

Table 5. Electroless Plating
Substrate Surfaces
Plated Not
Plating Reducing Directly Requiring Plated
Metals Agents (Activators) Activator (Poisons)
Cu hypophosphites Pd other metals Pb

Co borohydrides Ni ceramics Cd
Au hydrazine Co plastics Zn
Ni formaldehyde Fe graphite Bi
Pd amine boranes Al
Ag
(Taken from data presented in Ref. 64)
Electroless plating is a non electrical, autocatalytic, reduction

process that is used to deposit certain metals from aqueous solution
onto a solid surface. The plating solution contains a phosphorus
containing, boron containing or other chemical reducing agent, such
as those listed in Table 5. The surface to be coated must be catalytic
to the reduction process for deposition to occur. Surfaces that can be
plated directly without “activation” also are listed in Table 5. These
same materials can serve as “activators” for surfaces not catalytic to
the electroless process, listed under “Requiring Activator.” Deposi-
tion of a thin layer of one of these catalytic elements on the
component’s surface will initiate the electroless process. Once initi-
ated, the electroless deposited coating will continue to catalyze the
electroless process so that deposition becomes self sustaining or
“autocatalytic.” Certain metals act as “catalytic poisons” to the
electroless process and cannot be plated directly. A list of these
poison metals is given in the far-right column of Table 5. To

electrolessly plate these surfaces, an intermediate catalytic layer
must be deposited by other methods, perhaps electroplating.
There are several examples of electroless Ni-B hard coatings
used for wear and corrosion resistance. Secoa Technologies Inc.[66]
has reported the use of Ni-B coatings in wear and corrosion environ-
ments to increase the operating life of several critical components
such as the inlet-guide vane and the synchronization ring on Pratt &
Whitney’s F-100 jet engine, an automotive manual-clutch assembly
and a spherical bearing used in the operation of the manipulator arm
on the space shuttle.
Successful examples of electroless Ni-P hard coatings used for
both wear and corrosion resistance, include deposits on bronze
castings in deionized water and on valves coming into contact with
sour (H2S) crude oil.[67]
In another example, a 50% construction cost savings was
realized in a corrosion application[68] when a 10-µm coating of Ni-
8%P was applied to the inside surface of a 48-inch diameter, 22-foot
long, high-pressure (5,000 PSI), thick-walled (4 inches), steam-clad
reaction vessel for the production of low-density polyethylene. The
production of low-density polyethylene is very sensitive to contami-
nation by iron. Costly stainless steel construction was avoided by
depositing a Ni-P coating, by pumping the electroless nickel plating
solution through a low-cost, carbon steel tank. This vessel has
continued to operate for more than 3 years without contamination of
the iron-sensitive polymer.
Composite or “inclusion” hard coatings can be formed with
either electroless or electroplating techniques. Finely divided par-
ticulates, on the order of 0.5-to 1-µm in size, can be dispersed in the
plating solution by chemical dispersing agents or by mechanical
agitation. As the coating process proceeds, the particulates are in-
cluded into the metal-matrix coating. Primarily, electroless nickel
(Ni-P) coatings, with additions of SiC, Al2O3, WC, BN, and dia-
mond, have been prepared for improved wear resistance.[69]
In two recent reports by Changgeng, et al.,[70] electroless nickel

coatings with variations in phosphorus content were prepared with
and without inclusions of SiC. Hardness values were measured
before and after heat treatment at 400°C. Each coating, 60-to 80-µm
thick, was deposited onto separate steel blocks. The wear volume of
each coated block was determined by rotating a WTiV steel ring
under load against the lubricated steel block. Table 6 indicates that
the SiC inclusion coating is 2 to 5 times better in wear performance
than samples electroplated with hard chrome or electroless Ni-P.
Coatings of Ni-P and Ni-P-SiC, 12 µm thick, have been deposited on
molding dies used in the production of plastic components. The
operating life of the dies coated with electroless Ni-P was twice that
of the uncoated dies, while dies coated with the Ni-P-SiC inclusion
coating were showed an operating life improvement of 5 times.
Table 6. Vickers Hardness of Electroless Nickel Composite Coatings
Vickers Vickers Wear Volume

Hardness Hardness after H.T.
Coating As Plated 400°C Heat Treat on Hard Coating
ElsP Ni-4%P 666 1146 ≈ 4 × 10-3 mm3

ElsP Ni-6%P 509 1146 ≈ 4 × 10-3 mm3
Elsp Ni-8.5%P 497 1137 ≈ 4 × 10-3 mm3
Elsp Ni-10%P 487 1137 ≈ 4 × 10-3 mm3
Elsp Ni-6%–2.5%SiC 650 1146 0.8 × 10-3 mm3
EP Cr 952 ≈ 2 × 10-3 mm3
(Taken from data in Ref. 70)
5.0 DIAMOND COATINGS
Any discussion of hard coatings would not be complete with-

out mention of diamond coatings. Today, diamond-film technology
is very much in its adolescent stage of development. The deposition

mechanism is not well established and much work remains to be
done to understand how to deposit acceptable films on component
substrates in a reproducible and economical manner. There are few
application examples of hard-diamond coatings in the market place
today, and these are focused primarily on electronic and optical
applications. Wear examples of diamond coatings on cutting tools
have been discussed in Ch. 7. Nevertheless, because diamond coat-
ing technology offers such high expectations for wear prevention for
non-cutting tool applications, a cursory description of the processes
is given here without specific industrial examples.
Diamond, a metastable form of carbon, exhibits extremes in
many of its properties, not the least of which are its hardness
(≈ 10,000 kg/mm 2), and elastic modulus (≈10.5 × 1011 N/m2), the
highest for all known natural materials. Furthermore, the coefficient
of friction of bulk diamond and deposited films against steel surfaces
are in the 0.05 to 0.15 and 0.035 to 0.2 ranges, respectively.[71][72]
The very low coefficient of friction of diamond puts diamond in the
frictional sliding range of teflon. These properties make diamond
especially attractive for wear applications. Other critical properties,
such as thermal conductivity, thermal expansion, electrical resistiv-
ity, and optical transparency are particularly attractive for electronic
and optical applications. However, all of these properties are very
sensitive to impurity and structure deviations.
Diamond particles have been commercially manufactured since
the mid fifties[73] by high temperature (≈ 2000°C), and high pressure
(≈ 55 × 10 3 atm) processes. With this and other similar processes, 30
to 40 tons of diamond grit, containing impurities of graphite, metal
carbides, and other elements, are manufactured annually for world
wide industrial abrasive applications.[74] Impurity-free synthesis has
been accomplished, but bulk scale up is thought to be extremely
expensive and difficult. Therefore, impurity-free industrial diamonds
have experienced limited usage.
Vapor deposition offers a practical route to produce high-
purity diamond films on surfaces of components at low pressures and
reasonable temperatures. Diamond coating methods can be classified

as shown in Fig. 8. The more frequently used class is based on
plasma assisted CVD methods (PACVD), and the other lesser used
class is based on high energy, ion beam deposition methods.
Figure 8. Diamond coating methods. (Courtesy of Basic Industry Research

Laboratory.)
PACVD methods have been used more often because they

require less-costly equipment, and would appear to be more easily
scaled to production than ion-beam methods. Coating rates of tens of
microns per hour can be achieved with PACVD methods.[72][74][75]
The Japanese have reported deposition rates of 930-µm/hr with a
direct current, thermal plasma jet process.[76]
PACVD methods use hot-wire filaments, high-temperature
combustion, RF, DC, or microwave excitation to generate plasmas to
promote the diamond deposition processes. With these processes,
typical gas combinations of 0.2% to 2% hydrocarbon precursor with
mixtures of hydrogen, inert gasses, and oxygen in some cases, flow
through the reaction chamber at pressures in the range of 1 torr to 1
atmosphere.[75] With hot-wire systems, the filament, usually W, Ta,
or Re is resistance heated in the range of 2000°C to 2500°C to create
a plasma in the immediate vicinity of the substrate. The plasma

provides energy to the hydrocarbon radicals to form the sp3 elec-
tronic configuration necessary to produce the tetrahedral carbon,
cubic diamond structure. For high quality films, substrate tempera-
tures are kept between 850°C and 1050°C by radiation or an alternate
heating device.[74]
High temperature, flame combustion methods have been re-
ported recently that produce diamond coatings on silicon substrates
in air with a brazing torch that uses oxygen and acetylene precursors
to create a high-temperature (3000°C) combustion flame at pressures
from 25 torr to 50 torr.[77] Still in the early stages of development, the
stationary process has produced only small area (limited by the
diameter of the inner-flame tip), and non uniform coatings. Presum-
ably, large-area coating could be accomplished with torch or sub-
strate translation.
Ion beam methods, also in the early stages of development,
may be capable of producing diamond coatings by one of two basic
methods. In the first, called direct ion beam deposition, an ion
accelerator source capable of producing high-energy beams of car-
bon atoms in the 10 eV to 150 eV range, are focused on a substrate
surface. The thermal agitation and resultant shock wave of these
picosecond pulses are capable of producing localized depositions of
single-crystal diamond. It is estimated that during these short bursts,
localized temperatures may reach ≈ 75,000°C with localized pres-
sures in excess of 100,000 atms.[72] With the second method, called
ion-beam enhanced deposition (IBED), a primary ion beam deposits
a coating of diamond on the substrate, while a secondary ion beam
bombards the deposited coating. Because the energy of these two
beams can be varied and controlled independently, the type of
carbon bonding may be controlled to produce high quality dia-
mond structures. Furthermore, substrate temperatures may be
held below 100°C, although coating rates are reported to be very
low, < 0.1-µm/hr.[72][75] Because of the great expense of this equip-
ment and the very low coating rates, this technique has not been used
extensively.
In all of these methods, carbon is deposited in several structural

forms simultaneously, each of which depends upon the bonding
configuration of the carbon atom. The bonding configuration is
associated with the energy supplied in the coating process. The
covalent sp3 configuration is required for the cubic diamond struc-
ture, sp2 is needed for the graphitic structure, and sp 1 is generally
associated with amorphous or “discontinuous” carbon. Mixtures of
carbon atoms with sp3 and sp2 configurations (sp3/sp2) form struc-
tures that have been called “diamond like” carbon while sp2/sp1
mixtures have been associated with amorphous carbon, and “graph-
ite like” carbon.[72][75] Early attempts to deposit diamond using
vacuum evaporation, sputtering, and non-plasma CVD methods,
often produced amorphous, graphitic, and sometimes diamond-like
carbon coatings. It is the higher energy PACVD and ion-beam
processes that are necessary to produce diamond coatings. Even with
these higher energies, all bonding configurations are formed, which
can result in mixtures of carbon, graphite, and diamond. Atomic
hydrogen is utilized in all the diamond coating methods because: it
preferentially etches away the non sp3 carbons and therefore, pro-
motes growth of a “purer” diamond structure[72][75] and it is thought
to aid in the nucleation process. Oxygen is sometimes added to the
gas mixture, because (1) it can increase deposition rate, (2) it also
preferentially etches away the non sp3 carbons, and (3) it promotes
the formation of atomic hydrogen in the plasma.[78] Trapped hy-
drogen in the films can produce high tensile stresses, causing
buckling and delamination.[75]
The nucleation of diamond is influenced by the substrate sur-
face for reasons not fully understood. Scratched, or grooved surfaces
have been observed to promote diamond nucleation much more
readily than smooth surfaces.[79] Silicon has proven to be an excel-
lent substrate material for promoting diamond nucleation.[81] There-
fore, materials such as Si, SiC, Si3N4, and silicon aluminum
oxynitride have been used as diamond coating substrates.[74][75]
Other successfully coated substrates have included graphite,
WC/Co, cubic BN, and certain metals such as Ti, Mo, Ta, W, Cu,
and their carbides.[80][81] Also Cu, Au, Pt, and Rh substrates have
been diamond coated.[82] Diamond nucleates readily, and adheres
well on carbide forming substrates like Mo and W.[81] However,
many of the transition metals have not been coated successfully for
reasons yet unknown.[81] These metals have included Fe, Ni, Co, Cr,
and Mn.[82] However, with Ni, diamond has been observed to deposit
in the presence of gaseous Si.[80]
Finally, hard wear resistant coatings can be deposited by both
the PACVD methods, applicable for high-temperature substrates and
the ion-beam methods, which could be used for the specialty lower-
temperature substrates. No doubt that with continued development,
successful applications for wear resistant diamond coatings on such
varied components as those included in Table 2 will become frequent
in the near future.
REFERENCES
1. Rabinowicz, E., Dept. Mechanical Engineering, Massachusetts

Institute of Technology, private communication, (March, 1990);
Peters, S., The Corroding of America, Chicago Tribune, Sec. 5,
(Aug. 30, 1988); U.S. Dept of Commerce, Materials Performance,
57, (February, 1983); The MIT Report, (June, 1987); Proceedings
of a Workshop on Wear Control to Achieve Product Reliability, (M.
J. Devine, ed.), Sponsored by OTA, U.S. Congress, Naval Air Dev.
Center, Warminster, PA (1976)
2. Council of Economic Advisors, Boskin, M. J., Chmn., Economic
Indicators, U.S. Government Printing Office, Washington, D.C.
(1989)
3. Bunshah, R. F., “Deposition Technologies: An Overview,”
Deposition Technologies For Films And Coatings, (R. F. Bunshah,
ed.), p. 335, Noyes Publications, Park Ridge, NJ (1982)
Wear and Corrosion
4. Sibley, L. B., and Allen, C. M., “Friction and Wear Behavior of

Refractory Materials at High Sliding Velocities and Temperatures,”
Wear, 5:312–320 (1962)
5. Dow, T. A., and Burton, R. A., “The Role of Wear in the Initiation of
Thermoelastic Instabilities of Rubbing Contact,” J. Lub. Tech.,
Trans ASME, Series F 95 (1):71–75 (1973); Dow, T. A., and
Stockwell, R. D., “Experimental Verification of Thermoelastic
Instabilities in Sliding Contact,” J. Lub. Tech., Trans ASME, Series
F99 (1):359–364 (1977)
6. Dufrane, K. F., and Glaeser, W. A., “Performance of Ceramics in
Ring/Cylinder Applications,” SAE Technical Paper Series, SAE
Paper 870416 (1987)
7. Department of Trade and Industry, Wear Resistant Surfaces In
Engineering, Her Majesty’s Stationery Office, London (1986)
8. Fontana, M. G., Corrosion Engineering, 3rd Ed., McGraw-Hill
Book Co., New York (1986)
Materials Selection
9. Holleck, H., “Material Selection For Hard Coatings,” J. Vac. Sci.

Technol. A., 4(6):2661–2669 (1986)
10. Ramalingam, S., “New Coating Technologies for Tribological
Applications,” in: Wear Control Handbook, (M. B. Peterson, and
W. O. Winer, eds.), American Society of Mechanical Engineers,
New York (1980)
Thermal Spray Coating
11. Kay, A., “Successful Hypersonic Spray Applications in the Steel

Industry,” Thermal Spray Technology New Ideas and Processes,
Proceedings of the National Thermal Spray Conference, (D. L.
Houck, ed.), pp. 443–448, ASM International, Metals Park, OH
(1989)
12. Tucker, R. C., Jr., “Plasma And Detonation Gun Deposition

Techniques And Coating Properties,” in: Deposition Technologies
For Films And Coatings, (R. F. Bunshah, ed.), pp. 454–488, Noyes
13. Poorman, R. M., Sargent, H. B., and Lamprey, Method and Apparatus
Utilizing Detonation Waves for Spraying and Other Purposes, U.S.
Patent No. 2,714,553 (August 2, 1955)
14. U.S. Department of Commerce, NTIS Foreign Technology, 89, No.
26 (1989)
15. Morando, J. A., and Schultz, W., Theory and Application of Liquid
Stabilized Plasma Technology on Thermo Deposited Near Net Shaped
Bulk Bodies, Plasma Fusion Inc., Taylor, MI (1986)
16. Bernecki, T. F., Basic Industry Research Laboratory, Evanston, IL,
private communication
17. Hillery, R. V., “Coatings for Performance Retention,” J. Vac. Sci.
Technol., 4(6):2624–2628 (1986)
18. Levy, A., and Macadam, S., “The Behavior of Ceramic Thermal
Barrier Coatings on Diesel Engine Combustion Components,”
Chemical Vapor Deposition
19. Yee, K. K., “Protective Coatings for Metals by Chemical Vapor

Deposition,” Review No. 226, International Metals Reviews,
(1):19–42 (1978)
20. Bryant, W. A., “Review: The Fundamentals of Chemical Vapor
Deposition,” J. Mater. Sci., 12:1285–1306 (1977)
21. Powell, C. F., Oxley, J. H., and Blocher, J. M., Jr., (eds.), Vapor
Deposition, John Wiley and Sons, New York (1966)
22. Blocher, J. M., Jr., “Chemical Vapor Deposition,” in: Deposition
Technologies For Films And Coatings, (R. F. Bunshah, ed.), pp.
335–362, Noyes Publications, Park Ridge, NJ (1982)
23. Blocher, J. M., Jr., Browning, M. F., and Barrett, D. M., “Chemical
Vapor Deposition of Ceramic Materials,” in: Materials Science
Research, (R. F. Davis, H. Palmour, III, and Porter, R. L., eds.),
17:299–316, Plenum Press, New York, (1984)
24. Bonifield, B. D., “Plasma Assisted Chemical Vapor Deposition,”

in: Deposition Technologies For Films And Coatings, (R. F Bunshah,
ed.), pp. 365–383, Noyes Publications, Park Ridge, NJ (1982)
25. Pierson, H. O., (ed.), Chemically Vapor Deposited Coatings,
American Ceramic Society, Columbus, OH (1981)
26. Stinton, D. P., Besmann, T. M., and Lowden, R. A., “Advanced
Ceramics by Chemical Vapor Deposition Techniques,” Ceramic
Bulletin, 67(2):350–354 (1988)
27. Lackey, W. J., Smith, A. W., Dillard, D. M., and Twait, D. J.,
“Codeposition of Dispersed Phase Ceramic Composites,”
Proceedings of the Tenth International Conference on Chemical
Vapor Deposition, pp. 1008–1027, Electrochemical Society,
Pennington, NJ (1987); Stinton, D. P., and Lackey, W. J.,
“Simultaneous Chemical Vapor Deposition of SiC Dispersed Phase
Composites,” Ceram. Eng. Sci. Proc., 6(7/8):707–713 (1985)
28. Allen, S. D., “Laser Chemical Deposition,” Proceedings of the
Eighth International Conference on Chemical Vapor Deposition,
pp. 267–271; Allen, S. D., “Laser Chemical Deposition,” NATO
Advanced Study Instruction Series, B, 4:455–468 (1983);
Electrochemical Society, Pennington, NJ (1981); Solanki, R., Ritchie,
W. H., and Collins, G. J., “Photodeposition of Aluminum Oxide and
Aluminum Thin Films,” App. Phys. Lett., 43(5):454–456 (1983)
29. Garg, D., Dyer, P. N., Dimos, D. B., Sunder, S., Hintermann,
H. E., and Maillat, M., “Low Temperature CVD Tungsten
Carbide Coatings for Wear/Erosion Resistance,” Cer. Eng.
Sci. Proc., 9(9/10):1215–1222 (1988)
30. Cox, J. W., Basic Industry Research Lab, Evanston, IL, unpublished
work
31. Podob, M., “CVD Hard Coatings Lengthen Tool Life,” Metal
Progress, 121(6):50–53 (1982)
32. Hintermann, H. E., Perry, A. J., and Horvath, E., “Chemical Vapor
Deposition Applied In Tribology,” Wear, 47:407–415 (1978)
33. Hintermann, D. E., Boving, H., and Hánni, W., “Wear Resistant
Coatings For Bearing Applications,” Wear, 48:225–236 (1978)
34. Horvath, E., and Perry, A. J., “The Wear Resistant Coating Of Steel
By Chemical Vapor Deposition,” Wear, 48:17–224, (1978)
35. Caputo, A. J., Lackey, W. J., Wright, I. G., and Angelini, P.,
“Chemical Vapor Deposition of Erosion Resistant TiB2,” J.
Electrochem. Soc.: Solid State Science And Technology, 2274–2280
(Sept. 1985)
36. Yawsend, C. L., and Wakefield, G. F., “Scale up Process for Erosion
Resistant Titanium Carbonitride Coating,” Proceedings of the Fourth
International Conference on Chemical Vapor Deposition, pp.
277–287, Electrochemical Society, Pennington, NJ (1973)
37. Agaisse, R., Haessler, M., Hertz, D., Hánni, W., Hintermann, H. E.,
Heinzel, V., and Schumacher, G., “Internal Protective Coating of
Niobium Tubes,” Proceedings of the Third European Conference
on Chemical Vapor Deposition, (H. E. Hintermann, ed.), pp.
249–258, Zwahlen, Saint Blaise, Switzerland (1980)
38. Kotchick, D. M., and Dobos, C. J., “CVD SiC Coatings For Ceramic
Heat Exchanger Tubes,” Conf. Proc. of Met. Soc. of AIME, (S. C.
Singhal, ed.), pp. 281–292, 112th Annual Meeting of AIME, Atlanta,
GA (1983)
39. Archer, N. J., “The Preparation and Properties of Pyrolytic Boron
Nitride,” High Temperature Chemistry in Inorganic and Ceramic
Materials, (F. P. Glasser, and P. E Potter, eds.), pp. 167–180, The
Chemical Society, London (1977)
40. Oxley, J. H., Reactor Material, 6, 2, 1, (1963); Oxley, J. H., Browning,
M. F., Veigel, N. D., and Blocher, J. M., Ind. Eng. Chem. Prod. Res.
Develop., 1:102, (1963); Oxley, J. H., Hannah, J. F., Blocher, J. M.,
and Campbell, I. E., Ind. Eng. Chem., 51:1391 (1959)
41. Stinton, D. P., Ceramic Composites by Chemical Vapor Infiltration,
DOER&D Report CONF-871027-6, DE87 009128 (1987)
42. Lackey, W. J., “Review, Status, and Future of the Chemical
Infiltration Process for Fabrication of Fiber Reinforced Ceramic
Composites,” Ceram. Eng. Sci. Proc., 10(7/8):577–584 (1989)
43. Lackey, W. J., Georgia Tech Research Institute, Alanta, GA., private
communication
Physical Vapor Deposition
44. Bunshah, R. F., “Evaporation,” in: Deposition Technologies For

Films And Coatings, (R. F. Bunshah, ed.), pp. 83–167, Noyes
45. Thornton, J. A., “Coating Deposition by Sputtering,” in:
ed.), pp. 170–237, Noyes Publications, Park Ridge, NJ (1982);
Thornton, J. A., “High Rate Thick Film Growth,” Ann. Rev. Mater.
Sci., 7:239–260, (1977)
46. Mattox, D. M., “Ion Plating Technology,” in: Deposition
Technologies For Films And Coatings, (R. F. Bunshah, ed.), pp.
244–268, Noyes Publications, Park Ridge, NJ (1982)
47. Sundgren, J. E., and Hentzell, H. T. G., “A Review of the Present
State of Art In Hard Coatings Grown from the Vapor Phase,” J. Vac.
Sci. Tech. A., 4(5) (1986)
48. Rickerby, D. S., and Wood, M. I., “Evaluation of Sputter Ion Plated
CoCrAlY, and NiCrAlY Coatings for Gas Turbines,” J. Vac. Sci.
Technol. A., 4(6):2557–2564 (1986)
49. Munz, W. D., “Titanium Aluminum Nitride Films: A New Alternative
to TiN Coatings,” J. Vac. Sci. Technol. A., 4(6):2717–2725 (1986)
50. Sproul, W. D., Basic Industry Research Laboratory, Evanston, IL,
51. Sproul, W. D., “Reactively Sputtered Nitrides and Carbides of
Titanium, Zirconium, and Hafnium,” J. Vac. Sci. Technol. A.,
4(6):2874–2878 (1986)
52. Sproul, W. D., “Wear of Sputtered Deposited Refractory Metal
Nitride Coatings, Physics and Chemistry of Protective Coatings,”
(W. D. Sproul, J. E. Greene, and J. A. Thorton, eds.), in: The
American Vacuum Society, Series 2, (G. Lucovsky, Series ed.),
American Institute of Physics Conference Proceedings, No. 149, pp.
157–172, New York (1986)
53. Sproul, W. D., with permission from the Paragon Die Cast Company,
Chicago, IL.
54. White, J. R., “Summary Abstract: A Physical Vapor Deposition
Technique for Plating Gun Tubes,” J. Vac. Sci. Technol. A.,
4(6):2855–2856 (1986)
55. Cheng, H. S., Chang, T. P., and Sproul, W. D., “A Morphological

Study of Contact Fatigue of TiN Coated Rollers,” Proceedings of
the 16th Leeds-Lyon Conference, Lyon, France (1989)
Electroplating/Electroless Plating
56. Brenner, A., and Riddell, G., Proc. Am. Electroplat. Soc., 33:23
(1946)
57. Schwartz, M., “Deposition from Aqueous Solutions: An Overview,”
ed.), pp. 385–488, Noyes Publications, Park Ridge, NJ (1982)
58. Brenner, A., Electrodeposition of Alloys, Principles and Practice,
Academic Press, New York (1963)
59. Brenner, A. Plating, 52(12):1249 (1965)
60. Krohn, A., Bohn, C. W., Electrodep. & Surf. Treatment, 1(3):199
(1973)
61. Honey, F. J., Kedward, E. C., and Wride, V., “The Development of
Electrodeposits for High Temperature Oxidation/Corrosion
Resistance,” J. Vac. Sci. Technology A., 4(6):2593–2597 (Nov/Dec,
1986)
62. Safranek, W. H., The Properties of Electrodeposited Metals and
Alloys, Elsevier Pub. Co., New York (1974)
63. Lowenheim, F. A., (ed), Modern Electroplating, 3rd Ed., John
Wiley and Sons, New York, (1973)
64. Dubpernell, G., Electrodeposition of Chromium, Pergamon Press,
NY (1977)
65. Pearlstein, F., and Bellis, H. E., Electroless Plating of Metals,
American Electroplaters Society (1972)
66. Product Bulletin, Secoa Technologies Inc., West Palm Beach, FL
(1986)
67. “Using Electroless Nickel Plating,” American Machinist (Sep, 1977)
68. Nye Kote Electroless Nickel Plating in 29 Ton Process Vessel
Avoids Costly SS Construction, Electro Coatings Inc., Houston, TX
(1978)
69. Broszeit, E., “Mechanical, Thermal and Tribological Properties of

Electro And Chemodeposited Composite Coatings,” Thin Solid
Films, 95:133–142 (1982)
70. Changgeng, X., Zonggeng, D., and Lijun, Z., “The Properties of
Electrodeposited Ni-P-SiC Composite Coatings,” Plating and
Surface Finishing, 75(10):54–57 (1988); Changgeng, X., Xinmin,
H., Zonggeng, D., and Yanwen, W., “The Properties of Electroless
Ni-P, and Ni-P-SiC,” Plating and Surface Finishing, 76(6):90–93
(1989)
Diamond Coating
71. Wong, M. S., Meilunas, R., Ong, T. P., and Chang, R. H. P.,
“Tribological Properties of Diamond Films Grown by Plasma
Enhanced Chemical Vapor Deposition,” Appl. Phys. Lett.,
54(20):2006–2008 (1989)
72. Deutchman, A. H., and Partyka, R. J., “Diamond Film Deposition,”
Advanced Materials & Processes, 135:29–33 (June 1989)
73. Bundy, F. P., Hall, H. M., Strong, H. M., and Wentorf, R. H.,
Nature, 176:51 (1955)
74. Bachmann, P. K., and Messier, R., “Emerging Technology of
Diamond Thin Films,” Chemical and Engineering News, pp. 24–39,
(May 15, 1989)
75. Angus, J. C., and Hayman, C. C., “Low Pressure, Metastable Growth
of Diamond and Diamondlike Phases,” Science, 241:913–921 (1988)
76. Ohtake, N., and Yoshikawa, M., “Diamond Film Preparation by Arc
Discharge Plasma Jet Chemical Vapor Deposition in the Methane
Atmosphere,” J. Electrochem. Soc., 137(2):717–722 (1990)
77. Yarbrough, W. A., Stewart, M. A., and Cooper, J. A., “Combustion
Synthesis Of Diamond,” Surface and Coatings Technology, 39/
40:241–252 (1989); Carrington, W. A., Hanssen, Snail, K. A.,
Oakes, D. B., and Butler, J. E., “Diamond Growth in O2 + C2H4 and
O2 + C2H2 Flames,” Metallurgical Transactions A, 20A:1282–1284
(1989)
78. Meilunas, R., Wong, M. S., Sheng, K. C., Ong, T. P., and Chang, R.
P. H., “The Initial Stages of Plasma Synthesis of Diamond Films,”
presented in MRS Fall Meeting 1988, paper to be published in MRS
proceedings, Vol. 129, Laser and Particle Beam Chemical Processes
on Surfaces (1989)
79. Spear, K. E., “Growth of Crystalline Diamond from Low Pressure
Gases,” Earth and Mineral Sciences, 56(4):53–59 (1987)
80. Yarbrough, W. A., and Messier, R., “Current Issues and Problems in
the Chemical Vapor Deposition of Diamond,” Science,
247:688–696 (1990)
81. Badzian, A. R., Badzian, T., Roy, R., Messier, R., and Spear, K. E.,
“Crystallization of Diamond Crystals and Films By Microwave
Assisted CVD,” Materials Research Bulletin, 23:531–548 (1988)
82. Wong, M. S., Basic Industry Research Laboratory, Evanston, IL,
Cubic Boron Nitride and

Diamond-Related Thin
Films
Chris H. Stoessel, Rointan F. Bunshah
1.0 INTRODUCTION
Diamond and cubic boron nitride (c-BN) are materials of

exceptional and very similar properties that make them extremely
valuable not only for the aesthetics of diamond, but also for their
technological potential. The outstanding and unrivalled qualities of
these two materials are mechanical, thermal, optical, and electronic
properties, making them prime candidates for the respective fields of
applications; and often they excel by offering unique combinations
of their various properties. Whereas natural diamond has been the
center of centuries of scientific attention and is therefore well docu-
mented, c-BN is a material synthesized rather recently that has
rapidly spurred technological and scientific interest. Presently, both
materials have found commercial applications.
466
Cubic Boron Nitride and Diamond-Related Thin Films 467
2.0 CUBIC BORON NITRIDE FILMS
2.1 Structure and Properties of B-N Compounds
Similar to carbon, there are three phases in the B-N system, but
unlike the carbon system, all three are synthetic materials. Graphite-
like hexagonal h-BN (also called g-BN or alpha-BN) is a soft
material of white color. Several derived varieties of this structure
exist: amorphous a-BN, turbostatic (random-layer lattice) t-BN, and
rhombohedral r-BN. The graphitic structures are metastable.[1][2]
The equivalent to diamond in the B-N system is the cubic c-BN
(also called z-BN or beta-BN) with several outstanding qualities,
such as hardness (>4000 HV) surpassed only by diamond, low
chemical reactivity, and high thermal stability (up to 1400 K), high
thermal conductivity, high electrical resistance, and low density. A
comprehensive list of physical properties of c-BN can be found in
several material handbooks, and a short list of data is compiled in
Table 1. This combination of properties makes c-BN more suit-
able as an engineering material than diamond for many applica-
tions, especially for wear.[1] For example, diamond cannot be used
to machine steels (particularly hardened steels), because diamond
(carbon) dissolves into the steel at the high tool tip temperatures. c-
BN, in comparison, has a very low affinity to steel, therefore making
it a much better candidate for machining this material.
The isotypical phase to hexagonal diamond is the wurtzite w-
BN (of unknown stability/metastability).[1][2] This structure is the
least important for technological applications of B-N compounds,
and its properties are only sparsely documented.
For c-BN, there is a strong correlation between color and
composition: black indicates excess boron; amber indicates B
deficiency.[3]
Table 1. Properties of B-N Materials
Property c-BN h-BN a-BN
crystal structure cubic, zinc blende hexagonal amorphous
Lattice constant a = 3.61 Å a = 2.50 Å

c = 6.60 Å
Density 3.48 gcm -3 2.2 gcm-3 1.74 gcm-3
Hardness 6000–7500 HV 2 Mohs 200–600 HV
Resistivity - undoped 10 10 ohm·cm (at RT) 1.7 x 1013 ohm·cm 1015 ohm·cm
- p-type (Be) 10 2–104 ohm·cm
- n-type (B) 10 3–107 ohm·cm
Activation energy (p) 0.19–0.23eV (n) 0.05–0.41eV
Thermal conductiv. 2–9 Wcm-1K-1 (c) 0.68 Wcm-1K -1

(a) 0.36 Wcm-1K-1
Lin.therm. expans. 4.8 x 10-6K-1 (700 K) 3.8 x 10-6 K-1
Many analytical methods have been used for the investigation

of BN films. Several characteristic features are that BN compounds
are poor x-ray scatterers, therefore making x-ray analysis sometimes
difficult; electron diffraction is better for crystallography investiga-
tions. Infrared spectrophotometry has been used for investigation of
short range order of BN films.[4]
2.2 Synthesis of c-BN Materials
Synthesis of Bulk Materials. Various attempts have been

made to fabricate BN materials that would be feasible for industrial
applications in various fields (for example, wear protection, machin-
ing, semiconductors, and electronics, as well as in optics). The first
results were achieved in the synthesis of bulk materials.
The transformation of h-BN to c-BN can be achieved under the

following optimum conditions (pressure of 6 GPa, temperatures of
1470–1720 K) and results in polycrystalline, fine-grained particles
after a processing time of 15 mins. It is the best if h-BN is of small
particle size and poorly crystallized, thus lowering the high activa-
tion barrier for the c-BN transformation, whereas well-crystallized
h-BN starting material results in w-BN.
Synthesis from amorphous a-BN has been accomplished at a
pressure of 7 GPa and temperatures above 1070 K. It is assumed that
the reaction follows two possible conversion mechanisms: direct
transformation of a-BN into c-BN or via an intermediate transforma-
tion, a-BN→ h-BN→ c-BN.
The two methods described above have been performed on a
laboratory scale, and produced c-BN of highest purity and quality.
However, they are unsuitable for the production of c-BN crystals on
an industrial scale, especially when purity aspects are not as stringent
(for example, for wear protection or machining applications), or for
thin films of BN onto substrates such as cemented carbide tool bits.
The most economical production methods in that case are “cata-
lyzed” transformations of h-BN to c-BN. In these reactions, addi-
tives (several are available) act as flux precursors, forming an eutec-
tic with h-BN. The resulting c-BN powder is of small size (submi-
cron to 0.4 mm diameter) with crystals of tetrahedral or octahedral
shapes. The crystallites usually contain precursor residues as inclu-
sion contaminants. Another less common method for special appli-
cations of bulk c-BN (for example, in the semiconductor field) is of
interest for research purposes. The static high pressure method (Fig.
1) uses a temperature differential (for example, 1890K to 1540K) in
a powder-filled reactor to grow crystals from c-BN seeds. The
process is very slow, (50 mg after 60 h under 5.5 GPa and use of
flux precursors) but allows the growth of larger crystals than the
methods mentioned above.
Finally, the dynamic high pressure process (Fig. 2) uses mul-

tiple shock waves to provide the transformation energy for the
production of c-BN from amorphous BN. In industrial applications,
this method is mainly used for compacting c-BN powders into larger
volumes (as an alternate to sintering).
Figure 1. Static high pressure method.
Figure 2. Dynamic high pressure method.
Synthesis of Thin Films. Recently, there has been consider-

able interest in the production of thin c-BN (as well as h-BN or other
configurations) films which are expected to open new fields of
applications for this unique material (see below). Generally speaking,
thin film deposition methods for BN films can be described as the

degradation of BN precursors by hot filament, plasma, or laser.
Various processes have been developed.
The ion beam assisted synthesis of i-BN (unidentified B-N
structure with low yield of c-BN, presumably N-rich) from borazine
(B3N3H6), following a suggestion of Weissmantel,[5] has been per-
formed in two variations; ion assistance was provided by a borazine-
fed Kaufman-type ion beam source, and the so-called IBED tech-
nique (ion beam enhanced deposition, Fig. 3), which combines
electron beam evaporation of boron with N-ion bombardment from a
Kaufman-type ion source. IBED is believed to be beneficial for
adhesion, (high ion energy) nucleation and grain growth, internal
stress, and variable composition. Significant hydrogen and carbon
contamination (H from water vapor, hydrocarbons, and precursor, C
from inadvertent evaporation of the graphite crucible) has been
detected. The resulting thin films (450 Å) were amorphous; thicker
films have also been produced and contain crystallites of 200–500 Å
diameter in an amorphous matrix.[6]
Figure 3. Ion beam enhanced deposition (IBED).

Deposition with the activated reactive evaporation technique

(ARE, Fig. 4) with Tsub = 450°C resulted in c-BN films, and used the
evaporation of boric acid (H3BO3) in NH3 plasma. This process has
also been described as the activated dissociation reduction reaction
process, (ADRRP) and uses a filament-anode arrangement to in-
crease plasma ionization, as well as RF substrate biasing. The c-BN
film could be identified clearly in IR and UV transmission, band gap,
optical, and microhardness studies. Earlier attempts with ARE meth-
ods employing less plasma activation (until 1987) resulted in non-c-
BN films. A modified version of the ARE process employing a gas-
activated nozzle was employed by K. Inagawa, K. Watanabe, and
coworkers. The nozzle injected an argon/nitrogen gas mixture into
the plasma region between the evaporation source, (e-beam gun) and
the substrate (RF biased) enhancing the formation of a stoichiomet-
ric c-BN film of high structural quality. A wide range of substrates
was investigated, and in addition, interlayers were inserted to im-
prove the bonding strength of the films. Only silicon significantly
improved the adhesion of the c-BN films thicker than ~1 microme-
ter, and a graded interlayer (consisting of a mixed compound ranging
from pure silicon over silicon/c-BN to pure c-BN) was most success-
ful. Investigations of the films focussed on tribological properties,
determining a hardness of ca. 6200 HV, a friction coefficient as low
as 0.02 after the initial stage and under vacuum and 400°C, and 0.35
in air at ambient temperature.
RF sputtering was first reported in 1983, depositing from an h-
BN target under N2 and resulted in h-BN films. No c-BN could be
obtained with this technique at that time.[7] Later, films containing
some c-BN were successfully produced, confirming that composite
BN targets of 1:1 stoichiometry dissociate into their atomic com-
pounds, resulting in loss of nitrogen which has to be compensated
with N2 reactive gas additions for stoichiometric films.[8] Analysis of
target emissions using mass and optical spectrometry, as well as
electron probes have been attempted, but did not produce conclusive
results as to the detailed plasma analysis. Substrate temperatures
were 300–400°C.
Figure 4. Activated reactive evaporation (ARE/ADRRP).
Recently, films of predominantly c-BN were produced by RF

sputtering of h-BN under Ar + N2 (Fig. 5). The c-BN phase forms at
low N2 partial pressures at a sufficiently high RF substrate bias. High
compressive stresses result in peeling of c-BN from Si substrates,
which can be prevented with a h-BN buffer layer. Presumably, the
film formation path is that the h-BN target is mainly sputtered in the
atomic state (emission spectra of B2 and BN were not observed).[9]
Further research showed that unbiased substrates inhibit the forma-
tion of c-BN, in which case only h-BN was observed, but with
significant lattice distortion (several percent) due to Ar incorporation
or missing B or N atoms of non-stoichiometric films. Biasing of
substrates enhances c-BN formation, even though it is doubtful that
this is the only phase present.
Figure 5. Reactive RF magnetron sputter ion plating (RF-MSIP).
In early experiments (1980) with ion plating techniques,

Weissmantel successfully produced c-BN films. Pure boron was
evaporated in an electron beam in plasmas of N2, NH3, and a mixture
of NH3 + Ar. In later variations of the ion plating method, a process
called MEP-IP (magnetically enhanced plasma ion plating, Fig. 6),
developed by Watanabe, et al.,[10] consisted of a hot cathode plasma
discharge in a parallel magnetic field. The substrates were RF biased,
and the process gas was a mixture of N2 + Ar. Boron was supplied by
electron beam evaporation of a pure boron source. In a similar setup,
(EB-evaporated B in arc-like plasma of N2 + Ar gas mixture, plasma
near substrate), c-BN films were successfully deposited but delami-
nated on Si substrates due to high internal stresses. B-rich i-BN
(amorphous) has much lower compressive stresses and was pro-
posed as a buffer layer in that particular case.[11]
Several research teams examined pulsed laser ablation for
depositing BN films. First reports were published in 1987, where
decomposition of an h-BN target resulted in h-BN.[2] Later, production
of (unstrained) heteroepitaxial c-BN films was successfully accom-
plished, but particle emission from the pyrolytic h-BN targets was
detected in the films. The formation of c-BN instead of previously
reported h-BN was attributed to the use of an KrF excimer laser (Fig.
7), which has twice the photon energy of the previously used Nd:YAG
or frequency-doubled YAG laser. Nevertheless, the fundamental
relationship between photon energy and the crystal phase is not yet
understood.[12]
Figure 6. Magnetically enhanced plasma ion plating (MEP-IP).
Figure 7. Pulsed laser deposition (PLD).

Chemical vapor deposition (CVD) was one of the earliest

group of techniques employed for producing BN films of various
structures. The first reports of CVD (1979) for the production of
h-BN were based on the decomposition of trichloroborazole (BCl3),
and NH3-H2-Ar. Other possible precursors and processing tech-
niques[13] are: borazine (B3N3H6) at 300–650°C; beta-trichlorborazole
(B3N3H3Cl3) at 700–1100°C; decaborane (B10H14), and ammonia at
600–900°C; boron trichloride (BCl3) and ammonia at 600–900°C;
BF3 and ammonia at 250–600°C, diborane (B2H6) and ammonia at
250–600°C. Especially for CVD, ammonia is the preferred donor for
nitrogen because its higher reactivity as compared to the N2 molecule.
BCl3 based films often contain crystalline boron oxide resulting
from reactions with atmospheric moisture, and borazine based films
deposited at low temperatures dissolve in water. Processes using the
highly poisonous diborane gas require stringent safety measures.
Deposition temperatures below 800°C produce quasi-amorphous
films. B/N ratios close to 1 are achieved only at higher temperatures.
Main control parameters for CVD are deposition temperature and
gas composition. Most CVD films are of pure h-BN or contain a
large fraction of that structure along with c-BN.
As compared to “classical” (thermally activated) CVD, the yield
of c-BN can be improved significantly with plasma-enhanced CVD
(PECVD) methods. In many cases, a lower processing temperature
can be achieved as well. One early example of PECVD is the use of an
RF discharge (Fig. 8) at a deposition temperature of less than 300°C,
with precursor gases diborane, and a mixture of ammonia and hydrogen.
More recently, one example of PECVD employed a micro-
wave discharge at the electron cyclotron resonance (ECR) condi-
tion.[14] Both the self-bias or RF bias of the substrate were investi-
gated, the latter creating an RF plasma near the substrate. Precursors
in this case were diborane and N2. Both techniques produced c-BN
films, but significant substrate bias is necessary for the formation of
the cubic phase. In the case of the self-biased substrates, the sub-
strates were not connected to ground, but allowing them to “float” at
a time-dependant bias voltage determined by the charge carrier
concentration during deposition.
Figure 8. Plasma-enhanced chemical vapor deposition (PECVD), RF substrate

bias.
Among the less common techniques, formation of BN by

nitrogen ion implantation (100 keV) of a previously deposited boron
film has been reported.[15] It produced BN films of unspecified phase
composition, and a significantly graded film structure (depth profile
analysis of B/N ratio). The films were thermally stable up to ca.
1000°C, and proved to be successful oxidation barriers for sub-
strates exposed to a corrosive environment.
As a concluding remark, for most BN thin film deposition
techniques, it can be stated in general that reproducibility, and
preparation of pure c-BN is still difficult.[15][16] Most films contain
some h-BN, and the most common contaminants are unspecified
amounts of hydrogen (from most precursors and/or residual gases),
oxygen and carbon (depending strongly on the deposition process,
e.g., water contamination or use of graphite crucibles for boron
sources). Obviously, the actual hardness of the films depends strongly
on the crystal structure (due to the difference between the values for
c-BN [>4000 HV], and h-BN [1000 HV]). Consequently, the large
scatter in the reported hardness values must be in part attributed to

the difficulty of producing pure c-BN films. An additional factor is
the process dependant difference in c-BN nucleation and grain growth.
2.3 Applications
Wear Protection. The first industrial applications of c-BN

benefited from the extreme hardness of that structure and before
films were available, the most extensive use of c-BN was as a
grinding wheel abrasive. A major advantage over the even harder
diamond material is the fact that c-BN is chemically inert up to 1600
K, and can be used for the machining of ferrous metals. Neverthe-
less, the material has its limitations, mainly because it is hard to
sinter (the only process to produce bulk objects of c-BN). The high-
temperature stability of c-BN requires extreme sintering parameters
(i.e., temperature and pressure) that result in the tendency of recon-
version from c-BN to h-BN; therefore, shock wave compression
methods have been developed (as mentioned before), but they do not
result in high bulk density. Due to these limitations, there is wide-
spread use of unsintered, compacted c-BN powder embedded in a
binder matrix, or as an abrasive material on cemented carbide, or
ceramic substrates.
For films in the B-N system, c-BN films have a rather low
friction coefficient of 0.18–0.22; a-BN (amorphous, B-rich) is even
lower at 0.15–0.18, and h-BN higher (up to 0.58).[17] Well adherent
c-BN films have reportedly been deposited on several metallic sub-
strates without interlayers. The friction coefficient of c-BN changes
very little over time, as opposed to a-BN and h-BN, where the
friction coefficient increased by a factor of approximately 5 and 2.5,
respectively, during 100 test cycles. H-BN also showed a significant
load dependance regarding the friction coefficient; c-BN and a-BN
friction coefficients remained almost constant over a normal load
variation between 0.1 and 50 N. Scratch test evaluations in the same
study determined the highest critical load of up to 110 mN for c-BN,
followed by up to 100 mN for a-BN, and up to 57 mN for h-BN
(substrate: silicon). These films were deposited with the magneti-

cally enhanced plasma ion plating (MEP-IP, Fig. 6) method.
A major limitation to current commercial use of c-BN films is
debonding when applied to conventional cutting tools.[18] The maxi-
mum thickness of 1 µm without spalling on tool steels has been
reported, which is not sufficient for most applications. Tools of
complex geometry (e.g., drills, screw taps) have been coated suc-
cessfully, e.g., with the MEP-IP process (Fig. 6). To solve the
debonding problem, several suggestions have been made: for ex-
ample, multilayers are proposed with a ductile Ti layer on the
substrate, followed by a hard, wear resistant a-BN interlayer beneath
the final (softer) h-BN which proved somewhat better for sliding
wear vs. stainless steel (soft material). The friction coefficient of the
resulting structure is slightly higher, but more constant than for c-
BN, according to this investigation. Another approach suggests a
graded a-BN to c-BN film to reduce compressive stresses at the
substrate-film interface.[11]
For other wear protection applications in the engineering field,
c-BN is seen as a candidate for glass castings or plastic injection
molds. Due to its good thermal and chemical stability, it is expected
to improve the mold-release performance, and to reduce sticking on
the surfaces, thus increasing tool life and improving product surface
finish.
It has been pointed out that the low substrate temperature
(300–400°C, e.g., for sputter deposited films) enables coating of
various types of tool steels, widening the field of possible tool
applications. Therefore, c-BN coating technology is not limited to
cemented carbide tools, as is the case for many other hard coatings.
Semiconductor Electronics. It has long been established that
B-N materials are wide bandgap materials, making them excellent
insulators. At the same time, c-BN has high thermal conductivity,
and has been tested as a heat sink for semiconductor lasers, micro-
wave, and other electronic devices.
Doping c-BN with Si or Be results in n-type or p-type
semiconductors, respectively, which has led to the production of
light-emitting p-n junction diodes. As they retain their high thermal

stability, they are especially attractive candidates for high tempera-
ture electronic applications.
Films of c-BN are most promising for semiconductor applica-
tions, but the development of electronic devices of that material has
begun just recently. Some limitations due to process characteristics
will be encountered, such as particle contamination reported for
pulsed laser deposited c-BN films.[12] Up to now, good results have
been obtained with bulk materials, such as the large crystals grown
by the bulk temperature difference method that were used to produce
a p-n junction.[19] A Be-doped p-type crystal was used as the seed for
growing the n-type c-BN to it. The n-type was achieved by adding Si
as a dopant to the LiCaBN2 solvent. The resulting crystal shape and
geometry were difficult to control, but a diode, and a resulting
optoelectronic device (LED, emitting blue light) were successfully
fabricated, even though purity and crystal quality of the c-BN were
not perfect. For high-temperature applications, the device is ex-
pected to be operable at temperatures well above 1000°C.
Another application aims at wear protection on electronic equip-
ment, such as c-BN coatings for magnetic recording head media
interfaces (sliding contacts, tape heads, etc.).[20] A related compari-
son of wear characteristics of possible materials in the B-N system
determined that h-BN was the least wear resistant material (regard-
ing hardness, scratch test, reciprocating cycle tests, friction coeffi-
cient), followed by a-BN; best performance was displayed by c-BN.
Due to a favorable lattice match to both silicon and diamond, c-
BN is suggested as a buffer for epitaxial diamond coatings on Si for
improving adhesion of the diamond film to the silicon substrate.
Another multilayer application of c-BN films as an x-ray lithography
mask membrane (Fig. 9) has been investigated extensively.[13][21]
The properties of c-BN make it an ideal material for this device used
to fabricate very large scale integrated circuits (VLSI), requiring
feature dimensions of 0.5 µm or less. The c-BN membrane serves as
the substrate which has to transmit radiation; the actual patterning is
achieved with the absorber pattern of a high-absorbing material (e.g.,
Au, W). The benefits of the BN substrate are; a very low x-ray
absorption coefficient in the 4–12 Å wavelength range (such as
for Pd L-alpha, C K-alpha, B K-alpha radiation), good mechani-
cal and dimensional stability, adequate optical transparency for
alignment, similar thermal expansion coefficient to silicon which
is used as the substrate frame, chemical inertness, and low defect
density. It has been mentioned that films do not necessarily have
to be of c-BN composition to be operable, allowing for some h-
BN in the structure.
Figure 9. X-ray lithography mask membrane.
Optics. Only a few investigations focus on the application of

BN films for optical purposes,[18] the greatest attraction arising from
the tribological, and thermal properties of c-BN to be used in coat-
ings of optical windows (e.g., zinc selenide, zinc sulfide). Optical
properties of c-BN, and probably additional interlayers have to be
taken into consideration for the individual applications, such as laser
or spectral investigation windows. One example for multilayer opti-
cal windows is the combination of c-BN coatings with diamond
films on windows, designed for wear resistance at elevated tempera-
tures. Deposition of c-BN on a diamond film with good adhesion
between the two hard materials has been demonstrated.
3.0 DIAMOND-RELATED MATERIALS
3.1 Structure and Properties of Diamond-Related Materials
The discussion as to the correct name for the synthetic diamond

related materials is based on the findings that most DLC films are
“amorphous” (as characterized by x-ray diffraction analysis), which
by definition should not be called a “diamond film.” Nevertheless,
these films contain nanocrystalline diamond phases (that can give
the film highly diamond-like properties), along with graphitic and
other non-diamond phases. In some publications, the definition of a
“diamond”, or “diamond-like” film is based on the volume fraction,
and size of nanocrystalline diamond structures. When a clear x-ray
diffraction signal of polycrystalline diamond can be identified (how-
ever weak or distorted it may be), or the film shows the characteristic
Raman shift for diamond at 1332 cm-1, the film is said to be a
“diamond film.” If the film shows neither an x-ray diffraction signal
nor a Raman shift that can be attributed to diamond, but has obvi-
ously diamond-like physical properties (e.g., hardness, optical and
IR transmissivity, chemical resistance), the film is referred to as
“diamond-like”. Since there is no clear indication in each publication
as to which definition is used, we will speak of “diamond-related
films” in a very broad sense.
In the current literature, various names describe synthetic dia-
mond-related materials, such as diamond-like carbon (or DLC), a-C
(for amorphous carbon), diamond, a-C:H (for amorphous carbohy-
drate), h-DLC (for hydrogenated diamond-like carbon), just to name
the most common versions; additional nomenclatures also exist, and
are briefly mentioned later. Nevertheless, the major differences be-
tween the synthetic diamond-related materials is whether they do, or
do not contain significant amounts of hydrogen. This distinction is
not only useful to categorize two major approaches to produce these
materials, but it also permits differentiation between two groups of
materials that are significantly different in their physical properties
and the different resultant applications.
Elemental carbon forms two natural crystallographic states,

namely graphite, and diamond. They are different in almost every
physical property. In addition to the two pure forms of carbon, two
intermediate forms—amorphous and amorphic carbon—are of tech-
nological interest. These two subsets together as a whole are often
referred to as a-C. The major differences between the four forms will
be described briefly in the following paragraphs.
Graphite is composed of a planar structure of carbon hexagons.
The sheets of hexagons are comprised of weakly bonded C(sp2)-
C(sp2)3-Pi-units, contributing to the mechanical properties of the
soft graphite. The sheets are connected by van der Waal’s bonding
weaker than the in-plane Pi bonds, leading to a significant anisotropy
(e.g., regarding shear). Graphite has a density of 2.6 g/cm3, which is
much lower than that of diamond (3.5 g/cm3).
Natural diamond is usually crystalline (cubic), optically trans-
parent, and electrically insulating. It has the highest atom number
density of all known materials. Structurally, each carbon atom is
strongly bonded to four nearest neighbors. This diamond-like bond-
ing type has also been described as a tetrahedral coordination of
C(sp3) atoms, i.e., a C(sp3)-C(sp3)4 bonding unit. Natural diamond
contains very low amounts of impurities, especially gases such as
nitrogen, boron, and hydrogen (a few percent). Synthetical H-DLC
(described in detail below), in contrast, may contain up to 50 at.%
hydrogen.
In addition to these two states of carbon, some intermediate
forms exist, with physical properties between the two pure forms,
namely amorphous and amorphic carbon. These intermediate forms
are much easier to synthesize than crystalline diamond materials, but
often have physical properties close to diamond, making them inter-
esting candidates for technical applications. In this context, “glassy”
or “amorphous carbon” describes an amorphous material of pre-
dominately diamond-type sp3 bonds that lacks a long-range ordered
structure (hence the name “amorphous”). Amorphous carbon con-
tains little or no hydrogen, making it harder than H-DLC. It is
electrically insulating, chemically inert, and optically transparent.
Table 2. Properties of Diamond-Related Materials
Property Diamond Graphite
Crystal structure cubic, zinc blende hexagonal
Lattice constant a = 3.56 Å a = 2.47 Å

c = 6.707 Å
Density 3.5 gcm-3 2.6 gcm-3
Hardness 10000 HV 12 HV
Resistivity (RT) undoped: 1014 ohm·cm in-plane: .04 ohm·cm

p-type (B): 10–1000 ohm·cm c-axis: 0.2 ohm·cm
Activation energy p-type (B): 0.37 eV n-type (P): -0.2 eV
Thermal conductivity 20 Wcm-1K-1 0.8–2.5 Wcm-1K-1
Lin. therm. expansion (300K) 0.8 × 10 -6 K-1 8.8 × 10 -6 K -1 (in-plane)

27.3 × 10 -6 K-1 (c-axis)
Similarly, amorphic carbon can have many diamond-like prop-

erties. It consists of small (15–20 Å diameter) islands of sp2-bonded
(hexagonal) graphite that are linked by disordered diamond-type sp3-
bonds. Obviously, the smaller the graphitic islands, the larger is the
fraction of sp3 bonds, resulting in physical properties closer to diamond;
hence, the material sometimes is referred to as amorphic diamond.
When diamond contains considerable amounts of hydrogen (as
in H-DLC films), it significantly reduces optical transparency, hard-
ness, and electrical resistivity. There are important structural differ-
ences between DLC and H-DLC, as the hydrogen stabilizes the
(“diamond-type”) sp3 bonding typical for H-DLC, even though the
material may have properties very much different from diamond.
Recent research indicates that H-DLC films consist of a random
network of sp3 (diamond), and sp2 (graphite) C-C and C-H bonds,
with no specific order for the hydrogen sites, various voids, and
unbonded hydrogen. These films are almost completely amorphous,
whereas DLC films are often nano- to microcrystalline. For films

with very high hydrogen content, up to 95% of the C-H bonds are
sp3, whereas the majority of C-C bonding is sp2.
Diamond-related films are usually referred to as “hydroge-
nated diamond-like carbon” (H-DLC) when their hydrogen content
is ca. 20%, or more. The lower limit allows for some hydrogen in so-
called unhydrogenated films, mostly due to unwanted hydrogen
incorporation from background gases or contaminants, whereas H-
DLC films obtain their hydrogen from the reactive gases, e.g.,
methane or acetylene. The density of these films is lower than that of
DLC films, typically around 1.8 g/cm3.
In the current literature (especially when dealing with natural
diamond) a classification scheme for diamond has been established
that is based on purity and does not follow a strict scientific standard.
The most popular scheme divides diamonds into four classes (Ia, Ib,
IIa, IIb, with subclassifications possible into IaA and IaB) in the
order of increased purity. Starting at a nitrogen level of up to 0.3 at%
in the I class, a distinction is made as to the form of nitrogen
inclusion (Ia: pairs (A), and quadruplets (B) of nearest neighbor nitro-
gen atoms, and most common in natural diamond, Ib: single substitu-
tional nitrogen, rare in nature but common in synthesized diamond),
whereas the purer II-types are subdivided according to their electronic
properties (IIa: insulating, IIb: p-type behavior due to substitutional
boron). The distinction between type I, and type II is usually based
on the infrared phonon-absorption detectability of nitrogen.
3.2 Synthesis of Diamond-Related Materials
Synthesis of Bulk Materials. As early as 1954, bulk diamond

of small size has been synthesized from graphite by the catalyst-
solvent method, requiring temperatures of at least 1500°C and pres-
sures above 50 kbar, as well as growth-enhancing catalysts (e.g., Fe,
Ni, Co, Mg, Ta, Cr). Under these conditions, diamond is the thermo-
dynamically stable form of carbon. The morphology of these crystals
is quite good, they often contain graphitic contaminants (resulting in
black color), and the synthesis is by now a commonplace item in

industrial applications (predominantly as abrasives). At this time,
production of low-impurity crystals suitable for optical, or electronic
applications is still too expensive. Several laboratories have grown
large crystals (up to 17 mm in diameter), of high purity with this
technique, but their production cost is too inhibitive for commercial
purposes. In 1961, the shock wave method was another high-pres-
sure process to produce diamond-containing material from graphite.
The technique is similar to shock wave processing to produce, or
compact c-BN (as described above), but no application of this tech-
nique in order to compact diamond is known.
Another approach of considerable interest for the growth of high-
purity diamond, is the combustion or flame torch method, burning an
octacetylene (mix of oxygen and acetylene of near-unity gas flow
ratio) flame (Fig. 10). The process can be used for the deposition of
small size bulk crystals, as well as diamond-related thin films. The
diamonds are grown in a limited area (primary combustion zone) of
the flame cone (characterized by a temperature of ca. 3000°C, and a
gas composition of predominantly CO and H2, and distinguishable
by a light incandescent glow) on a substrate. The substrate has to be
cooled from the rear in order to maintain a temperature between 800
and 1050°C. With substrate scanning, large-area deposition is possible.
Figure 10. Flame torch method.

The technique was first successfully applied in 1988, when it

was performed in ambient atmosphere, employing a common weld-
ing torch. Several advantages of this method quickly became obvi-
ous, namely the simplicity and availability of equipment, operability
in ambient air without vacuum chambers, a very high growth rate (up
to 200 µm/hr for diamond particles, and up to 900 µm/hr for dia-
mond-related films), and possibility of large-area coverage. How-
ever, the growth characteristics of this method are difficult to con-
trol, since the pointed flame creates severe temperature gradients in
the substrate (resulting in warping or cracks), and the radially inho-
mogeneous composition of the flame leads to the deposition of
unwanted non-diamond materials in the perimeter areas of the flame.
Also, effective substrate cooling is essential, thus, limiting usable
substrate geometries somewhat.
Since its introduction, several improvements of the technique
have been reported, such as improved burner design, use of carbon-
carrying gases other than acetylene, precise gas composition control,
and exclusion of ambient air by using a protective sheath of argon, or
running the experiment in vacuum. These refinements have resulted
in a significant increase of the diamond fraction (by reducing graph-
ite and soot formation), high density and polycrystallinity of large-
area films, and even doping for the production of semiconducting films.
Synthesis of Thin Films. As mentioned above, there is a
distinction regarding the hydrogen content of diamond related films.
To produce H-DLC films, hydrocarbon precursor gases such as
methane or acetylene are used (sometimes together with argon or
pure hydrogen), whereas the deposition of H-free DLC films is
achieved by evaporating pure carbon (e.g., graphite), or by
mass-selecting plasmas of hydrocarbon gases, or by fluorocarbon
chemistry.
All film deposition methods operate at low pressures, where
diamond is metastable (as opposed to high pressures), and under
these conditions competes with graphite formation. When deposit-
ing diamond-related films, several effects should be considered
because the resulting structural properties of the films are controlled
by fundamental deposition parameters. One important aspect for

many applications is the magnitude of stress and strain, which is a
result of the link between hydrogen content and ion bombardment.
Some studies reported a correlation between hydrogen content and
compressive stress for films prepared with hydrocarbon gases, claim-
ing that very low hydrogen contents are responsible for the high
compressive stress values, whereas recent studies attribute the stress
mainly to high-energy ion bombardment, e.g., by additional argon
plasmas. That does not contradict the previous findings, as this
bombardment eliminates much of the hydrogen in the film, thus
lowering the hydrogen content. On the other side, DLC films grown
from hydrogen-free precursors have shown extremely low internal
stresses. Therefore, hydrogen-free films of low compressive stress
should be grown by hydrogen-free methods rather than a combina-
tion of hydrocarbon precursors and ion bombardment for hydrogen
removal.
Regarding the thermal stability of diamond-related films of
high hydrogen content (14 vol% diamond-like, rest polymeric), it
has been demonstrated that annealing up to 650°C causes a complete
transformation of that phase into graphite, as the sp3-stabilizing
hydrogen presumably leaves the structure, via an increasing volume
of voids. This graphitization (conversion to sp2 bonds) goes along
with a loss of transparency, hardness, and changes in other physical
properties. Similar findings were observed when hydrogenated dia-
mond-related films were exposed to laser radiation, and post-deposi-
tion ion and electron bombardment. For laser irradiation, both pulse
energy, and laser power determine the damage threshold, depending
on the photon energies above or below the bandgap, respectively. In
that context, an investigation on the effects of irradiation energy
levels showed that laser irradiation caused the most drastic changes
of the film structure (i.e., graphitization), whereas ion bombardment
caused merely a loss of hydrogen (affecting electrical and optical
properties) without significant graphitization. Electron bombard-
ment affected the films only mildly (no detectable hydrogen effusion
or graphitization), modifying the optical and conducting properties
of the film. It has been pointed out that hydrogen-containing dia-

mond related films deposited at elevated temperatures undergo a
self-annealing during deposition, thus resulting in a graded structure
and hydrogen content depending on film thickness.
Numerous deposition methods have been developed and inves-
tigated, of which the most important can be categorized into such
process groups as physical vapor deposition (PVD), plasma-en-
hanced chemical vapor deposition (PECVD), pyrolysis processes,
hot filament methods, and flame torch methods.
Physical Vapor Deposition (PVD). Pure, solid carbon (graph-
ite) has been decomposed to provide the carbon required for dia-
mond-related films by several methods of PVD, such as magnetron
sputter ion plating (MSIP, similar to Fig. 5), activated reactive
evaporation (ARE, similar to Fig. 4), ion beam deposition (IBD), and
pulsed laser deposition (PLD, similar to Fig. 7). Most techniques can
operate with or without additional hydrogen or hydrocarbon gases,
deposit films at rather low substrate temperatures and produce mostly
amorphous or microcrystalline films with varying fraction of the
diamond phase. The highest reported fraction of diamond-type sp3
bonding of 85–90% (in an amorphous film) has been synthesized by
mass-selected ion beam deposition (Fig. 11). Even though the depo-
sition rate of some techniques is quite low, commercial and scientific
interest in these deposition methods is high, mostly due to their high
degree of flexibility, (in regard to parameter variations) and material
purity.
ARE (in similar setups to the one shown in Fig. 4) has proven
to be a powerful process for the deposition of diamond-related films,
as it provides complete decoupling of relevant deposition parameters
that control the structure of the films. Variable parameters include
the means of evaporating the graphite (usually an electron beam
gun), interactions between the evaporated carbon and the process
gas, substrate temperature, plasma activation (RF DC power), sub-
strate biasing etc. The process is capable of depositing nanocrystalline
diamond films at a growth rate of over 1 µm/hr at substrate tempera-
tures below 500°C.
Figure 11. Mass-selected ion beam deposition (MS-IBD).
Sputter deposition of a graphite cathode (MSIP, general setup

shown in Fig. 5) has demonstrated that it is possible to vary the film
structure from purely graphitic, to amorphous carbon to diamond-
like by varying the deposition parameters (namely substrate tem-
peratures of 300–600°C, and gas pressures of 0.1 to 1.5 Pa). Both
pure Ar and Ar + H2 atmospheres have been applied. Film properties
depend primarily on hydrogen content, and the energy of the ion
bombardment, where hydrogen supports the formation of the sp3
bondings typical for diamond, and increased Ar ion energy results in
more diamond-like structures.
At Stanford, sputtering of a graphite cathode in a helium
atmosphere (without hydrogen) was applied to produce a film of
presumably amorphous or graphitic carbon, which in a sequential
step was exposed to a pure hydrogen plasma (generated by a fila-
ment) to form the desired diamond film. This is one of the few
examples demonstrating the importance of hydrogen to diamond
formation, because of the clear separation from other, usually super-
imposed surface effects which occur during the sputtering with gas
mixtures of e.g., Ar or Ar+H2.
In ion beam deposition (IBD), (as early as 1969 and with setups
similar to Fig. 3), graphite electrodes were dissociated by argon
sputtering into carbon atoms in the gas phase which were subse-
quently ionized in an argon plasma, and the ions were accelerated
onto the substrate. The films were amorphous with diamond-like
physical properties, such as transparency, electrical insulation, resis-
tance against acid attack and electronic structure. The argon pressure
was kept low to minimize its effect on film contamination, but at the
same time the deposition rate was slow, because graphite is a mate-
rial of low sputter yield. In attempts to increase the deposition rate,
later developments focused on the use of sputtering gases other than
argon, such as to apply other techniques as ion sources, such as
cathodic arc evaporation or high-rate arc decomposition of methane-
argon gas mixtures.
Ion beam deposition (IBD) is also a very flexible process
applied to evaporate solid graphite for the production of mostly
hydrogenated diamond-related films, employing either hydrocarbon
or hydrocarbon/argon ion beams. It has been reported that applica-
tion of sufficiently high bias voltages to the substrate causes a
resputtering effect, thus, removing much of the hydrogen (as well as
some of the carbon, thus, reducing deposition rate), and creating low
hydrogen films with diamond like characteristics. Nevertheless, the
significant amount of ion bombardment onto the substrate due to the
bias voltage results in rather high substrate temperatures, thereby
limiting the substrate materials available for deposition. This
resputtering effect has been observed to be the key factor for the
deposition of non-hydrogenated films with diamond like properties,
making the (kinetic) energy of the impinging ions the most signifi-
cant parameter for the distinction of depositing DLC or H-DLC films
even from hydrocarbon gases. The optimum energy for C+ ion
beams seems to be about 100–120 eV. Regarding the upper limit,
energies as high as 600 eV have also been successfully applied to
produce diamond-like films, indicating that the low sputter yield of
carbon tolerates high-energy ion bombardment. Extremely high ion
fluxes and energies lead to graphitization, due to overheating of the
growing film during deposition. However, for actual applications,
one wants to keep the ion energy as low as possible, because of the
close correlation between the energy of the impinging ions, and

resulting substrate temperature. Deposition temperatures of 30°C
have been reported.
In the face of the importance of ion energies for the deposition
of diamond related materials, IBD techniques can be divided into
two major categories: Indirect and Direct Deposition.
In indirect deposition, an ion gun (preferably Ar) irradiates a
carbon target which decomposes and deposits carbon onto the substrate
(similar to sputter deposition). Occasionally, an additional gun may be
used to modify the growing film, e.g., by irradiation of Ar or Ar/H2.
In direct deposition, the ion source produces carbon ions (e.g.,
from pure carbon or methane) which are extracted and directly
deposited onto the substrate. Both systems provide quite different
ion energies for the growing film (indirect being lower than direct),
and direct IBD enables better control over the energy and species of
the material arriving at the substrate.
The quest for very low impurity, and high-sp3 yield diamond
films has led to the development of mass separated IBD (Fig. 11),
which applies a magnetic field to the beam exiting the ion source to
direct only selected species of ions onto the substrate. The spatial
separation of the ion source from the substrate permits the variation
of several key parameters, especially ion energy (10 eV to 1 keV),
and substrate temperature. It also enables deposition at very low
background pressures near the substrate, allowing for several in-situ
UHV characterization methods such as Auger, x-ray, and UV photo-
electron spectroscopy, low energy electron diffraction, residual gas
analysis, electron energy loss spectroscopy, and ionization loss spec-
troscopy. Combinations of several different ion beam sources within
the same deposition chamber make it possible to irradiate the grow-
ing film with a controlled beam of argon, hydrogen or other gases to
study doping or surface effects. It has been suggested that this
method might be one of the first to produce heteroepitaxial diamond
films. Several drawbacks of this method (low deposition rate, line-
of-sight deposition, hardware cost, etc.) do not make it feasible as a
commercial production process, but do not diminish its qualities as a

powerful research tool.
First results of the application of high-intensity laser irradia-
tion for the formation of diamond were reported in Russia, consisting
of the polymorphic transformation of graphite to diamond. Whereas
this process is rather a posttreatment method, than a pure film
deposition process (because the graphite was already on the sub-
strate, and no material transfer was involved), recent developments
in laser technology have resulted in various attempts to apply the
laser to the deposition of diamond films. Two main approaches of
direct laser application are currently being investigated, one using
the laser as an evaporator for the graphite (similar to the configura-
tion in Fig. 7), and the other generating the plasma that is to decom-
pose hydrocarbon gases. Recent research in indirect laser (assisted)
diamond film growth includes the production of seed films on
copper from liquid benzene with subsequent diamond film deposi-
tion via hot filament CVD.
Plasma-Enhanced Chemical Vapor Deposition (PECVD).
In the initial phase of diamond film research, a major setback of most
hydrogen free PVD techniques was a low deposition rate and a
comparably low diamond fraction in an otherwise mostly graphitic
film. Even though this is no longer true (due to constant improve-
ments in process technology, plus the fact that hydrogen, and hydro-
carbon gases are widely used by modern PVD techniques), it sparked
the development of chemical vapor deposition (CVD), which relies
solely on gas phase precursors, omitting graphite as the source for
carbon. In order to break up the hydrocarbon gas and to generate
activated particles that facilitate growth of a diamond-related film, a
plasma is applied by coupling of additional energy in most cases.
This definition also distinguishes PECVD from pyrolysis or the hot
filament method. Nowadays, it is difficult to draw a strict line
between PVD and CVD deposition methods for diamond related
films, mostly due to the complex plasma chemistry involved in the
decomposition of hydrocarbon gases frequently applied in what
often started out as a “classical” PVD technique.
In principle, three approaches of energy induction can enhance

a plasma (Fig. 12): (i) direct current (DC), (ii) radio frequency (RF),
and (iii) microwaves, (especially ECR at suitable process parameters).
All three have been investigated extensively. DC plasma activation is
limited to the deposition of conducting diamond-related films, such as
doped (semiconducting) structures, or coatings with higher graphitic
contents. It is not very common in high-purity diamond film research.
(a)
(b)
Figure 12. Plasma-enhanced chemical vapor deposition: (a) DC enhancement;

(b) RF enhancement; (c) microwave enhancement.
(c)
A DC Plasma Assisted CVD process is called HCDCA-Hol-

low Cathode DC Arc.* The apparatus is the same as the PVD hard
coating apparatus developed more than a decade ago by Balzers for
hard coatings like TiN and published extensively in the technical
literature. The substrates are located around a long DC arc column
and are not heated. The reactant gases are argon, hydrogen, and
methane. The species needed for diamond deposition are produced
in the plasma column and diffuse to the substrate. Substrate tempera-
ture is a balance between heating by H ion recombination and
thermal radiation loss.
* Ref. H. Bollier and J. Karner of Balzers Lichtenstein—presented at the Gorham

Conference on Hard and Superhard Materials, Atlanta, March 1996 and pub-
lished in the Proceedings of that Conference.
RF has proven to be a very versatile technique, as it usually can

be generated over larger areas as compared to microwaves, and is
more reliable for the deposition of thick insulating films (such as
diamond), as compared to constant current activation. RF is also the
technique most commonly used to bias substrates for effective ion
bombardment, often at frequencies above ion plasma frequency (ca.
3–5 MHz) in order to employ the mobility difference between (high-
inertia) ions and electrons. Even though substrate biasing is not
acknowledged as a plasma activation method in the traditional sense,
it influences gas phase and surface reactions considerably. As with
DC activation, RF depends on electrodes to induce the electrical
energy into the plasma and field effects usually have to be taken into
account when designing experimental setups. In most cases, the
vacuum chamber serves as the grounded electrode.
Microwave activation is capable of achieving the highest pro-
cess temperatures. Thus, it is possible to drastically increase reaction
temperatures (e.g., 5000 K in microwave plasmas, as compared to
2300 K for the hot filament method), thereby changing the decompo-
sition conditions significantly. Accordingly, several research teams
have investigated the effects of increased plasma temperature by
thermodynamical calculations, and plasma analysis. One major find-
ing is that a methane hydrogen precursor decomposes into CH3
radicals, atomic hydrogen, acetylene (C2H2) plus smaller quantities
of various other carbon related radicals, and the ratio of these reac-
tants is largely a function of the plasma temperature: at lower
temperatures, CH3 is the major reaction product, whereas at 5000 K
it is almost undetectable, and acetylene is the dominant species. Gas
phase composition and substrate surface energy are the key factors
controlling the formation of the growing film structure.
Frequently, combinations of plasma activation methods are
applied, as for instance the combination of microwave activation and
RF biasing. The microwave energy is primarily focused on the gas
reactions to produce a certain ratio of highly energetic, and ionized
gas decomposition radicals, and the RF bias controls the energy level
at which the ions arrive at the substrate surface. This makes it
possible to decouple individual process parameters in order to tailor

deposition conditions.
Pyrolysis Processes. In contrast to the aforementioned PECVD,
pyrolysis processes are capable of depositing epitaxial diamond
films without additional gas phase excitation such as electric fields,
plasmas, or filaments. Although both diamond and graphite are
simultaneously deposited, due to the very high etching rate of graph-
ite in this process, it is eliminated on the heated substrate during film
growth, leaving only the diamond (which has a much lower etch rate)
behind. It has long been known that hydrogen ions can achieve that,
but much more effective etchants are fluorocarbons.
Several precursor gases have been applied (some of them
enabling the growth of hydrogen-free films), such as methyl fluoride
(CH3F), mixed with H2, and diluted in helium, a mix of CS2 and
fluorine (F2), or CF4 and F2. Growth rates of over 0.5 µm/hr at
substrate temperatures below 900°C have been reported. In addition
to the etching mechanism, the fact that diamond films could be
deposited without using hydrogen has received much attention and
spurred the discussion on the reaction mechanisms for diamond
formation in general. Many uncertainties still remain despite intense
efforts to explore the thermodynamics of diamond growth kinetics
(often by means of calculations and simulation), but it seems that
atomic fluorine (similar to hydrogen), promotes the formation of sp3
bonds that are essential to diamond formation. Thermodynamic
calculations support the finding that the partial pressure of molecular
fluorine required for diamond formation is at least two orders of
magnitude lower than hydrogen, thus, explaining why less energetic
deposition conditions (no added energies except substrate heating)
are successful.
Hot Filament Methods. The hot filament method was one of
the first viable means to deposit diamond related films. It has re-
ceived much scientific attention, and substantial improvements to
the original experimental setup have been made since then. The key
component in this technique is a tungsten filament mounted close (a
few mm distant) to the substrate, providing thermal energy and
electron emission to generate a plasma that decomposes a mix of a

carbon-carrying gas and hydrogen.
The major radicals resulting from the fragmentation of the
precursors most commonly used, methane and hydrogen, are CH3
(methyl radicals), C2H2 (acetylene), and large amounts of atomic
hydrogen. The gas phase chemistry leading to reliable diamond
formation has been discussed broadly. The resulting films respond
very sensitively to filament to substrate distance, and it has been
determined that this is mostly due to the spatial variation of the
decomposed gas phase products in the vicinity of the filament. A
major role in this radial distribution is attributed to atomic hydrogen
which is produced or consumed in various reactions depending on
the local energy level, ranging from endothermic atomic hydrogen
formation close to the filament to exothermic recombination to
molecular hydrogen at larger distances. However, a significant amount
of atomic hydrogen is required at the substrate surface for a number
of reasons (heat transfer, sp3-bond stabilization, “graphite etching,”
conservation of diamond-forming carbon carrier species), and there-
fore the substrate to filament distance is limited.
A considerable increase of the deposition rate (up to 8 mm/hr,
as opposed to 0.5 mm/hr previously), was achieved when Matsubara,
et al., used a tantalum filament, instead of tungsten to decompose a
methane-hydrogen gas mix. This was attributed to the higher melting
temperature of TaC (which forms as a reactive product on the
filament), which is 3985°C, as compared to 2785°C for WC (or
3380°C for pure W), and makes it possible to run the experiment at
higher methane concentrations, and it was also observed that deposi-
tion rates increase with higher filament temperatures. The maximum
deposition rate was obtained at 2600°C, which was about 600°C
above the limit for a tungsten filament.
The methyl radicals are believed to be the sole contributor to
diamond film formation (assisted by atomic hydrogen), whereas
acetylene may have an inhibiting influence (at least in the relevant
temperature range of up to ca. 2300 K). Therefore, modifications in
the hot filament technique aim at a preferred generation of CH3,
which was achieved by adjusting gas flow (speed and direction), in

the filament/substrate region and by controlling substrate tempera-
ture (decoupling from filament radiation by cooling/heating). Plasma-
enhancement by means of microwave or RF excitation is also widely
investigated.
Flame Torch Method. The flame torch method for growing
films is essentially the same simple process as described in the bulk
processing of diamond earlier in this chapter. It employs the same
slightly sub-stoichiometric mix of acetylene and oxygen in a burner
(Fig. 10). Diamond growth takes place in the primary combustion
zone of the torch flame, with gas temperatures of ca. 3100°C, and a
gas composition of primarily H2 and CO.
For classical thin film applications, this method has several
drawbacks, particularly the small deposition area, noncontinuous
film morphology comprised of mostly individual and loosely packed
particles, and the difficulty to control the purity, and structural
composition. The first constraint can be solved by improved burner
nozzle design, and guiding or sweeping burner rigs, but the second,
and third problems are inherent to the technology and limits its
scientific and practical application in comparison to other thin film
techniques available. Nevertheless, the technique has received much
attention, and considerable success has been achieved in improved
burner design, optimized gas composition, and reduced fuel con-
sumption. Also, the interest in this method remains high, due to its
simple practical application, and the highest deposition rate of all
diamond-related thin film processes (up to 900 mm/hr), marking it as
a very economical way to produce “rough” diamond films for appli-
cations such as thermal management substrates, where the draw-
backs of this technique are not important.
Plasma-Enhanced Chemical Transport Reaction. A recent
process development at UCLA, uses the high surface mobility of
activated species to deposit diamond films from a DC hydrogen
plasma generated between two graphite electrodes. A major differ-
ence to most CVD processes is the closed-system characteristic,
because the deposition does not require a continuous flow of
precursor materials; the process chamber is filled with hydrogen at

the beginning of the experiment and then sealed, drastically reducing
operating costs. The process can also be operated under interrupted
flow or continuous flow conditions. The electrodes are of different
diameter in order to maintain a specific discharge characteristic. In
this very simple and straightforward experimental setup, the sub-
strate is directly exposed to the plasma discharge by positioning it on
one of the electrodes. Highly crystalline diamond films have been
grown on substrates of carbides and carbide-forming elements such
as tungsten, molybdenum, germanium, tantalum, and silicon. SEM
investigations verify the presence of a continuous diamond film. The
morphology of the film surface can be modified by varying the
process conditions and dense films of well faceted crystals can be
grown. The deposition rate is quite high, with a maximum of 1 mm/
hr. There is no limitation to thickness of the films.
The direct exposure to the plasma opens the potential for easily
doping the diamond film, and it has successfully been demonstrated
that films with varying dopant levels of sodium, phosphorous, lithium,
and other elements can be grown. In most cases, the dopant is
supplied from a side-mounted crucible. With this approach, p-type
semiconducting, highly conducting diamond films have been depos-
ited, and the potential for synthesizing n-type diamond is being
investigated.
At present, the substrate size is limited by the diameter of the
electrode (or the extension of the resulting plasma), but a potential
for large-area deposition of diamond films is to be expected by
upscaling the existing setup, while at the same time preserving the
potential for doping. This might give this technique an advantage
over most other processes that have an obvious small-area character-
istic, such as the flame-torch-/arc jet methods, hot filament CVD or
microwave plasma CVD.
Laser Assisted Diamond Deposition at Ultra High Rates.
Pravin Mistry, in his attempts to deposit titanium by a new laser
process, discovered the deposition process for diamond when a
nitrogen tank was accidentally replaced by a carbon dioxide tank
(New York Times, March 12, 1996). This new process enables the
deposition of diamond films at ultra high rates in the absence of
hydrogen commonly a reactant in most CVD and PVD processes.
This QQC Diamond Synthesis technique uses a mixture of carbon
dioxide (source of carbon), and nitrogen in an open atmosphere (no
vacuum or container). The materials processing cell has three lasers
focusing on the object to be coated—a UV excimer laser, an infrared
Nd:YAG laser, and an infrared CO2 laser. Objects from 1 to 30 cm in
dimension can be coated at a rate of 1 mm/sec with coating thickness
up to 45 mm. The coatings have been tested on cutting tools. They
have excellent adhesion and their cutting performance is equal to
polycrystalline diamond sintered bodies.*
Substrate Considerations. Nucleation and crystal growth are
very important, as they influence the morphology of the diamond-
related films, resulting in a wide range of films structures and surface
topographies. It has been proven very viable to prepare the substrate
surface accordingly, and in many cases nucleation seeds are in one
way or another, applied on the substrate prior to the actual DLC or H-
DLC film deposition. Successful seeding materials are graphite, dust
of natural diamond or silicon carbide, or carbides that are deposited
on the surface. In most practical applications, the substrate surface is
polished with these or similar abrasive materials, where diamond
paste is by far the most effective material. This supports the assump-
tion that polishing itself may provide some nucleation seeds due to
increased surface energy or microscratches, but that the major effect
is homoepitaxial growth of “diamond on diamond” (diamond dust
residue from the polishing paste that remains even after thorough
cleaning), indicating the importance of the nature of the substrate. A
recent suggestion is the conditioning of the substrate surface to
promote diamond-related sp3 bonding, be it by reversible chemisorption
* Ref. P. Mistry, M. C. Turchan, S. Liu, G. O. Granse, T. Baurmann, and M. G.

Shara, in: Innovations in Materials Research, 1(2):192–207 (1996).
of hydrogen or halogens. Hydrogen is also believed to etch or gasify

graphite faster than diamond, which may play a role for the initial
surface conditioning for plasma-assisted deposition methods, thus,
eliminating graphite from the substrate surface, leaving diamond
seeds for further nucleation.
Suitable substrates promoting the growth of diamond can be
divided into three categories: (i) lattice matched and chemically
compatible materials (such as diamond or c-BN), (ii) carbides and
carbide-forming materials (such as Si, Mo, Ta, W, Al2O3, Si 3N4,
selected cemented carbides, etc.), and (iii) some non-carbide form-
ing materials (such as Au, Pt, Ni, Cu). Nevertheless, a strong interde-
pendency exists between the substrate material, substrate prepara-
tion (e.g., nucleation and promotion of sp3 hybridization), and pro-
cess characteristics (e.g., presence of hydrogen).
The growth of heteroepitaxial films is a prerequisite for many
potential applications particularly in microelectronics, and much
effort has been spent to achieve it for diamond films. So far, hetero-
epitaxy can only be maintained in small areas and under certain
circumstances. Some such requirements are lattice match, sp3 hy-
bridization, and a precursor and process parameters that avoid graphi-
tization. The best results have been reported for c-BN substrates, a
material that has chemical and physical properties very similar to
diamond. Very localized heteroepitaxy has been reported for Ni,
SiC, and BeO substrates, and it is hoped that certain buffer layers
may facilitate heteroepitaxial diamond films for a wide range of
substrate materials.
3.3 Applications
Wear Protection. Wear protection is so far the biggest com-

mercial application for diamond-related films, and Japan has taken
the lead in that field. Diamond-coated tools are predominantly used
in the machining of nonferrous (mostly high-silicon aluminium)
alloys in the automotive industry and significant growth rates of this
application are expected in the future.
As pointed out earlier, the properties of diamond-related films

vary widely, depending on structural composition. A good overview
over several basic mechanical properties (Young modulus, bulk
modulus, shear modulus, Poisson ratio, and yield stress) of various
types of synthesized diamond-related materials has been compiled in
the literature.
Diamond-like films show some of their best adhesion on car-
bide-forming substrate materials, and therefore are expected to be
quite suitable as wear-resistant films on (e.g., tungsten-based) ce-
mented carbides or high-alloy tool steels. Nevertheless, sufficient
adhesion can only be achieved on a limited number of substrates or
with the deposition of additional interlayers, the thermal mismatch
between diamond (ca. 1 ppm), and common cobalt-matrix cemented
carbides (4–6 ppm), is quite large. Adhesion of diamond like films is
quite dependent on proper nucleation (the importance of surface
preparation methods has been described above), but the mere pres-
ence of carbides is not a guarantee for diamond film adhesion. In the
case of cemented carbides, the cobalt content of the substrate pro-
motes graphitization of the initial film and prohibits the expected
diamond nucleation on the tungsten carbide phase, thus resulting in
poor adhesion. Ways to avoid that effect are: a reduction of the cobalt
matrix in cemented carbides or etching the cobalt from the substrate
surface prior to diamond film deposition, depositing interlayers, or
choosing altogether different substrates such as SiAlON or silicon
nitride ceramics.
An emerging new application for wear-resistant diamond films
will soon bring considerable improvements in the performance of
data storage media, increasing data density, and enhancing miniatur-
ization. Currently, contact storage media are being developed that
allow the magnetic read-write head to slide directly on the magnetic
disc, resulting in a higher data storage density, and reduced disk
speeds. In conventional disc drives, an air cushion separates the disc
from the sliding head to prevent wear, but that requires a stronger
field to write, or read the data (hence the higher disk speed to
generate an appropriate induction signal), and limits the resolution
for the reliable separation between two neighboring signal bits. In

contact media technology, both the head and the disc will be pro-
tected by a diamond film, and its physical properties (magnetic
permeability and good optical transmission) also make it compatible
for read/write laser disc data storage technology.
Films are usually very smooth when they are amorphous or
nanocrystalline, and can be grown to useful thicknesses of ca. 10 µm,
but develop significant internal compressive stress with increasing
film thickness, which can lead to delamination. It has been demon-
strated that intrinsic stress of polycrystalline diamond films of con-
stant thickness varies significantly with deposition parameters, where
high bias voltages and higher deposition temperatures reduce stress
down to the point of tensile stress. Of course, the application of low-
stress deposition conditions is usually limited by the temperature
tolerance of the substrate. From the viewpoint of deposition pro-
cesses, sputtered and mass-selection ion beam deposited films pos-
ses the highest stress levels.
Due to its relatively low thermal stability (regarding the work-
ing temperatures of up to 620°C of high-speed cutting tools, and
compared to boron nitride), synthetic diamond is not suitable for the
machining of ferrous metals. H-DLC films lose their hydrogen (and
consequently their desirable properties) at ca. 400°C. Most of its
applications can be found in the low to medium speed (but high-
feed) machining of aluminum or copper base alloys, as well as
ceramics. For these applications, the high thermal conductivity of
diamond related materials is regarded as another positive character-
istic, as it enhances the heat transfer from the work area into the
cutting tool, thus, avoiding overheating of both tool and workpiece.
The chemical stability of diamond-related coatings against
many aggressive acids, bases, and all known solvents is high, and the
material is considered biocompatible. This has led to several proposals
to incorporate these films in artificial implants or electrical and thermal
biosensors, but no significant applications have yet been reported.
The friction coefficient and wear behavior of DLC and H-DLC
are quite dependent upon the humidity of the ambient. The presence
of water vapor increases friction and wear in general, and especially

when a-C:H films are deposited at low plasma powers (thus, making
them more diamond-like). Diamond related films of very graphitic
or polymeric structure are less affected by changes in humidity, but
the difference between these films and very diamond-like structures
becomes quite apparent in dry air or nitrogen atmospheres. Similar
dependencies were observed when the influence of annealing and
deposition temperature were investigated, leading to the conclusion
that deposition and wear conditions (humidity, temperature) that
result in graphitization of the films significantly reduce their wear
resistance, shifting properties closer to the lubrication effect typical
for graphite.
Semiconductor Electronics. DLC, and H-DLC films can be
deposited on many important electronic substrates (e.g., silicon,
germanium) with good adhesion, and the lattice parameters of nickel,
copper, silicon carbide, beryllium oxide, and boron nitride allow
epitaxial growth of diamond. These materials have also been used as
buffer layers for the deposition of diamond-related films on sub-
strates of low adhesion (e.g., ZnS or ZnSe). Yet, the biggest problem
for the widespread use of diamond films in semiconductors is the
inability to grow defect-free heteroepitaxial films of high crystal
perfection, which are required for maximum thermal conductivity,
carrier mobility, electrical resistivity, and electrical insulation. An-
other limitation is the difficulty to deposit films with smooth sur-
faces, a requirement when additional films are to be deposited on top
of the diamond film. Nevertheless, some techniques have been
developed to modify diamond surfaces (ion etching, oxidation meth-
ods at temperatures above 700°C). Another positive aspect is the fact
that diamond surfaces can be drilled and cut with high-power lasers.
For semiconductor patterning of diamond films, nitrogen dioxide
etching is available, replacing the traditional oxygen etching
which has an extremely low etch rate for this particular material, but
is very successful in removing graphite or other non-diamond carbon
compounds.
Diamond-related films are wide band gap materials (diamond:

5.45 eV), and therefore of high resistivity, and for H-DLC films, this
parameter increases with increasing hydrogen content (from ca. 107
ohm·cm to 1016 ohm·cm). Both film types have a very high break-
down voltage, making them useful insulating materials for semicon-
ductor applications. The electrical conductivity of hydrogenated
amorphous carbon films occurs by the mechanism of electron hop-
ping between defects (as opposed to metallic type band conduction
via hole migration, which characterizes p-type semiconductors such
as doped diamond, see below), the conductivity increases with in-
creasing temperature and is independent of the bandgap size. Dia-
mond Schottky diodes have been fabricated that operate reliably
above 700°C, as compared to the limit of 150°C for the equivalent
device made with the conventional silicon techniques.
As the thermal conductivity (more than 650 W/m·K) of these
materials is among the highest of all known materials, DLC and H-
DLC are ideal heat-sink materials. Together with the high resistivity,
this has led to the laboratory-scale development of several electronic
high-temperature and high-power density devices. The bandgap of
perfect diamond is 5.45 eV, whereas Si is only 1.3 eV, thus, indicat-
ing the potential for high power diamond devices. In this context,
beta silicon carbide single crystal films of high quality have been
produced, and are diamond’s potential competition for high power
devices. For large scale production, reproducibility (among other
technological problems), of the diamond film quality and perfor-
mance is not yet sufficient. A further limitation for applications
subject to cyclic temperature variations is the large difference in
thermal expansion coefficients of diamond (1–1.5 ppm), and silicon
(2.6 ppm), or GaAs (5.9 ppm), creating difficulties in conventional
packaging techniques. Nevertheless, some novel design concepts
have been proposed to facilitate the superior thermal dissipation
provided by diamond films by incorporating thermal “greases” that
would allow for differential thermal expansion between the diamond
layer and the signal and power plane structures.
Diamond related films have been successfully doped with

boron to form p-type semiconductors. Dopant donor gases may be
diborane (up to 1% dilution in gaseous hydrocarbons), resulting in a
tenfold increase in conductivity. The dopants substitute carbon sites
in the amorphous diamond like structure, but the new structure
creates only a weak shift of the Fermi level, making it a rather
ineffective semiconductor that has not yet found application in
devices. Whereas p-type doping is widely used and has led to the
development of photodiodes and other semiconducting devices in
several laboratories, n-type doping has not yet been successfully
achieved. Based on calculations, sodium and lithium are the most
promising dopants; phosphorus is also possible, but creates a bandgap
too small for practical applications. Most efforts focus on sodium,
but stabilizing the necessary dopant level, and element distribution
while maintaining a high-purity diamond structure (without graphite
formation) is difficult. A major difference between the doping strat-
egies of boron (p-type) and sodium (n-type), is that boron diffuses
easily in the diamond structure and can be directly applied e.g., by
ion beam implantation after diamond film deposition, whereas so-
dium has to be supplied by complex precursors that have to be
broken down in a plasma, and the sodium implanted during diamond
deposition since there is no mobility once the sodium is embedded in
the film structure. A diamond transistor has been successfully fabri-
cated, but the p-n-characteristic was achieved not by creating a
“true” n-type layer, but by varying the dopant level of boron in the
three respective layers (overdoping of emitter and collector, light
doping of the base).
Another proposed application of diamond films is based on
their negative electron affinity. Reviving the principle of cathode ray
tubes, a cold diamond cathode operated at a fifth of the current
density required by a conventional hot cathode device, and thus, has
the potential to provide “cooler” or more powerful cathode ray tubes
(CRTs) or vacuum tube amplifiers.
Optics. Hydrogen free DLC films are transparent to visible
light and slight hydrogen impurities lead to a yellow discoloration in
thicker films. Transmission is also high for infrared (IR) frequencies,

and for H-DLC films, C-H bond vibrations limit this good IR
transparency only slightly. The rather poor adhesion of diamond-
related films on some optical materials (e.g., ZnS, ZnSe) can be
improved with thin buffer layers of germanium without significantly
impairing the optical properties.
The refractive index of diamond-related films (1.6 to 2.4), is
intermediate between that of air and germanium, and diamond coat-
ings have been used to reduce reflective losses in infrared optics, and
solar cells. Low reflectance also can be achieved on substrates with
refractive indexes between 2.5 and 5.8, such as ZnS, TiO2, Se, and
As2S3 glasses, but not uncommon Si-based glass with refractive
indexes near 1.5. The lower hydrogen content of the diamond related
films goes along with a higher index of refraction.
Early applications of diamond related films were in the field of
wear protection for optical lenses, drawing from the superior proper-
ties such as hardness, smoothness, good optical transmission, and
low friction coefficient. However, it is important for these applica-
tions to be successful that the substrate (glass or plastic), provides
sufficient support for the film. Recent applications of these films in
laser windows aim at the mechanical and especially chemical protec-
tion at minimal transmission losses required by some laser environ-
ments. This also includes the interface between the lens and the
interior of the laser to protect the laser’s plasma gas from contamina-
tions from disintegrating lenses. In aerospace research, diamond-
related films have been tested as heat, and wear protective coatings
for canopy windows and radar nose cones (particularly for the
protection of the underlying Germanium optics), but no large scale
applications have emerged at this time.
Other proposed applications for diamond related coatings are
wear protective coatings for optical storage media (laser discs), also
providing long time corrosion protection for the aluminum layer, as
well as the coating of optical sensors immersed in aggressive fluids
and gases. A similar application is that of a diamond film to provide
a hermetic seal for fiber optics devices and is expected to perform
better than the polymer coatings currently used. Investigations have

determined that diamond related films provide very effective protec-
tion against moisture penetration.
Laser technology may also benefit from developments in syn-
thetically grown diamond crystals, as the thermal conductivity in
high power pulsed lasers is a limiting design factor. Heat up during
laser operation causes the crystal to expand, thus, pushing it out of
phase with the designated wavelength, reducing efficiency, and the
cyclic wear of pulsed lasers may eventually destroy the crystal. The
low thermal expansion coefficient, good thermal stability, superior
thermal conductivity, and strength provide excellent conditions for a
diamond laser, but diamond synthesis at this time falls short in
respect to crystal size, purity, and perfection (all of which are
required to prevent graphitization), and such laser applications will
not be available in the near future.
3.4 Emerging Materials
Considerable attention has recently been given to a new syn-

thetic material that may be even harder than diamond–C3N4. In 1988,
an empirical scaling of the relationship between hardness and bulk
modulus led to the suggestion by A. Y. Liu and M. Cohen[1] at
Berkeley, that a material with a larger bulk modulus than that of
diamond (4.43 mbar), might display higher hardness. This presum-
ably can be achieved when substituting silicon with carbon in the
beta-Si3N4 structure, resulting in beta-C3N4 with a calculated bulk
modulus of 4.83 mbar. According to these predictions, the com-
pound is hexagonal (thus possibly displaying anisotropic elastic
properties), with a unit cell consisting of two formula units. The
atomic coordination suggests sp3 hybrids on the C atoms and sp2
hybrids on the N atoms; cohesive energy calculations indicate the
material will be at least metastable, which increases the possibility of
synthesizing it in thin film processing.
Since this initial work, further theoretical studies focused on

the bond structure, but attempts to synthesize this material were not
successful for several years. Recently though, several teams reported
the deposition of thin carbon nitride films indicating possible pres-
ence of C3N4, and initial efforts have been made to verify the
existence, and examine the structural properties of the new material.
The reportedly first carbon nitride film that exceeded the hard-
ness of diamond was deposited by Fujimoto and Ogata[2] with a
process called ion assisted dynamic mixing (IVD). This technique
evaporates carbon in an electron beam evaporator in the presence of
an ionized nitrogen gas atmosphere. Based on ion current measure-
ments, the carbon/nitrogen ratio was examined over a wide range by
varying the carbon evaporation rate and films were deposited on
tungsten carbide and silicon substrates. The substrates were cooled
to a deposition temperature of less than 100°C.
The films produced by Fujimoto and Ogata were amorphous
(whereas Liu and Cohen predicted a crystalline material), and were
never clearly identified as beta-C3N4, but the maximum hardness of
6500 HV (Knoop) was achieved at ratios of C/N = 1 and high-ion
energy deposition conditions. XPS analysis revealed a higher peak
maximum (286.3 eV) than that of diamond (285.8 eV), suggesting
the presence of a new (though unspecified) material. For the samples
of a composition that displayed maximum hardness, this “new”
material is dominant, with traces of diamond, i-carbon, and/or car-
bon also present.
Niu, Lu, and Lieber[3] at Harvard University deposited solid
carbon nitride films by pulsed laser ablation of a graphite target in
the presence of atomic nitrogen at low temperatures. The approach of
using atomic nitrogen (which was obtained as a beam emitted from a
nozzle that processes the gas in a helium-assisted RF discharge) was
chosen due to thermodynamical considerations of the predicted
metastable nature of the material. Films were deposited on Si(100)
and Ni substrates, and the nitrogen content was varied over a range
that had a possibility to provide the desired stoichiometry. Adhesion
on both substrate materials was good and the films were thermally
stable, maintaining their composition up to 800°C. XPS investiga-
tions revealed binding energies of 284.6 (for C), and 399.1 eV (for
N). TEM showed small crystallites in a mostly poorly crystalline
film, and x-ray diffraction patterns are consistent with calculated
data for the theoretical structure of beta-C3N4; nevertheless, the
authors state that further investigations are required to substantiate a
claim of actually having synthesized the desired compound.
Important initial investigations in the field of tribological prop-
erties of C-N thin films have been conducted by Y. W. Chung, W. D.
Sproul, and coworkers at Northwestern University.[4] They prepared
films by DC high-rate magnetron sputtering of a pyrolytic graphite
target in nitrogen, and nitrogen/argon gas mixtures on zirconium,
NaCl, and glass substrates. The films were of high quality, with very
low contamination, absence of pinholes, and clear IR absorption
spectra, and consisted of both crystalline and amorphous regions.
The content x of nitrogen in these CNx-films could be varied in wide
margins depending on processing parameters, and IR absorption
spectra indicate the presence of a carbon-nitrogen triple bond, even
though the films were never referred to as C3N4 by the authors.
Although the films were quite thin for tribological applications (ca. 1
mm), they performed excellently in first tribological tests. A friction
coefficient of 0.16 under unlubricated conditions was determined,
which is close to that of diamond (ca. 0.1), and remained unchanged
during the wear test. These first results are first evidence of the high
potential of C-N films for tribological applications.
Several other attempts to deposit beta-C3N4 include plasma
decomposition of methane (CH4), and nitrogen or pyrolytic decom-
position of C-N-H organic precursors; they produced amorphous
carbo-nitro-hydrides and did not indicate the presence of a material
that could likely be the desired carbon nitride. These results lead to
the assumption that the presence of hydrogen inhibits the formation
of this structure, and clearly indicate that the reaction kinetics are
poorly understood and/or difficult to control. Nevertheless, this
should not lead to the conclusion that there will be no valuable

contributions these C-N-H films can make. Quite in comparison to
diamond and diamond like materials, amorphous and hydrogenated
carbon nitride films were found to have some diamond like proper-
ties, and may thus, have considerable potential for various technical
applications, especially when their production is easily achieved,
reliable, economical, or otherwise attractive.
The range of possible applications of carbon nitride is difficult
to fathom; aside from the superior hardness which lends itself to
many tribological applications, most physical properties of the mate-
rial are yet unknown. Once there is clear evidence for the existence
of the material and a reliable process has been established to synthe-
size the necessary amounts for further investigations, the electronic,
thermal, optical, and further properties of the material will reveal the
technological potential of C3N4.
4.0 CONCLUDING REMARKS
This contribution presents a broad overview over the current

topics in thin film research on boron nitride, and diamond related
materials. The focus on matters of film deposition processing, and
applications was not meant to underestimate the tremendous efforts
and contributions of the theoretical work in this field, and we encour-
age the interested reader to learn more about the impressive achieve-
ments in thermodynamical calculations, film growth models, theo-
retical aspects of solid state physics and process simulations. With-
out them, we would lack the guidelines that spur both deposition
development, and application potential.
Only a few of the applications for c-BN and diamond related
materials that have been discussed in this article have reached the
stage of commercial applications. At present, by far the biggest
market for these materials are wear related applications (especially
fabrication and machining), and the introduction to optic and data
storage technologies (wear protection of lenses, sensors, and record-

ing heads) is currently taking place. However, considering the tre-
mendous efforts that are being made to utilize the electronic proper-
ties of these materials in research labs worldwide, a tremendous
impact on the semiconductor sector can be expected a few years from
today, setting new standards for high power, and high temperature
devices. The prerogative for that is finding solutions to the most
pressing problems (such as structural composition, crystal and film
quality, epitaxial growth, control of semiconducting properties, con-
tamination, and purity), which then have to be incorporated in
classical concepts of semiconducting device fabrication to make
them available in larger quantities, and at competitive prices. With
that, another glamorous sparkle will be added to the noblesse of these
outstanding materials.
REFERENCES
1. Vel, L., Demazeau, G., and Etourneau, J., J. Materials Science and
Engineering B, 10:149 (1991)
2. Bhushan, B., (ed.), Handbook of Tribology-Materials, Coatings,
and Surface Treatments, McGraw-Hill, New York (1991)
3. Caveney, J. Materials Science and Engineering B, 11:197–199
(1992)
4. Pouch, J. J., Alterovitz, S. A., (ed), in: Synthesis and Properties of
Boron Nitride, Materials Science Forum 54 & 55, Trans Tech
Publications, Brookfield, VT (1990)
5. Shanfield, S., Wolfson, R., J. Vacuum Science, and Technology A,
1:323 (1983)
6. Halverson, W. D., Tetreault, T. G., and Hirvonen , J. K., in: Synthesis
and Properties of Boron Nitride, Materials Science Forum 54 & 55,
Trans Tech Publications, Brookfield, VT (1990)
7. Bhushan, B. in: Handbook of Tribology, McGraw-Hill, New York
(1991)
8. Aita, C. R., in: Synthesis and Properties of Boron Nitride, Materials

Science Forum 54 & 55, Trans Tech Publications, Brookfield, VT
(1990)
9. Mieno, M., and Yoshida, T., Jpn. J. Applied Physics, 27:L1175–L1177
(1990)
10. Watanabe, S., Miyake, S., and Murakawa, M., Surface and Coatings
Technology, 49:406–410 (1991)
11. Ikeda, T., Kawate, Y., and Hirai, Y., J. Vacuum Science & Technology
A, 8:3168–3174 (1990)
12. Doll, G. L., Sell, J. A., Taylor, C. A., and Clarke, R., Physical
Review B, 43:6816–6819 (1991)
13. Dana, S. S., in: Synthesis, and Properties of Boron Nitride, Materials
(1990)
14. Osaka, Y., Chayahara, A., Yokoyoma, H., et al., in: Synthesis and
Properties of Boron Nitride, Materials Science Forum 54 & 55,
Trans Tech Publications, Brookfield, VT (1990)
15. Sugyiama, K., and Itoh, H., in: Synthesis and Properties of Boron
Nitride, Materials Science Forum 54 & 55, Trans Tech Publications,
Brookfield, VT (1990)
16. First European Conference on Diamond and Diamond-like Carbon
Coatings, Crans-Montana, (Sep. 17–19, 1990)
17. Watanabe, S., Miyake, S., and Murakawa, M., Surface and Coatings
Technology, 49:406–410 (1991)
18. Murakawa, M., and Watanabe, S., in: Application of Diamond Films
and Related Materials, (Y. Tzeng, M. Yoshikawa, M. Murakawa,
A. Feldman eds.), pp. 661–668 Elsevier (1991)
19. Mishima, O., in: Synthesis and Properties of Boron Nitride, Materials
(1990)
20. Miyake, S., Watanabe, S., and Murakawa, M., in: Application of
Diamond Films and Related Materials, (Y. Tzeng, M. Yoshikawa,
M. Murakawa, A. Feldman, eds.), pp. 669–674, Elsevier (1991)
21. Karnezos, M., in: Synthesis and Properties of Boron Nitride,
Materials Science Forum 54 & 55, Trans Tech Publications,
Brookfield, VT (1990)
22. Liu, A. Y., Cohen , M. L., Science, 245:841 (1989)

23. Fujimoto, F., Ogata, K., Jpn. J. Applied Physics, 32:L420–L423
(1993)
24. Niu, C., Lu, Y. Z., Lieber, C. M., Science, 216:334–337 (1993)
25. Chen, M. Y., Lin, X., Dravid, V. P., Chung, Y. W., Wong, M. S.,
Sproul, W. D., Surface & Coatings Technology, 54/55:360–364
(1992)
ADDITIONAL READING
Synthesis and Properties of Boron Nitride, (J. J. Pouch, S. A.

Alterovitz, eds.), Trans Tech Publications, Brookfield, VT (1990)
Krar, S. F., and Ratterman, E., Superabrasives: Grinding and
Machining with CBN and Diamond, Gregg Division, McGraw-Hill,
New York (1990)
Vel, L., Demazeau, G., and Etourneau, J., J. Materials Science and
Engineering B, 10:149 (1991)
Paine, R. T., Narula, C. K., Chem. Rev., 90:73 (1990)
Synthetic Diamond: Emerging CVD Science and Technology, (K. E.
Spear, J. P. Dismurkes, eds.), The Electrochemical Society Series,
John Wiley & Sons, New York (1993)
Prins, J. F., Ion-Implanted Structures and Doped Layers in: Diamond
Materials Science Reports, 7:271–364 (1992)
Diamond Films and Coatings, (R. F. Davis, ed.), Noyes Publications,
Park Ridge, NJ (1993)
Dresselhaus, M. S., and Kalish, R., Ion Implantation in Diamond,
Graphite, and Related Materials, Springer Verlag Berlin, New
York (1992)
Proceedings of the First International Conference on the Application
of Diamond Films, and Related Materials, Auburn, Alabama (1991),
Elsevier, New York (1991)
Proceedings of the First International Symposium on Diamond, and
Diamond-Like Films, Los Angeles, CA (1989), (J. P. Dismurkes, et
al., eds.), Electrochemical Society, Pennington, NJ (1989)
Proceedings of the Second International Symposium on Diamond

Materials, Washington, DC (1991), (A. J. Purdes, et al., eds.),
Electrochemical Society, Pennington, NJ (1991)
NATO Advanced Study Institute on Diamond and Diamond-Like
Films and Coatings, Castelvecchio Pascoli, Italy (1990), (R. E.
Clausing, et al., eds.), Plenum Press, New York (1991)
Proceedings of the Second International Conference on New
Diamond Science, and Technology, Washington, DC (1990), (R.
Messier, et al., eds.), Materials Research Society , Pittsburgh, PA
(1991)
Journal of Diamond Films and Technology, Myu, K. K. (publisher),
Tokyo, Japan
Journal of Diamond and Related Materials, (R. Messier, ed.),
Elsevier, The Netherlands
On Carbon Nitride Materials
Liu, A. Y., Cohen, M. L., “Prediction of New Low Compressibility

Solids,” Science, 245:841 (1989)
Liu, A. Y., Cohen, M. L., Physical Review, B, 41(15):10727–10733
(1990)
Fujimoto, F., Ogata, K., Jpn. Journal of Applied Physics, 32:L420–
L423 (1993)
Niu, C., Lu, Y. Z., Lieber, C. M., Science, 261:334–337 (1993)
Summary, Developments, and Outlook 517
10
Summary, Developments,
and Outlook
Rointan F. Bunshah
This chapter is divided into nine sections, summarizing devel-

opments occurring during the last few years and the future outlook.
1.0 CURRENT INDUSTRIAL PVD PROCESSES
There are four industrial reactive PVD processes used to de-

posit hard coatings. They have been described in detail in Ch. 2.
They are
1. Low voltage electron beam evaporation (Balzers
process, or Ulvac process).
2. Triode high voltage thermionic beam (Tec Vac
process).
In both of these, the electron beam evaporates the metal which
reacts with the reactive gas plasma to deposit a compound. Ion
bombardment of the substrate is obtained by putting a negative bias
on the substrate. These processes can deposit coatings such as the
mononitrides of Ti, Zr, and other species, the carbides such as TiC
and carbonitrides such as TiCN. It is not possible to use these
processes to deposit alloy nitrides such as TiAlN2.
517
However, a recent development in the high voltage EB process

has overcome this deficiency. Garside[1] reported on a system con-
taining two evaporant sources, for example, Ti, and Al.
By controlling the EB impingement parameters on each source,
he is able to deposit any composition of Ti Al nitride, which might be
a major advantage over arc evaporation, and magnetron sputtering
methods.
3. Cathodic arc evaporation process (random, or steered
arc) as practiced by Multi-Arc Scientific Coatings,
Hauzer Vac Tec.
One of the main advantages is the ability to deposit alloy
nitrides, and other complex compounds. This process produces a
very thin zone, 200–300 Å in thickness on the substrate surface
where the coating and substrate materials are intermixed. No other
process produces such a mixed zone which is thought to result in
better tool performance in punching and stamping applications. The
main drawback of the process is the production of liquid droplets or
macroparticles which deposit along with the coating. More advanced
techniques such as the steered arc, or filtered arc which reduce, and
in the case of filtered arc, eliminate the macroparticles but at the
expense of reduced deposition rate.
4. Unbalanced magnetron process, as practiced by
Teer Coatings Ltd., and Hauzer Vac Tec. This is
perhaps the most versatile process from the
perspective of a number of materials that can be
sequentially deposited without breaking vacuum.
Systems with 2, 4, 6, or 8 rectangular cathodes
spaced around the chamber. Another virtue is that
the cathodes can be used for arc evaporation, or
magnetron sputtering, or for ion bombardment
cleaning. The multiple cathodes also permit
deposition of layered superlattice hard coatings such
as TiAlN/TiNbN.
T. Harkmann, et al.,[2] describe a large volume PVD coating

system using advanced controlled arc, and combined arc/unbalanced
magnetron deposition. In this machine, the cathodes are rectangular
and 1700 mm long. They are arranged in a closed field array.
Different cathode layouts are possible depending on the application.
For example:
1. Four advanced controlled arc (ACA), and two

unbalanced magnetron (UBM) cathodes could be
set up for a decorative application of a hard coating
such as TiN or ZrN with a gold flash as the top coat.
2. With two ACA cathodes, and four UMB cathodes,
a metal ion etch to clean the substrate surface can be
performed using the two Ti arc cathodes. This could
be followed by UMB magnetron sputtering from
the targets. For example TiAlCN can be deposited
using Ti and Al targets, and reactive gases containing
carbon and nitrogen.
3. Using the cathodes alternately, superlattice layered
coatings can be produced.
2.0 CVD AND PACVD PROCESSES
Thermal CVD is predominantly used to deposit hard coatings

of carbides, nitrides, oxides, and multilayer coatings onto cemented
carbide tools. Very often the coating consists of two or three layers,
e.g., TiC/aluminum oxide/TiN. The exact sequence varies with the
tool, and the application. All of these layers require a high deposition
temperature to carry out the reaction, using TiCl4 as the source of Ti,
for example. For corrosion applications, the presence of Cl ion at the
coating substrate interface deteriorates the performance of the coating.
For engineering parts of complex shapes, it is very expensive to

fabricate them out of cemented carbides. High speed steel is a
preferred substrate material, since it can withstand temperatures
up to 550°C before tempering (softening). The deposition tem-
perature therefore becomes an important parameter for such applica-
tions. PACVD is used since TiN, and other hard coatings can be
deposited at a low temperature, 400 to 600°C, as compared to
CVD processes, where the deposition temperatures are 1000°C, and
above. The Cl content at the interface is sufficiently high at
400°C to cause flaking. Above 500°C, the Cl content decreases
leading to an improvement in properties.[3]
3.0 TRIBOLOGICAL COATINGS AND THEIR

APPLICATIONS
The evolution and variety of coatings for tribological ap-

plications is shown in Table 1, courtesy of Dennis Teer. The early
days, 1968 onwards for CVD coatings, and 1972 onwards for PVD
coatings, featured simple compounds such as TiN or TiC. A 2–3
micrometer thick coating on cemented carbide tips produced a 100%
improvement in tool life. In 1973, a CVD multilayer coating TiC-
Ti(CN)-TiN was marketed, which produced remarkable improve-
ments over the simple coatings. Later developments incorporat-
ing the incorporation of alumina in multilayer films has enabled
optimum lifetimes to be attained.
Subsequent major developments occurred with the advent, and
maturing of PVD technologies. As shown in Table 1, complex PVD
coatings produced dramatic improvements in wear resistance of
tools, and engineering surfaces. A more remarkable concept was the
use of soft, very low friction coatings such as molybdenum disulfide
on high speed steel drills and end-mills (developed by Rechberger,
Brunner, and Dubach).[4] These coatings produced good machining
results with platinum alloys, titanium and titanium alloys, wrought
Table 1. Coatings for High Performance Machining
Coating Bulk Friction Adhesion Uses/Application

Vickers Coeff. Critical Load
Hardness Comparative Minimum
TiN ~2200 ~0.6 >55N A general coating for various tools and
dies. The most widely used and well
known coating.
Low temp.
TiN ~2200 ~0.6 >55N Useful for coating temperature sensitive
tools, e.g. brazed carbides, D2 01 dies,
bearing steels, etc. Coating temperature
200–300°C.
TiAlN ~3500 >50N For higher speed machining, particularly
carbide tooling where operational tem-
peratures are increased. Known to work
much better than TiN on machining
tough steels, e.g. D2 and cast irons.
TiZrN ~3500 <0.6 >50N Similar to TiAlN, hardness of up to
4000Hv can be achieved.
ZrN ~3300 <0.6 >45N Very good for the machining of alu-
minium alloys, always significant im-
provement over TiN.
DLC ~2500 <0.15 >40N A hard low friction coating excellent in
machining Al, copper, and brass. Also
outperforms TiN in drilling stainless
steel. Very good performance in any ap-
plication requiring low friction, e.g.
extrustion dies, engine wear parts (cam
shafts,rocker arms, valve guides, etc.),
fishing rod guides. A decorative black
coating.
MoS 2 very low <0.1 Excellent performance machining alu-
minium alloys. Also used as a dry, self-
lubricating film for aerospace bearings,
mechanisms operating under vacuum
and worm gears, etc.
TiCN ~2800 0.6 >50N Tough coating suitable for punches. Bet-
ter than TiN in some applications where
friction is important.
TaN ~2400 0.6 Stable at much higher temperatures than
TiN.
CRM ~2400 0.6 >70N Tough, for copper m/c.
aluminum alloys, cast aluminum, and electrolytic copper, the main

improvements being in the workpiece surface finish, productivity,
and tool life. In the initial work, titanium was used to produce a
gettering effect in the chamber to reduce the water vapor content
(since these coatings are susceptible to degradation in the presence of
water vapor). Titanium was also used as an interlayer between the
substrate, and the coating which led to an improvement in coating
adhesion. A natural progression of this work was the incorporation
of titanium in the coating itself, resulting in improved friction and
wear properties, i.e., increased load capacity, greater reproducibility,
and less susceptibility to humidity. Promising future applications
will be in the protection of many aerospace, automobile, and general
wear components.
4.0 ALUMINA COATINGS BY CVD, PACVD AND PVD

PROCESSES AND FIRST MACHINING STUDIES
Alumina is a very interesting crystalline material exhibiting

many crystalline phases. The alpha, kappa, and xeta phases are
called the “alpha series” with hcp stacking of oxygen atoms; whereas,
the gamma, theta, eta, and delta phases form the “gamma series”
with fcc stacking of oxygen atoms.
Multilayer CVD coatings have been in production since the
70s. The enhanced thermodynamic stability of the oxide, as com-
pared to the nitride, or carbide coatings is the driving force. CVD
alumina coatings that contain the hard crystalline alpha phase and the
metastable orthorhombic kappa phase are deposited at temperatures
above 1000°C.[5] Because of the high substrate temperature required
for the CVD process, only a few materials such as cemented carbides
can be used as substrates.
This provided the impetus for the development of processes to
deposit crystalline alumina at lower substrate temperatures. This can
be achieved both by PACVD and PVD processes.
Ling, Wan and Hon[6] studied the formation of alumina films in

an RF Plasma Enhanced CVD process. The temperature range of
deposition was from 300 to 500°C. Aluminum oxide deposition was
achieved in a AlCl3/CO2/H2 gas system by the hydrolysis of AlCl 3.
XRD patterns show that the films were essentially amorphous. The
film composition was near stoichiometric. There was considerable
Cl content in the films deposited at temperatures below 500°C. It is a
pity that the temperature range in these experiments was not ex-
tended above 500°C to attempt to produce crystalline alumina films.
PVD alumina deposition was studied by several authors. Kelly,
et al.,[7] used reactive unbalanced Magnetron sputtering in the pulsed
mode. The coatings deposited without intentional substrate heating
were fully dense amorphous alumina with microhardness values
between 1940 and 2480 kg/mm-2.
Zywitski and Hoetzsch[8] used dual magnetron pulsed reactive
magnetron sputtering to deposit hard crystalline alumina layers on
steel substrates over a temperature range of 290–770°C, at various
sputtering powers. The results of their investigation are given in
Table 2. At the highest sputtering power, textured gamma alumina
starts depositing around 350°C. At lower sputtering powers, the
deposit transforms from amorphous alumina, to gamma alumina at
450–480°C. The formation of Alpha alumina starts at 670°C and
practically pure alpha alumina deposits at 750–770°C. The substrate
bias has a substantial influence on ratio of the phase fractions of
gamma, and alpha alumina. Linked to the formation of the crystalline
phases is the increase in hardness from 10 to 22 GPa. The same
authors,[8] in a subsequent investigation, studied the structure with
XRD, and cross-section TEM. They observed that the phase trans-
formation from amorphous, to alpha alumina occurred via gamma
alumina. Also, an enhancement of the plasma density shifts the
formation of the crystalline phases to lower temperatures.
Table 2. XRD results on alumina phase as a function of deposition

temperature and sputtering power by PMS sputtering. (Zwitzki and
Hoetsch.)
Mean Sputtering Power

TS
/°C 11 kW 13 kW 16 kW 17 kW
290–350 amorphous amorphous amorphous γ-Al2O3

(+50 V bias) with (110)-
fibre texture
450–480 amorphous –––– γ-Al2O3 ––––

(+50 V bias) (+50 V bias)
550–560 γ-Al2 O3 with (110) - fibre texture ––––

(+50 V bias)
670–690 10% α-Al 2O3 45% α-Al2 O3 10% α-Al 2O3 ––––
90% γ-Al 2O3 55% γ-Al2 O3 90% γ-Al 2O3
(+50 V bias) (+50 V bias)
750–770 α-Al2O3 60% α-Al2 O3 α-Al2O3 ––––

40% α-Al2 O3
(+50 V bias)
Schneider, et. al.,[9] studied very high rate sputtering of alu-

mina hard coatings using pulsed reactive magnetron sputtering at
deposition temperatures less than 250°C. The coatings were amor-
phous, and near stoichiometric with a hardness of 12 GPa, and an
elastic modulus of 140 GPa. Subsequently, the same group studied
alumina coatings deposited by ionized reactive magnetron sputter-
ing, using an RF coil to excite the plasma in the space between the
source and, substrate. XRD results revealed that the structure was
amorphous at 300°C and at 450°C deposition temperatures. The
influence of ion bombardment at 472°C was to transform the
amorphous structure to the kappa phase. A mixture of the kappa, and
theta phases at higher ion current density was observed at 430°C.
The first machining studies of pulsed plasma deposited alu-

mina by PACVD and pulsed bipolar dual magnetron reactive sput-
tering (DMS), were reported by Sjostrand and Soderberg.[10] The
Pulsed plasma PACVD technique produced hard crystalline coatings
(20 GPa), with a dense, smooth finish without any thermal cracks.
Preliminary cutting tests of alumina PACVD coated carbide inserts
in the turning of a regular steel showed that the intrinsic wear
resistance of the coating was comparable to conventional CVD
alumina.
With the pulsed DMS method, alumina, as well as coatings
such as TiAlN could be deposited. Crystalline alpha, as well as very
fine grained crystalline gamma phase were deposited on cutting tools
in the temperature range of 500–800°C. The coatings are dense, and
smooth with a hardness around 20 GPa. The results of preliminary
cutting tests show that the PVD alpha/gamma alumina coated inserts
in the turning of a regular steel show superior cutting performance of
the gamma alumina coating as compared to PVD TiN(TiCN) coated
inserts. Furthermore, the gamma alumina coated inserts show an
intrinsic wear resistance close to that of CVD alpha, or kappa
alumina coatings.
5.0 ENVIRONMENTAL CONCERNS IN MACHINING

AND FORMING
Stricter pollution control requirements demand “green” manu-

facturing processes. Many machining processes use cutting fluids to
cool the workpiece, the tool, and the machine during cutting. The
coolant also reduces friction, and the chips created during the opera-
tion are removed from the cutting area. Coolants, however, create
high costs in production, and can be hazardous to the environ-
ment, and to human health. For example, the cost of cutting fluid
technology (amortization, equipment, staff, service, energy, dis-
posal, etc.) represents about 17% of the costs of machining an engine
cylinder head, and a cam shaft.[11] The number of laws regulating
waste oil in Germany have increased from 2 in 1950, to more than 30

currently. The disposal costs for cutting fluids in Germany have to be
paid as tax per ton. They have increased from 100 DM per ton in
1992, to 200 DM per ton in 1993, and are expected to increase
further. Lahres and Jorgenson studied the dry machining perfor-
mance in turning, and milling of diamond coated tools on an AlSi10Mg
hypereutectic casting alloy commonly used for automobile compo-
nents. Their results showed that optimized diamond coated ce-
mented carbide tools have great potential in dry turning and milling
of the aluminum alloys, as compared to PCD (polycrystalline dia-
mond), and K10 (uncoated cemented carbide tools). The results
indicated a consistent, qualitative correlation between the properties
of the diamond films, and their cutting performance. In dry turning,
with a continuous tool load, the morphological properties (e.g., a
well defined sp3/sp2 ratio, and good film homogeneity) are the
critical parameters. On the other hand, in dry milling with an impact
load, the critical property is the adhesion of the diamond film to the
substrate. Vetter, et al.,[12] studied hard coatings for lubrication
reduction in metal forming. This article contains an excellent discus-
sion of the various lubricants used in metal forming operations, their
advantages, and limitations. PVD hard coatings are effective not
only in wear reduction, but also in better workpiece surface finish,
higher productivity, and reduced use of lubricants. TiN, TiNC,
AlTiN, CrN, and other coatings have been used to enhance the
performance, and life of cold forming tools in processes like stamp-
ing, bending, rolling, extrusion, embossing, etc. This also reduces
the consumption of lubricants. Vetter, et al., found that CrN coatings
deposited by the cathodic arc process to be the most promising. They
report that neat stamping oils may be substituted by water diluted
emulsions when working with coated forming tools.
Four main effects were observed: (i) reduction of lubricants up
to a factor of 85% of the additive concentration; (ii) increase of tool
life to a factor of 169 ( first polishing); (iii) increase of working speed
up to a factor of 2; and, (iv) reduction of washing processes. These

effects lower the production costs, and decrease the environmental
charge.
6.0 DECORATIVE HARD COATINGS
Cost effective value addition to base metals for decorative, and

ornamental applications is a subject of major economic importance.
Electrochemical methods have been used for a long time for various
decorative coatings. The advent of PVD TiN coatings with the
characteristic “gold like” color in the early seventies opened up a
new technology for hard decorative coatings. Today, decorative
coatings are deposited by various PVD methods, including electron
beam evaporation, cathodic arc evaporation, or magnetron sputter-
ing, or in some machines by a combination of cathodic arc, and
magnetron sputtering. The colors available for decorative hard coat-
ings are shown in Table 3. They range from metallic tones of CrN, to
various bright yellow and gold tones of TiN, TiCN, and TiZrN
alloys, as well as other colors such as yellow brown, grey violet,
neutral blue grey based on the TiAl alloys, light grey to dark grey for
TiC, and deep black for DLC coatings, or DLC with incorporation of
metals such as Ti.
Since the hard gold colored nitride coatings do not have the
exact sheen or color of pure gold, particularly for costume jewelry
applications, they are overcoated with a thin layer of pure gold. This
is very conveniently done in multicathode machines where some of
the cathodes are made from gold. Other decorative applications
include the metallic greys and blacks. Another application which
combines the abrasion resistance, corrosion resistance, and a brass
color are ZrN coatings on zinc castings for bathroom fixtures and
door hardware.
Table 3. Decorative Hard Coating Colors (Adapted from Kopacz

and Schulz--Proc. Soc. of Vacuum Coaters 34th Annual Technical
Conference 1991)
Coatings Colors
Gold gold
CrNx metallic gray
TiNx light yellow-gold to brownish yellow
TiCxNy gold color to reddish brown
TiZrNx gold colors
Gold alloys gold colors on identically colored TiN xCy coatings
Decodoat 031-03 brownish yellow/violet grey/bluish grey
(based on TiAl)
Decodoat 036-038 reddish brown/copper colored

(based on TiAl)
TiCx light grey to dark grey
i-C (DLC) metallic black
7.0 SUPERHARD COATINGS FOR CUTTING TOOLS
Superhard materials are significantly harder than other refrac-

tory compounds such as the nitrides, carbides, etc. A Knoop hard-
ness of 40 GPa or greater characterizes superhard materials. Super-
hard materials are diamond, cubic boron nitride (cBN), boron car-
bide with high hot hardness, boron suboxides which are boron rich
phases, or boron oxide, polycrystalline superlattice coatings with
alternating nano-layers of compatible materials such as TiN/NbN,
and the elusive beta-carbon nitride C3N4 phase.
Inspector, et al.,[13] have reviewed this subject in a publication
entitled Advances and Technology Issues in Superhard Coatings for
Cutting Tools, published in The Proceedings of the Conference on

Wear and Superhard Materials, March 2–4, 1998, sponsored by
Gorham Advanced Materials Inc. Polycrystalline diamond (PCD),
and polycrystalline cubic boron nitride (PCBN), were the only su-
perhard tools on the market until the appearance of high quality
diamond films in 1981/82 prepared by low pressure CVD, and
plasma CVD techniques. The cutting edge of a superhard tool is a
superhard tip or blank cut from a disc and bonded onto cemented
carbide inserts. CVD diamond cutting tools are either diamond sheet
(thick film) tools or thin film diamond coated cemented carbide
inserts. The diamond sheet tools are fabricated much like the PCD
tools.
Diamond coatings can also be deposited directly onto ce-
mented carbide inserts. They generally adhere poorly to the substrate
due to mismatch of thermal expansion coefficients causing large
thermal stresses at the diamond-carbide interface, as well as the
interaction between carbon in the coating, and the cobalt layer on the
surface of the tool. Several approaches have been reported to limit
this carbon cobalt interaction.[13]
Although diamond is the hardest material known, it has some
problems. They are: brittleness, oxidation at temperatures above
600°C, rapid oxidation above 1000°C, and chemical reaction at
elevated temperatures with elements from groups IVa, Va, VIa, and
VIIa of the periodic table. Thus, diamond tools cannot be used for
machining ferrous materials. The major use of PCD tools is in
machining of nonferrous materials, and principally in machining of
aluminum-silicon alloys in the automobile industry for parts, such as
pistons, engine heads, blocks and manifolds, wheels, and transmis-
sion parts. Other applications of PCD tools is in the machining of
graphite, carbon-carbon composites, metal matrix composites, and
fiber reinforced plastics. PCD tools are also used in the woodwork-
ing industry in Europe and Asia.
The critical problem with the deposition of cBN films is the
build up of high compressive stress in the film, which leads to poor
adhesion. As the film thickness grows, the compressive stress also
grows, which leads to cracking and peeling off of the coating. Thus,
the key technical goal is the development of a deposition process, or
the modification of an existing deposition process, which will enable
the growth of a commercially viable thickness (at least 3 to 5 micron
meters) of cBN, on a cutting tool with sufficient adhesion to with-
stand the shearing forces of the machining process.
The commonly investigated deposition methods are ion beam
assisted deposition (IBAD), plasma assisted CVD (PACVD), and
unbalanced magnetron reactive sputtering. Energetic ion bombard-
ment, a high substrate temperature, and film stoichiometry are the
necessary conditions for the deposition and stabilization of the cubic
phase.
Cubic boron nitride is chemically inert to hot iron, steel, and
oxidizing environments. Therefore, PCBN tools are used to machine
hardened steels (Rc 50–65), high temperature nickel, cobalt, and iron
base alloys, grey and ductile cast irons, sintered ultrafine powdered
metals. Example of cast and ductile iron parts are: pumps, impellers,
shafts, engine blocks, brake rotors, and drums. Examples of PM
parts are: pressure plates, sprockets, cam shafts and gears. Although
the parts are generally Rc 40 to 50, they contain very hard phase
particles which can be as hard as Rc 60 to 65. Conventional tungsten
carbide tools wear rapidly due to the abrasion from these particles,
but PCBN tools can withstand the abrasion of these hard phases, and
give the desired long tool life with higher speed capability. Hard
turning to replace grinding is another expanding application of
PCBN tools. Another advantage of PCBN tools is their compatibility
with dry machining, which will become increasingly important with
the stringent environmental regulations on cooling fluids used in
machining. Recently, Badzian, et al.,[14] have proposed that silicon
carbonitride may be a rival to cBN. Much work needs to be done to
realize the viability of this proposal.
8.0 POLYCRYSTALLINE SUPERLATTICE COATINGS
A repeating layered structure of two different materials with

nanometer scale dimensions is called a superlattice. Another term for
the same type of material is nanolayered composites (NLC). The
reader is referred to a good introductory article on these materials by
Scott Barnett and Anita Madan.[15] Metallic superlattices (e.g., Cu-
Ni), were studied in the early 80s to investigate the super-modulus
effect, i.e., an enhancement in the elastic moduli of 150 to 500%,
when the superlattice period was reduced to a value around 2 nm.
More recent attempts to reproduce these results have failed. The
superhard superlattices are constructed from materials with the fol-
lowing properties, usually by sputter deposition from multiple tar-
gets. These properties are:
1. The materials must be initially hard by themselves,
e.g., nitrides, oxides, refractory metals.
2. There should be a large difference in shear moduli
between the two materials.
3. They are miscible, and the crystal structure is the
same.
4. Interface widths between the layers should be small.
Examples of these are the nitride/nitride superlattices NbN/
TiN, AlN/TiN. A NbN/TiN superlattice can have a Vickers hardness
up to 50 GPa, which is 2.5 times greater than what would be expected
from the rule of mixtures of the constituent materials. The hardness
enhancement is dependent on the superlattice period and the inter-
face width. It goes through a maximum with the superlattice period.
For example, for the NbN/TiN superlattice, the maximum in hard-
ness occurs a superlattice period around 8 to 9 nanometers. The
hardness enhancement is associated with the difficulty in moving a
dislocation across the interface. It is also possible to prepare
superlattices from materials with different crystal structures, which
would typically be immiscible. Such “non-isostructural” superlattices
could have substantial advantages over the former, where the layers
have the same crystal structure. Examples are TiN/BN NLC, where
one of them is cubic and the other is hexagonal. A thin layer of the
normally hexagonal material can be forced into a cubic structure
when sandwiched between structural template layers. Other ex-
amples are metal/nitride, for example, Mo/NbN, W/NbN. The met-
als are body-centered cubic and the nitride is face-centered cubic.
The practical question of using these superhard superlattices as
coatings for cutting tools raises other issues, namely the hot hardness
of the coating and the temperature threshold for interdiffusion be-
tween the nanolayers, which would destroy the periodic nanoscale
structure. For the TiN/BN materials, the hardness remains at a value
between 3400 and 4000 kg/sq mm after thermal cycling up to 700°C.
It drops down to 1000 kg/sq mm when the temperature is raised to
800°C. Since typical cutting tool temperatures during use range from
800 to 1000°C, suggesting that this coating would be unsuitable in
this thermal regime.
In spite of the above comment, Selinder, et al.,[16] tested the
performance of PVD TiN/TaN, and TiN/NbN superlattice coated
cemented carbide tools in stainless steel machining. The coatings
were grown on cemented carbide inserts by a combination of ion
plating of TiN, and reactive magnetron sputtering of NbN and TaN.
Due to the relatively rough surface of the carbide substrate, the
lamellae coatings were macroscopically disordered, but they did
exhibit a microscopic superlattice structure. For both material sys-
tems, the two constituent layers of the superlattice crystallized in the
cubic B1 structure. The hardness values at a load of 0.5 N were 32
GPA for the TiN/NbN samples, and 39 GPa for the TiN/TaN samples.
The tools were tested in a one tooth face milling operation under dry
cutting conditions in the machining of austenitic stainless steel,
Types 303 and 304. The results indicated a superior performance of
the tools coated with the harder lamellae coatings, as compared to
tools coated with single layer PVD or CVD coatings, which are
softer. This investigation has shown that nanoscale multilayer coat-
ings may open up new avenues for the development of high perfor-
mance stainless steel cutting tools.
9.0 NOVEL SUPERHARD COATINGS
A novel concept of superhard coating has been advanced by

Veprek, and his co-workers.[17] They propose a nanocrystalline/
amorphous composite material. The bulk of the material consists of
2–5 nm crystallites surrounded by amorphous material of 0.3–0.6 nm
thickness. Dislocations cannot generate in the small sized crystal-
lites, and crack propagation is stopped in the surrounding amorphous
mass. They discuss various methods to prepare the nc/a composite,
and settle on the low temperature Plasma assisted CVD process to
minimize interdiffusion. Ideal candidates for the nanocrystalline
phase are the transition metal nitrides such as TiN, VN, ZrN, etc.,
with amorphous Si3N4 being an ideal candidate for the amorphous
phase. The hardness increased rapidly with the decrease in the
crystallite size, reaching a value of 50 GPa, at crystallite size of 3–4
nm, along with an elastic modulus >500 GPa, which could be
reached in a film of thickness 4 to 5 micrometers, and load of 10–50
mN. The nc-TiN/a-Si3N4 composites showed a remarkable stability
against oxidation up to 800°C.
REFERENCES
1. Garside, International Conference on Metallurgical Coatings and

Thin Films, (April, 1988)
2. Harkmann, T., et al., Surface and Coatings Technology, 92:62
(1997)
3. Lunarska, E., et al., Surface and Coatings Technology, 85:125
(1996)
4. Rechberger, Brunner, and Dubach, Surface and Coatings Technology,
62:393 (1993)
5. Lux, B., et al., Thin Solid Films, 138:49 (1986); Vuorinen, S. and
Karlson, L., Thin Solid Films, 214:132 (1992)
6. Ling, Wan and Hon, Surface and Coatings Technology, 90:106
(1997)
7. Kelly, et al., Surface and Coatings Technology, 86/87:28 (1996)

8. Zywitski and Hoetzsch, Surface and Coatings Technology,
94/95:303 (1997)
9. Schneider, et al., Surface and Coatings Technology, 96:262 (1997)
10. Sjostrand and Soderberg, International Workshop on Pulsed Plasma
Surface Technology (PPST), San Diego, CA, (April 24–26, 1998)
11. Lahres and Jorgenson, Surface and Coatings Technology, 96:198
(1997)
12. Vetter, et al., Surface and Coatings Technology, 86/87:739 (1996)
13. Inspector, et al., Surface and Coatings Technology, 68/69:359 (1994)
14. Badzian, et al., Diamond and Related Materials (June, 1998)
15. Barnett, S., and Madan, A., Physics World (January, 1998)
16. Selinder, et al., Surface and Coatings Technology, 105:51–55 (1998)
17. Veprek, et al., Thin Solid Films, 268:64 (1994); Surface and Coatings
Technology, 86/87:394 (1996)
Index 535
Index
pressure-sensitive tape test 167

A scratch test 167
Abrasive wear 4, 91, 413 Adhesion tests 195, 196, 197, 210
applications 426 indentation 197
Absorption characteristics indenter 197
laser 48 Adhesive bond failure 166
Accelerated test apparatus 283, 295 Adhesive bonding mechanism 83
Acceleration (body force) test 167 Adhesive friction 267
Acetylene flame 86 Adhesive type wear 4
Acoustic emission 202 Advanced ceramics 412
curves 269 Advanced controlled arc (ACA) 519
detection 209 AES 135
sensor 240 Air plasma spraying 425
Acoustic imaging tests 167 Alloys 44
Activated centers 9 Alpha alumina coatings 525
Activated reactive evaporation 39 Alpha series
Adatom phase 522
energy 121 Alternating stress fields 164
mobility 136 Alumina films 55, 523
Adhesion 323 Amorphic diamond 484
behavior 204 Amorphous carbon
coating 222 hydrogen free 335
coating/substrate 194, 207 Amorphous carbon coatings 302, 306
enhancement 167 friction and wear 341
forces 158 mechanical properties 323
Adhesion strength 263, 266 Anisotropic columnar structure 160
Adhesion testing Annealing temperature 340
acceleration (body force) test 167 Anode 12, 29
electromagnetic stressing 167
535
Applications 3 Ball-on-disk wear tests 431

BN 467 Balzers process 389, 517
corrosion 519 BARE 60
decorative 527 Basic adhesion
high temperature 427 (BA) 194
non cutting tool 420 Beam
PCBN tools 530 current 35
PVD coatings 394 mode 35
tribological 520 Benzene-based precursor gas 310, 320
wear 426, 479 Berkovich
Ar concentration 320 hardness 249
Arc 18 indenter 241, 244, 259
anodic 40 pyramids 245
cathodic 491 Bi-axial stress 157, 213, 216
commercial coatings 39 state 222
decomposition 491 Bipolar 18
deposition advantage 41 BN
evaporated coatings 136 a-BN to c-BN 469
evaporation-deposition method 39 BN substrate
plasma 38 benefits of 481
PVD process 389 Boltzmann constant 68
regime 38 Bond coat 423, 424
techniques 38 Bonded ceramic materials 426
trajectories 39 Bonding effects 81
ARE 39 Bonding mechanisms 81
biased process 60 Boric acid
process 42, 52, 59 evaporation of 472
Argon ions 443 Boron fibers 438
Artificial implants 505 Boron hydride
ASTM definition 194 formation of 86
Atmospherical plasma spraying Boron nitride 300, 466, 528
(APS) 95 Brale 234
Atomic force microscope indenters 197
(AFM) 230, 283, 288 Brillouin scattering 164
Atomic hydrogen 303 Brinell hardness 233
Atomic structure 110 Brinell number 233
hardness 112 Brittle coatings
Atomistic phenomenon 23 resistance of 201
Austenitizing temperatures 432 Brittle failure 437
Autocatalytic 450 Brittle η-phases 378, 386
Automotive industry 503 Brittle materials 235, 259
Bulk
B coating 181
Balanced magnetron 16 diffusion 119
material 108
Ball bearings 233, 433
phase separation 141
Index 537
C Cathodoluminescence 133
CBN films
a-C coatings deposition 529
friction and wear properties 335 CCVD 61
a-C films Ceramic coatings 105
deposition conditions 307 Ceramic compounds 2
c-BN Ceramic matrix composites 437
deposition of 482 Cermets 426
engineering material 467 Characterization 2
properties of 481 Chemical
a-C:H films activation 9
wear-protective 135 bonding 194
β -C3N4 510 etching 434
Capacity 26 inertness 378
CAPS system interaction 24
advantage 99 removal process 415
Carbide ball bearings 433 wear 414
Carbide films Chemical vapor deposition 1, 5,
synthesis 51 60, 476
Carbide tools 532 Chemical vapor infiltration 438
uncoated cemented 526 Chemisorption 83
Carbon 456 Chlorides 386
Carbon coatings 302, 306 Chromium’s hardness 448
adhesion 323 Closed-system 500
friction and wear 341 Cluster growth
mechanical properties 311 cone-shaped 121
mechanical properties of 323 Co-evaporation 44
PVD techniques 306 Co-sputtering 44
Carbon films Coal liquification 433
hydrogen free 317 Coarse-grained film 328, 331
β -Carbon nitride 528 diamond 312
Carbon species 306 Coated surface properties 417
Carbon-nitrogen 300 Coating adhesion 166
Catalyst-solvent method 486 Coating material selection 416
Catalytic poisons 450 Coating methods 420
Catalyzed transformations 469 Coating process 87
Catastrophic failure 325 Coating rates 441, 444
Cathode 11, 12, 29, 32, 519 Coating systems 439
ACA 519 Coating technology 421
UMB 519 Coating temperatures 430
Cathode ray tubes 508 Coating/substrate
Cathodic arc adhesion 207
carbon coatings 327 interface 158
continuous 39 system 181
deposition 317
evaporation 491, 518
pulsed 39
Coatings Cosine Law 24

decorative 527 Covalent bonding 112
fundamental properties 182 Cr Κα radiation 216
omni directional 430 Crack formation 159
thin vs thick 446 Crack-tip plasticity 206
Cobalt bonded tungsten carbide 426 Cracking 258
CoCrAlY 444 Creep test 257
Cohesive energy 112, 113 Critical load determination 268
Cohesive failure 166 Cross-sectional imaging 110
Cold Cathode Plasma Electron Beam Cross-sectional TEM
Gun 34 XTEM 116
Cold wall reactor 64 Crossed cylinders test apparatus 284
Colors of coatings 527 Crucible
BN 467 contamination 29
Columnar grains 119 material 28
Combustion flame spraying 89 Crystal
Combustion methods 424, 486 growth 501
Combustion process 91 orientations 382
Combustion spraying 84 structure 146
Composite hard coatings 451 Crystalline alumina
Composite mixtures 412 deposition of 523
Composite powders 434, 435 Crystallograph planes 218
Composites 426 Crystallograph states 483
micro-laminate 1 Crystallography 468
nano-laminate 1 Cu Κα radiation 216
Composition contamination 146 Cubic boron nitride 2, 466, 528
Composition modulation 144, 162 Cubic phase 530
Compressive stresses 152, 154, 162 Current range 26
parallel planes 214 Cut-machining orientated coating
Conductive substrate research 401
heating 61 Cutting
Contact fatigue wear 414 continuous and interrupted 373
Contamination 146 diamond tools 387
Continuous cutting 373 edge chip-off 384
Continuous fibers 438 material 370
Controlled Atmosphere Plasma Spraying speed 376
(CAPS) 99 temperature 375
Conventional CVD 60, 61 Cutting fluid 525
Coolants 525 Cutting tool temperatures 532
Core filament materials 438 Cutting tools 181
Correlations CVD 5, 60, 133, 134, 142, 149,
hardness 149 168, 198, 218
intrinsic stress 149 advantages 6
microstructure 149 assisted techniques 303
Corrosion cemented carbide 392
applications 519 coated cemented carbides 391
fatigue 416 coatings 378, 405
protection 412 conventional 61
resistance 433 deposits 70
Index 539
diamond coatings 387 Deposition conditions 333

fluidized bed 435 a-C or a-C:H films 307
ion plating 153 Deposition rate 13, 24, 49, 58, 66
laser induced 62 limitation 51
metal organic 430 of diamond 303
plasma assisted 62 operating conditions 13
process 389, 405, 428 Deposition rates
process sequence 63 high 8, 41
sputtering 153 Depth of indentation 243
techniques 61, 160, 303 Depth-sensing apparatuses 237
thermal 519 Detonation gun process 89
CVI Diamond 302
diffusion 439 classification scheme 485
methods 438 crystalline 133
Cylindrical deposition 304
cathode cavity 34 formation 497
hollow magnetron technique 14 micromechanical properties 312
magnetrons 14, 50 natural 333
post magnetron 14 nucleation 456
sheet tools 529
D single crystal material 387
synthesis 101, 302
D-gun technique 424
Diamond coatings 100, 453
Data density 504
Data storage adhesion 323
characteristic of 134
media 504
DC surface mobility 133
Diamond films 342
plasma activation 495
commercial application 503
power source 36
sputter deposition 18 deposition 507
friction and wear 327
sputtering 388
Debond layer 437 on silicon 329
Diamond growth
Decaying plasma channel 38
thermodynamics of 497
Decorative coatings 527
Deformation 231 Diamond synthesizing
n-type 500
elastic-plastic 243
Density of defects 162 Diamond-like carbon 2, 134, 456
Diatomic gases 92
Deposited film size 43
Die casting operation 445
Depositing species source 9
Deposition Dielectric materials 26
Diffuse arc 40
compounds 47
conditions 109 Diffusibility 65
Diffusion 378
process 7
processes 440
process flexability 8
processes 109, 171, 323 Diffusion bonded coating 432
Diffusion bonding 431
techniques 325
technologies 2 Diode
ion plating 58
temperature 476, 520
sputtering process 13
volume 16
system 30
Direct deposition 492 emission 49

Direct evaporation 47 emitter assembly 33
processes 51 energy loss spectroscopy 133
Direct imaging 115, 246 ranges 9
Direct ion beam deposition 455 traps 14
Disappearing anode Electron beam 517
effect 19 guns 31
Disc cathode 33 heated sources 29
Discharge current 13 heating 54
Disk-on-disk test apparatus 286 methods 442
Dislocation loops Electron cyclotron resonance 476
generation of 219 Electronic
Dislocation networks 126 applications 453, 480
Displacement reactions 62 structure 110
Disposal costs for cutting fluids 526 Electroplating 447
Disproportionation reactions 63 Electrostatic
Dissociation energy 67 clamping 295
Droplet ejection 31 field 32
Dry corrosion 415 forces 194
Dual magnetron sputtering 525 Electrostatical
configuration 19 bonds 166
Ductility 380 Element evaporation 25
Duoplasmatron 21 Energy conversion efficiency 43, 48
Engineering applications 182, 213
E Environmental charge 527
E-beam gun 472 Environments
soft vacuum 435
EB process 518
Erosive wear 414, 433
Elastic deformation 206
Elastic properties 112 Evaporation 8, 440, 441
activated reactive 472
Elastic solid 242
Elastic waves 240 coefficient 24
comparisons 38
Elastic-plastic deformation 229, 231
gas scattering 38
Elastic-plastic solid 242
Elasticity 231 heat sources 22
process 22
Electric wire-arc spray process 79
Electrical processes 44
rate 23, 26, 29
biosensors 505
reactive 38
potential application 34
resistance 26 sources 24
thickness 26
Electrically hard materials 2
Electrochemical redox process 415 Expanding spherical cavity
model 187, 252
Electrode
Experimental adhesion
auxiliary 59
positively biased 53 of energy. See Adhesion
of forces 195
Electroless plating 447, 450
Electromagnetic stressing 167
Electron 17
bombardment 489
diffraction 116
Index 541
F Free standing shapes 426, 434

Friction
Failure mechanisms 373 coefficient 325, 328, 453, 505
Fatigue 262 measurement 295
Fatigue wear 413 micro-measurement 290
FBCVD profiles 325, 331, 342
applications of 435 tests 279
Feedback control techniques 51 Friction and wear 278
Feedstock materials 423 carbon coatings 341
Fermi level 507 measurement 287, 299
Ferromagnetic material 14 polycrystalline diamond films 327
Ferrous metals tests 279
machining of 478 Friction force 282, 294
Fibers 434 measurement 289
continuous 438 Friction force microscope
Film 42 (FFM) 230, 288, 292
formation 8 Friction-torque model 294
forming process 68 Frictional force measurement 202
growth 7, 68 Fusion process 79
stoichiometry 68, 530 FWHM 217
Film deposition 4
ion beam 135 G
laser techniques 135
parameters 109 Galvanic corrosion 415
reactive plasma 135 Gamma series 522
Filtered arc 518 Gas
evaporation process 40 composition 476
Fine-grained films 328, 331 ion bombardment 49
diamond 312 scattering evaporation 22, 38
Finite element method 252 vapor sources 70
Flame combustion 421 Gas precursors
methods 455 decomposition of 430
Flame spray Gas pressure 31
coatings 89 Gasification plants 433
processes 86 Geometric conditions of thermal
Flame torch method 486 spraying 95
Flash evaporation 45 Germanium optics 508
Flow stress 267 Getter surfaces 51
Fluidized bed 435 Gettering 522
Fluorocarbons 488, 497 Glow discharge 37, 58, 443
Focus adjustment 35 sputtering technology 21
Formation Gold like color 527
microstructures 400 Grain size 160, 328, 331
Fracture initiation 153 diamond film 314
Fracture resistance 197 Graphite cathode
Fracture toughness 258, 265 sputtering of 490
equation 260 Graphite like carbon 456
Graphitic carbon 307 High velocity oxy-fuel

Graphitic structures 467 (HVOF) 90
Graphitization 488 High-resolution TEM
Green manufacturing process 525 HRTEM 139
Griffith energy balance 204 High-speed steel substrates 147, 150
Grit blasting 84 Highly oriented pyrolytic graphite
Gross National Product 411 (HOPG) 288
Growth mode HK 235
2-D layer-by-layer 143 Hollow cathode activated deposition 53
3-D island 143 Homoepitaxial growth 502
Growth temperature 143 Hot Hollow Cathode Discharge Beam
Gun system 29 Gun 36
Hot plate 61
H Hot wall reactor 64
Hot wire systems 454
H-DLC films
production of 487 Hot-filament CVD technique 69
HRB 234
HAD process 55
HRC 235
Hall-Petch relation 114
Hall-Williamson plot 218 Humidity
effect on friction and wear 337
Hard coatings 1, 4
characterization of 115 HVOF 91, 425
spray technique 425
microstructures 110
Hydrocarbon precursors 488
superlattice system 143
Hard facing welding process 79 Hydrocarbon-hydrogen mixtures 133
Hydrogen 303
Hard materials 2
Hardness 230, 312 depletion 339
loss 333
apparatus 237
plasma 491
testers 233
Hardness number 231 Hydrogen free
amorphous carbon 335
Hardness values 448
Hastelloy 104 carbon films 317
methods 488
HB 249
Hydrogenated carbon coatings 306
HDA composite 28
Heat of formation 113 I
Heat sinks 100
Heat-sinks I-carbon 192
materials 506 IBED technique 471
Helix sources 26 Impingement limitation 51
Heteroepitaxial films Impinging ions 154
c-BN 474 Impression fatigue 262
growth of 502 Impurity atoms 9
High rate Inclusions
arc decomposition 491 hard coatings 451
reactive sputtering 444 Indentation 185, 202
High speed steels 387 adhesion test 197
High temperature oxidizing cracking 258
environments 437 creep test 257
Index 543
curve 242 Ion beam deposition 309, 491

depth 240 Ion beam energy 335
depth of 232 Ion beam enhanced deposition 455
fatigue test 262 Ion beam methods 455
hardness 230 Ion beam sputtering 20, 21, 22
hardness tests 231 Ion bombardment
load relaxation (ILR) 257 energy 68
method 114 Ion gun 16
plastic theory 203 Ion milling preparation technique 117
size effect (ISE) 160 Ion plating 9, 38, 56, 440, 443
spalling criterion 202 methods 441
testing 200 process 58
tests 167 Ion-induced intermixing 169
Indentation behavior Ion-surface interactions 307
elastic/plastic 203 Ionizable gas range 35
Indentation stress Ionization
elastic/plastic 205 efficiency 13
Indenter 185 energy 67
moving 208 plasma 472
penetration 188 Ions 17, 68
tip radius wear 207 bombardment 35, 37, 120, 150, 307
Index of refraction 508 collection 13
Indirect incorporation 124
deposition 492 IR absorption spectra 511
Induction ISE 160, 187, 192, 194
heated sources 29 Isothermal gradient pulse methods 440
heating 28
Industrial applications 468, 486 J
Industrial diamonds 453
Jet-Kote TM process 90
Inert-Gas Plasma Spraying (IPS) 98
Jewelry applications 527
Infrared spectrophotometry 468
Infrared spectroscopy 133 K
Injection conditions 94
Insulating ring 39 Kappa alumina coatings 525
Interdiffusion 166 Kappa phase 523, 525
Interface region Kaufman source 21
coating-substrate 117 Kinetic friction measurement 298
Interfacial adhesion 201, 230 Kinetic limitations 136, 143
Interfacial properties 181 Knoop hardness 235, 528
Intergranular cracking 415 Knoop indenter 236
Intermediate transformation 469 Knudsen’s cosine law 24
Internal strain gauge 211
Internal stress 205 L
thin films 220 Laser ablation 45, 48
Interplanar distance 140 Laser evaporation 42
Interplanar spacing 211, 213 Laser induced CVD 62, 69
Interrupted cutting 373 Laser irradiation 488, 493
Ion assisted dynamic mixing 510
Ion beam assisted deposition 530
Laser sources 431 Mass transfer rate 65

Laser techniques tests 167 Materials compatibility 433
Laser technology 509 Materials selection process 416
Laser-solid interactions 42 MD
Lateral force microscope (LFM) 288 2-D 124
Lattice Mean free path 22, 66
distortion 213 Measuring
mismatch 145 adhesion 220
parameter 140 friction force 202
strain 223 internal stress 211
Lattice defects Mechanical bond 423
grain boundaries 108 Mechanical deflection techniques 211
vacancies 108 Mechanical interlocking 166
LCVD processes 69 Mechanical loads 370
Line-of-sight processes 304, 441 Mechanical method 196
Linear cathode gun 33 Mechanical properties
Load displacement curve 244 diamond films 316
Load frame compliance 246 microprobe 186
Loadability of the tool 377 Melting 14, 26
Low friction coatings 522 systems 30
Low pressure CVD technology 65 Melting points
Low pressure plasma spray 425 high 26
Low vapor pressure 26 MEMS 294
Low-beam current 22 Metal
Low-load technologies 135 alloys 412
LPCVD filaments 438
advantage 66 sources and evaporates 28
Lubricants 526 Metal concentrations
effect on hardness 322
M Metal organic CVD 430
Machining technology 371 Metallic chlorides 386
Metallic colors 527
Macro arcs 18
Metallurgical grain 119
Macrohardness testers 233
Macroparticle Metastable phase 136
Methane
component 40
density 43 plasma decomposition 511
precursor gas 310, 316, 320
formation of 39
Meyer hardness 234
Macroscratch testers 268
Macrostrains 211 Meyer Index 187
Micro arcs 18
MAD process 54
Magnetic field 32 Micro-crack propagation 114
Microchemistry 109
strength 14
Microcutting 414
Magnetron activated deposition 53
Magnetron sputter ion plating process Microdroplets
formation of 442
(MSIP) 388
Magnetron sputtering 523 Microelectromechanical systems
(MEMS) 229
configurations 14
Microelectronic applications 16, 63
sources 17
Index 545
Microfriction measurements 290 Nickel coatings 451

Microhardness 235 NiCrAlY 444
equipment 183 Nitride superlattices 165
testers 184 Nitridic hard materials 400
Microindentation hardness apparatus 237 Nitrogen dioxide etching 506
Micromechanical elements 294 Nitrogen ion implantation 477
Micromotor 295 Nitrogen-to-metal ratio 121
Microprobe 230 Non-equilibrium microstructures 150
Microscratch testers 271 Non-isostructural superlattices 531
Microstrains Non-reactive supports 26
anisotropy 212 Nonconducting sputtering target 17
dislocations 212 Nuclear magnetic resonance
Microstructure 146, 149, 171 spectroscopy 133
films 109 Nucleation 456, 501
Microwave activation 496 density 148, 324
Microwave discharge-assisted CVD 306 method 196
Microwave plasma CVD 67
Miniaturization 504 O
Modulus of elasticity 251
Oils 526
Molecular dynamics 124, 150
Optical applications 51, 453
Molybdenum disulfide 522
Optical imaging 246
Monatomic gases 92
Optically hard materials 2
Monolayer superlattices 143
Organometallics 430
Monte Carlo technique 124
Oxide coatings 434
Morphology 146
Oxide crucibles 28
Multilayer coatings 109, 380, 382
Oxidizing environments
Multilayer structures 142
high temperature 437
N P
NaCl structure 136, 147 P-n junction diodes 480
Nanofatigue 262
P-type semiconductors 506
Nanoindentation 165, 237
PACVD 66, 68, 454
apparatus 237 advantages 68
fatigue 263
Parallel axial magnetic field 37
hardness 185 Particle bombardment 151
Nanoindenter 184, 193, 271
Particle diameters 435
testers 186
Particle size distribution 79
Nanolayered composites 531 Particle size feedstock 423
Nanoscratch measurements 277
PCBN tools 530
Near net shapes 434 PCD tools
Near stoichiometric films 44
applications of 529
Negative bias 17
Peierls stress 112
Negative pulse 18 Phase identification 110
Neutral radicals 67
texture 117
variables 67 Photochemical reaction 70
Ni-B hard coatings 451
Photoelectron spectroscopy 133
Ni-P hard coatings 452
Photolytic LCVD 69
Nickel bonded chromium carbide 426
Physical vapor deposition 440. See also Polycrystalline diamond 302,

PVD 387, 526, 529
Physisorption 83 films 314
Pierce type gun 33 friction and wear 327
Pin-into-bushing test apparatus 284 Positive pulse 18
Pin-on-cylinder test apparatus 284 Potential difference 29
Pin-on-disk test apparatus 284, 300 Powder flame spraying process 85
Pin-on-flat test apparatus 284 Powders 102, 423
Planar DC discharge 17 composite 435
Planar diode 50 mass flow rate 95
glow discharge 13 properties 94
Planar magnetron 14, 50 Power density 25
Planetary substrate holders 22 Power losses
Plasma 34, 51, 52 high 18
activation 66 Precursor gas 309, 320
density 52 benzene-based 310
discharges 9 Preform fibers 438
e-beam gun types 34 Pressure
generation 8 plating 22
gun 31 range 22
source 21 Pressure coating systems 439
spray process 95 Pressure-sensitive tape test 167
state 92 Processing parameters 8
Plasma assisted Processing techniques 476
CVD 60, 66, 530 Production costs 527
evaporation process 52 Pseudomorphic growth 136
reactive evaporation process 42, 56 Pull test 431
Plasma ionization 472 Pumping speed 51
Plasma jet process 454 PVD 5, 135, 150, 160, 183, 218
Plasma sprayed coatings 423 amorphous carbon coatings 306
Plasma spraying 84, 92 cemented carbides 391
low pressure 425 coated tools 391, 393, 401
Plasma temperature 425 coatings 117, 168, 392, 400
Plastic components deposits 70
production of 452 films 218
Plastic deformation 153, 231 methods 440, 489
Plastic indentation strain rate 258 process 44, 126, 136, 391, 393,
Plastic solid 242 400, 405
Plastic zone radii 187 reactive 48
Plating solution 450 spalllation 213
PLD 42 techniques 8
Ploughing 267, 325 techniques amorphous carbon
Point defect aggregates 126 coatings 306
Poisoning effect 49 titanium nitride 212, 220
Poisson’s ratio 198, 248, 253 Pyrolytic LCVD 69
effect 214 Pyrolytic reaction 62
Polyatomic molecules 70
Polycrystalline cubic boron nitride 529
Index 547
Q hollow cathode low voltage EB

guns 38
QQC Diamond Synthesis 501 induction 64
Quasi-amorphous films 476 induction power 320
power 38
R
power source 17
Radiant heat 61 signal 17
Radiation hard materials 2 sputter deposition 17, 18
Radical generation rate 67 sputtering 472
Radical mobility 68 sputtering frequencies 17
Radioactive decay 283 RIP process 59
Raman shift 482 Rockwell hardness test 234
Raman spectroscopy 133 Roller bearing applications 433
Random arc Rolling contact fatigue 446
disadvantage 39 Rolling contact fatigue (RCF) tests 286
sources 39 Rotating cylinder test apparatus 284
Rate limiting step 64, 66 Rotating four ball test apparatus 286
Rate monitor 26 Roughness 327, 333
RB 234 RS technique 49
RC 235
S
Reaction 25, 28
energy 69 SAD process 53
Reactive evaporation 38, 47, 51, 442 SAED patterns 137
Reactive ion plating Sand blast tests 431
process 59 Scanning electron micrographs 188
variation 60 Scanning electron microscope 115, 184,
Reactive plasma spraying 100 202, 232, 283
Reactive species 66 Scanning tunneling microscopy
Reactive sputtering 50 (STM) 283
problem 49 Scattering
process 49 backscattered neutral particles 152
Reactors 64 Scherrer’s equation 117
Recoil implantation 124 Scratch
Recrystallization 119 adhesion 202
Reduction reaction 62 depth profiles 325, 326
activated dissociation 472 depths 275
Reflected light microscopy 209 hardness 230
Refractory hard materials 442 loading 273
Refractory metal 36 resistance 266, 277
Refractory oxides 29 Scratch resistance/adhesion 342
sources 28 carbon coatings 327
Reichert wear tester 285 Scratch test 167, 168, 208, 210,
Residual stress 211, 222, 400 222, 266, 267
Resistant coatings 4 Scratch tester 268, 271
RF commercial 209
cycle 17 Selected-area electron diffraction
field 68 (SAED) patterns 137
glow discharge 389 Self-accelerated gun 30
glow discharge-assisted CVD 306
SEM 118 Sputter deposited films 479

Semiconductor applications 506 Sputter
Semiconductors deposition 11, 13, 44, 45, 49, 136
p-type 506 competitor 41
Separation method Sputter depostion
crystallite size 218 mid-frequency 18
Shapes Sputter etching 135
complex 520 Sputtered coating
Shear stress 265, 432 composition of 444
Shell four-ball tester 285 Sputtering 8, 9, 11, 440, 443
Shock wave method 478, 486 cathode 16
Shock wave test 167 gas pressure 119
Short bond length 112 high rate 14
Shrouded Plasma Spraying (SPS) 99 rate 13
Silica coatings 55 very high rate 524
Silicon carbide fibers 438 Stainless steel machining 532
Silicon carbonitride 530 Stamping oils 526
Silicon oxide films 339 Static friction measurement 295
Silicon scratch test 276 Static high pressure method 469
Sin2y 216 Static indentation 230, 231
method 213 Static pressure coating systems 439
plot 216, 222 Steel balls 233, 339
Single source evaporation 44 Steered arc 40, 518
Single-crystal superlattice film 143 Stoichiometry 47
Sliding tests 279 Stress corrosion cracking 416
Sliding wear 91, 284, 432 Stress enhanced corrosion 415
Source materials 28 Stresses 232
Sp3-bonded carbon 341 Structural characterization 135
Sp3-bonding 302 Structural densification 114
Space age materials 437 Structure zone models (SZM) 118
Spallation 201, 202 Structures
Spalling resistance 446 graphitic 467
Spectrophotometry Sublimation source 27, 29
infrared 468 Substrate biasing 310, 320, 339
Spinodal decomposition 141 a-C:H films 337
Spitting 14, 442 Substrates 21
Splashing effect 43 temperatures 68, 322, 455
Splat cooling 81 thermal management 499
Spotless arc deposition 53 Superconducting films 8
Spray coating Superhard
thermal 421 coatings 533
Spray environment 424 materials 2, 528
Sputter tools 529
coating process 11 Superlattice 143, 531
etching 167 coatings 109
influences 20 hardening 163
process 388 Supermodulus effect 164
pulsed magnetron 54 Surface
target erosion rate 11 contamination 146
Index 549
energy 113 characteristics 394

imaging tools 279 coatings 386
morphologies 329 TiAlN Coatings 136
pretreatments 324 TiC
properties 417 coated cemented carbide tools 377
reaction rate 65 CVD deposition 382
roughening 84 TiN 153, 527
Surface roughness 327, 328, 333 TiN/BN materials 532
profile 290 Titanium 522
Symmetrical pulsed DC 19 Titanium carbonitride 433
SZM 120, 147, 150 Titanium diboride 433
Tool applications 479
T Topographic imaging 289
Tangential Total friction force 206
Transfer layers 339
frictional stress 205
Tantalum filament 498 Transformation
catalyzed 469
Tc superconducting films 42
intermediate 469
Tec Vac process 517
TEM 118, 135, 144, 153, 220 Transition behavior analysis 50
Transition metal carbides 109
Temperature
cutting 375 Transmission electron microscopy
(TEM) 110, 115
cutting tool 532
Transmitting /focusing systems 42, 48
deposition 520
distribution 79 Transverse linear cathode guns 33
Triaxial stress 216
Tempering 520
Tensile and shear test 167 Tribaloy 104
Tribochemical wear 403
Tensile stress 159
Tribochemistry of RF PECVD carbon
Textile industry 105
Thermal activation 9 coatings 341
Tribological applications 222, 511, 520
Thermal barrier coatings 427
Thermal biosensors 505 Tribological properties 149
Tribological tests 294
Thermal conductors 29
Tribologically hard materials 2
Thermal DC plasma jet 101
Thermal decomposition reaction 62 Tribomechanical system 373
Triode sputtering 60
Thermal gradient methods 439
Thermal loads 370 U
Thermal mismatch factor 211
Thermal spray coating 81, 421 U beam 238
Thermal spraying 77, 79, 107 Ulvac process 517
Thermal stress 150, 159 Unbalanced magnetron 16, 518, 519
Thermionic gun 31 sputtering 530
Thermodynamics 23 Union carbide 28
Thermoelastic instability 414, 427 Unipolar pulse 18
Thickness monitor 26 Unstressed titanium nitride 215
Thin films
anisotropic 109 V
Thrust-washer test apparatus 284
Vacuum arcs 39
Ti-C-N
Vacuum deposition system 43
Vacuum environments 435 Wear-resistant coatings 183

Vacuum plasma spraying 97 Weibull distribution 209
Vacuum pulse cycles 440 Welding process 414
Vacuum tube amplifiers 508 Wet corrosion
van der Waals bonds 166 galvanic and nongalvanic 415
van der Waals forces 194 Wire-arc spraying process 77
Vapor Wobble micromotor 295
distribution 24 Work-accelerated gun 30
flux 22 Wurtzite 136, 467
high density 26
sources 25 X
transportable 441
X-ray analysis 167
Vapor deposition 4, 115
X-ray diffraction 116, 149, 212, 482
Vapor phase 23, 24, 147
X-ray lithography 481
reactions 49
X-ray techniques 211
specie 7
X-Y-Z table 238
Vapor pressure 24, 26
XPS analysis 510
Vaporization rate 47
XRD 118, 144, 164
Vibrational entropy 113
patterns 136
Vickers 187
XTEM 126, 147, 164, 169
hardness 531
micrograph 141, 143
indentation measurement 193
renucleation rates 117
indenter 197, 235, 245, 259
indents 188 Y
microhardness 163
pyramids 245 Young’s modulus 198, 213, 229,
Viscoelastic/viscoplastic materials 232 231, 248, 312
Voltages
asymmetric and symmetric 18 Z
Volume law of mixtures 186, 251 Zirconia 427
VPS 98, 100 Zone 1 150
advantages 97 structure 121, 146
disadvantages 98
VPS/LPPS processes 98
W
Warren-Averbach’s method 117
Wear 413
and corrosion 411, 420
depths 291
measurements 282, 287
micro-measurements 291
protection 412, 479
resistance 279
resistance and ductility 377
test 282
Wear rates 300
of counterface materials 329

(Materials and Processing Technology) Rointan F. Bunshah, Rointan F. Bunshah - Handbook of Hard Coatings - Deposition Technologies, Properties and Applications-William Andrew (2002)

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

(Materials and Processing Technology) Rointan F. Bunshah, Rointan F. Bunshah - Handbook of Hard Coatings - Deposition Technologies, Properties and Applications-William Andrew (2002)

Uploaded by

Copyright:

Available Formats

HANDBOOK OF HARD COATINGS

Deposition Technologies, Properties and

Published in the United States of America by

Library of Congress Cataloging-in-Publication Data

Handbook of hard coatings / edited by Rointan F. Bunshah

HANDBOOK OF SPUTTER DEPOSITION TECHNOLOGY: by Kiyotaka Wasa and

Professor Rointan F. (Ron) Bunshah

Distinguished Teaching Award of the Engineering Graduate Student Association

Stephen Rossnagel June, 2000

This volume covers a wide range of topics relating to the fabrication,

Bharat Bhushan Otto Knotek

Steve J. Bull G. Krämer

Rointan F. Bunshah F. Löffler

To the best of our knowledge the information in this publication is

2 Vapor Deposition Technologies ............................................. 4

5.0 CHEMICAL VAPOR DEPOSITION (CVD) .................................... 60

3 Thermal Spraying and Detonation Gun Processes............ 77

4 Structure/Property Relationships For Hard Coatings .... 108

4.0 INFLUENCE OF MICROSTRUCTURE ON COATING

5 Characterization of Hard Coatings................................... 181

6 Macro- and Micromechanical

4.0 MECHANICAL AND TRIBOLOGICAL PROPERTIES OF

7 Applications to Cutting Tools ............................................ 370

8 Wear and Corrosion Resistant Hard Coatings

9 Cubic Boron Nitride and Diamond-Related Thin Films. 466

3.0 DIAMOND-RELATED MATERIALS ............................................ 482

10 Summary, Developments, and Outlook ............................ 517

Index .......................................................................................... 535

Mankind has used hard surfaces on various tools since antiquity.

the development of modern technology in the areas of optical, optoelec-

3. Applications. The chapters are: “Applications to Cutting

1. Aitchison, L., A History of Metals, Interscience Publishers, 2:351

This chapter deals with the vapor deposition technologies used

2.0 CLASSIFICATION OF VAPOR DEPOSITION

Vapor deposition processes may principally be divided into

Table 1. Classification of PVD and CVD Processes

Basic PVD Processes Basic PAPVD Processes

Basic CVD Process Basic PACVD Processes

Table 2 shows the characteristics of basic PVD and CVD

Table 2. Characteristics of Basic PVD and CVD Processes

Evaporation Sputtering Chemical Vapor

Mechanism of Thermal energy Momentum Chemical reaction

As with all processes, one is concerned with the process itself,

Figure 1. The three steps in film deposition.

Step 1: Creation of Vapor Phase Specie—Material can

The degree of independent control of these three steps deter-

3.0 PHYSICAL VAPOR DEPOSITION (PVD)

Thermal activation of the condensing species resulting

Thus, for thin continuous films, the electron energies should be

3.2 Sputter Deposition

Sputtering Phenomena. Sputtering is the phenomena of mo-

where J is the ion current density in mA/cm2, S is the sputtering yield

where: Mi: mass of the incident atom

Planar Diode Glow Discharge Sputter Deposition. This is

Figure 4. Planar diode sputter deposition.

The discharge current increases with the applied voltage, thus,

In magnetron sputter deposition process, an applied magnetic

Figure 5. Cylindrical magnetron (post cathode) sputter deposition.

Figure 6. Planar magnetron sputter deposition.

Balanced and Unbalanced Magnetrons. The magnetic field

Figure 7. Balanced and unbalanced sputter deposition.

RF Sputter Deposition. The development of the RF sputter

on reactive sputtering in Sec. 4.3). Professor S. Schiller and his