Microtechnology and MEMS

Martin Dienwiebel
Maria-Isabel De Barros Bouchet Editors

Advanced
Analytical
Methods in
Tribology
Microtechnology and MEMS

Series editors
Hiroyuki Fujita, CIRMM, Institute of Industrial Science, Tokyo, Japan
Zhong Lin Wang, School of Materials Science and Engineering, Georgia Institute
of Technology, Atlanta, GA, USA
Jun-Bo Yoon, School of Electrical Engineering, Korea Advanced Institute of
Science and Technology, Daejeon, Korea (Republic of)
The series Microtechnology and MEMS comprises textbooks, monographs, and
state-of-the-art reports in the very active field of microsystems and microtechnology.
Written by leading physicists and engineers, the books describe the basic science,
device design, and applications. They will appeal to researchers, engineers, and
advanced students.

More information about this series at http://www.springer.com/series/4526

Martin Dienwiebel
Maria-Isabel De Barros Bouchet
Editors

Advanced Analytical
Methods in Tribology

123
Editors
Martin Dienwiebel Maria-Isabel De Barros Bouchet
Fraunhofer Institute for Mechanics Laboratoire de Tribologie et Dynamique des
of Materials Systèmes, CNRS-UMR5513, Ecole
Freiburg, Baden-Württemberg, Germany Centrale de Lyon
Université de Lyon I
and Écully, France

Institute for Applied Materials

Karlsruhe Institute of Technology KIT
Karlsruhe, Germany

ISSN 1615-8326
Microtechnology and MEMS
ISBN 978-3-319-99896-1 ISBN 978-3-319-99897-8 (eBook)
https://doi.org/10.1007/978-3-319-99897-8

Library of Congress Control Number: 2018952882

© Springer Nature Switzerland AG 2018

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

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

Friction- and wear-related phenomena are experienced by almost everybody in

daily life. Whether it is by walking on slippery ground, commuting to work, or
when travelling by train, airplane or on a cruise ship, the science known as tri-
bology is omnipresent but mostly hidden to the eye. However, although we
appreciate low friction, for example, when we go inline skating or skiing during the
weekend or high friction when we want to get our car to stop in front of a red traffic
light, we usually do not know how to control the coefficient of friction using
established theories. Scientists strive to unravel the mysteries of friction and wear
processes since the very first systematic studies more than 500 years ago yet uni-
versal guidelines for the selection of the best materials, coatings and chemicals are
not easily at hand. The reason for this is not simple as frictional properties are
governed by mechanisms that can appear on many different length scales from
several metres to the molecular scale and can be related to the topography of sliding
surfaces, to the composition and distribution of phases or to the surface chemistry
directly. Because of this complexity, an experimentalist needs to use, besides a
well-designed tribological apparatus, also other instruments in order to characterize
the surfaces, lubricants and materials.
A century ago, the optical microscope was probably the main and only tool for
the analysis of sliding surfaces. Many important properties of tribological systems
were still hidden at that time. Since then, many new scientific instruments have
been invented that can and need to be utilized to improve our understanding of what
causes high or low friction in various environments. These analytical instruments
undergo constant improvements in resolution, data acquisition and analysis speed.
The constantly filling toolbox of analytical methods available to tribologists
enabled novel technologies such as coatings and lubricants that provide ultra-low
friction and very small wear rates or that allow changing the friction by applying a
potential. It was possible to understand the fundamental steps that lead to these
improvements by a combination of suitable analytical techniques and also by a
combination of experiments and computer simulations. In many recent develop-
ments, those techniques have proven to be particularly important that could be
applied in situ during the experiment or shortly afterwards.

v
vi Preface

The aim of this book is to provide a broad overview of the possibilities that
modern state-of-the-art analytical and simulation techniques give in combination
with tribological experiments. In 11 chapters, leading experts introduce the ana-
lytical methods and present examples from the recent research on how the method
can be successfully applied.
Chapters 1 and 2 deal with microstructural characterization of sliding interfaces.
The first chapter presents the method of focused ion beam analysis and related
techniques in order to obtain information on shear-induced changes in the
microstructure. The second chapter demonstrates the use of the transmission elec-
tron microscope (TEM) for the analysis of ultra-low wear tribocouples. The fol-
lowing six chapters are devoted to various techniques for the investigation of
tribochemical mechanisms. The application of Near Edge X-Ray Absorption Fine
Structure Spectroscopy (NEXAFS) and related Synchrotron-based spectroscopic
techniques for the investigation lubricant chemistry and surface termination are
presented in Chap. 3. Chapter 4 is giving an overview over the possibilities of
gas-phase lubrication experiments in combination with in situ X-ray photoelectron
spectroscopy (XPS). Chapter 5 demonstrates the possibilities of the in situ analysis
of tribochemical reactions using Auger electron spectroscopy and low-energy
electron diffraction (LEED). The analysis of anti-wear tribofilms by X-ray
absorption near-edge spectroscopy (XANES) is covered in Chap. 6. Chapter 7
demonstrates the possibilities of micro-FTIR for the in situ analysis of lubricants
and greases. Chapter 8 is devoted to the analysis of the use of synchrotron radiation
in order to measure real-time wear dynamics. The following three chapters deal
with in situ microscopy techniques. Electrochemical atomic force microscopy and
atomic-scale friction measurements in an electrolytical environment are introduced
in Chap. 9. Examples for friction measurements inside an electron microscope are
presented in Chap. 10, and Chap. 11 then reviews instrumentation and the appli-
cation of in situ topography measurements. The final chapter of this book is ded-
icated to simulations as a numerical analytical tool.
Obviously, this book is not intended to be an exhaustive review of all analytical
techniques available nor does it cover all research in tribology as this would go
beyond the bounds of the content of a single book. At this point, the editors are
extremely thankful to all authors who contributed to the chapters of the book and
their time spent in carefully selecting interesting portions of their work. The editors
gratefully acknowledge support from their funding agencies (Centre National de la
Recherche Scientifique de France and Deutsche Forschungsgemeinschaft).

Freiburg/Karlsruhe, Germany Martin Dienwiebel

Écully, France Maria-Isabel De Barros Bouchet
Contents

Part I Microstructural Characterization

1 Electron Microscopy and Microanalysis for Wear Surface
Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 3
Joseph R. Michael, Paul G. Kotula and Somuri V. Prasad
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 3
1.2 Focused Ion Beam Preparation and Scanning Electron
Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 5
1.2.1 Focused Ion Beam (FIB) Preparation of Samples . .... 5
1.2.2 Electron Backscatter Diffraction and Transmission
Kikuchi Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3 Transmission Electron Microscopy and Microanalysis . . . . . . . . 12
1.3.1 Transmission Electron Microscopy . . . . . . . . . . . . . . . 14
1.3.2 Scanning Transmission Electron Microscopy . . . . . . . . 15
1.3.3 Microanalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.3.4 Spectral Image Analysis . . . . . . . . . . . . . . . . . . . . . . . 17
1.3.5 Multivariate Statistical Analysis . . . . . . . . . . . . . . . . . 18
1.4 Tribology Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.4.1 Diamond-Like Carbon Coating Characterization . . . . . . 20
1.4.2 Wear-Induced Recrystallization . . . . . . . . . . . . . . . . . . 20
1.4.3 Tribology of Nanocrystalline Metals . . . . . . . . . . . . . . 23
1.4.4 Transition Metal Dichalcogenides . . . . . . . . . . . . . . . . 25
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic
Materials by Transmission Electron Microscopy . . . . . . . . . . . . . . . 29
Alfons Fischer, Wlodzimierz Dudzinski, Birgit Gleising
and Priska Stemmer
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.1.1 TEM Analytics of the Cyclic Strain Gradient . . . . . . . . 33

vii
viii Contents

2.1.2 TEM Analyses of the Tribomaterial . . . . . . . . . . . . . . . 38

2.1.3 Wear Particle Analyses by TEM . . . . . . . . . . . . . . . . . 47
2.2 Conclusions and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Part II Tribochemical Characterization Techniques

3 Near Edge X-Ray Absorption Fine Structure Spectroscopy:
A Powerful Tool for Investigating the Surface Structure
and Chemistry of Solid Lubricants . . . . . . . . . . . . . . . . . . . . . .... 63
Filippo Mangolini and J. Brandon McClimon
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 64
3.2 Principles of Near Edge X-Ray Absorption Fine Structure
(NEXAFS) Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.3 Experimental Methods in NEXAFS Spectroscopy . . . . . . . . . . . 70
3.3.1 Data Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.4 Applications of NEXAFS Spectroscopy . . . . . . . . . . . . . . . . . . 76
3.4.1 NEXAFS Spectroscopic Investigation
of the Surface Structure and Chemistry
of Thin Films for Tribological Applications . . . . . .... 76
3.4.2 NEXAFS Spectroscopic Investigation of the
Thermally-Induced Structural Evolution
of Solid Lubricating Films . . . . . . . . . . . . . . . . . .... 83
3.4.3 NEXAFS Spectroscopic Investigation
of the Tribochemistry of Solid Lubricants . . . . . . .... 87
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 96
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 96
4 Tribochemistry of n-Alkane Thiols Examined by Gas-Phase
Lubrication (GPL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Maria-Isabel De Barros Bouchet and Jean-Michel Martin
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4.2.1 Gas-Phase Lubrication (GPL) Coupled
with In Situ XPS/AES Surface Analyses . . . . . . . . . . . 109
4.2.2 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4.3 Adsorption of Hexanethiol on Steel . . . . . . . . . . . . . . . . . . . . . 113
4.3.1 Procedure for Adsorption Study . . . . . . . . . . . . . . . . . 113
4.3.2 Adsorption of Hexanethiol on the Native Oxide
Layer of Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.3.3 Adsorption of Hexanethiol on Metallic Iron
(Etched Steel Surface) . . . . . . . . . . . . . . . . . . . . . . . . 115
4.4 Model GPL Experiment for Understanding the Formation
Mechanism of Hexanethiol-Based Tribofilm . . . . . . . . . . . . . . . 116
Contents ix

4.4.1 Tribochemical Reactions of Hexanethiol

on Oxidized Steel and Metallic Iron . . . . . . . . . . . . . . 117
4.4.2 Study of the Role of Nascent Metallic Surfaces
on the Tribochemistry of Hexanethiol . . . . . . . . . . . . . 118
4.5 Ex Situ Analyses of the Iron Sulfide Tribofilm
Formed by Thiol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.5.1 Formation of a 1-Hexanethiol Tribofilm by GPL . . . . . 121
4.5.2 ToF-SIMS Chemical Analysis of the 1-Hexanethiol
Tribofilm Formed by GPL . . . . . . . . . . . . . . . . . . . . . 122
4.5.3 TEM Observations and EDS Chemical Analyses
of the 1-Hexanethiol Tribofilm Cross Section . . . . . . . . 123
4.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5 In-Situ Measurement of Tribochemical Processes
in Ultrahigh Vacuum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Wilfred T. Tysoe
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.2 In-Situ Analyses of Sliding Interfaces . . . . . . . . . . . . . . . . . . . 131
5.2.1 In-Situ Analyses with Opaque Interfaces . . . . . . . . . . . 131
5.2.2 In-Situ Analyses of Transparent Interfaces . . . . . . . . . . 135
5.2.3 Pseudo In-Situ Analyses . . . . . . . . . . . . . . . . . . . . . . . 137
5.3 Elementary Steps in Tribochemical Reactions . . . . . . . . . . . . . . 144
5.3.1 Gas-Phase Lubrication of Copper by DMDS . . . . . . . . 146
5.3.2 Thermal Chemistry of DMDS on Copper . . . . . . . . . . 147
5.3.3 Identifying Elementary Steps During Sliding . . . . . . . . 149
5.3.4 Microkinetic Modelling of Tribofilm Formation
for DMDS on Copper . . . . . . . . . . . . . . . . . . . . . . . . . 152
5.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6 Tribochemistry and Morphology of P-Based Antiwear Films . . . . . 159
A. Dorgham, A. Neville and A. Morina
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
6.2 Chemical Nature of ZDDP and DDP . . . . . . . . . . . . . . . . . . . . 161
6.3 Reactions of ZDDP and DDP . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.3.1 Ligand Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.3.2 Decomposition of Peroxides and Peroxy-Radicals . . . . 163
6.3.3 Formation of Tribo- and Thermal Films . . . . . . . . . . . . 164
6.3.4 Hydrolytic Decomposition . . . . . . . . . . . . . . . . . . . . . 169
6.4 Composition of Antiwear Films . . . . . . . . . . . . . . . . . . . . . . . . 170
6.4.1 Phosphorus, Zinc and Iron Species . . . . . . . . . . . . . . . 171
6.4.2 Sulphur Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
6.4.3 Factors Affecting the Tribofilms’ Composition
and Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
x Contents

6.5 Mechanical Properties of Antiwear Films . . . . . . . . . . . . . . . . . 187

6.5.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
6.5.2 Hardness and Elastic Modulus . . . . . . . . . . . . . . . . . . 192
6.5.3 Tenacity and Durability . . . . . . . . . . . . . . . . . . . . . . . 196
6.6 Rheological Properties of Antiwear Films . . . . . . . . . . . . . . . . . 197
6.7 Tribological Properties of Antiwear Films . . . . . . . . . . . . . . . . 198
6.7.1 Friction Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6.7.2 Antiwear Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . 201
6.8 Antiwear Films in Humid Environments . . . . . . . . . . . . . . . . . 203
6.8.1 Effect of Water on the Composition of Tribofilms . . . . 204
6.8.2 Effect of Water on the Tribological Properties
of Tribofilms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6.9 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
7 In Situ Observation of Lubricating Films by Micro-FTIR . . . . . . . 215
Shigeyuki Mori
7.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.1.1 Tribological Contact . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.1.2 In Situ Observation . . . . . . . . . . . . . . . . . . . . . . . . . . 216
7.2 Principle and Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.2.1 Infrared Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.2.2 Lubrication Tester . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.2.3 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
7.3 Structure of EHL Film . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.3.1 Film Thickness and Pressure . . . . . . . . . . . . . . . . . . . . 219
7.3.2 Component of Lubricant . . . . . . . . . . . . . . . . . . . . . . . 220
7.3.3 Chemical Information . . . . . . . . . . . . . . . . . . . . . . . . . 222
7.4 Molecular Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
7.4.1 Interaction of Components . . . . . . . . . . . . . . . . . . . . . 223
7.4.2 Effect of Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.5 Molecular Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
7.5.1 Effect of Shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
7.5.2 Effect of Molecular Structure . . . . . . . . . . . . . . . . . . . 225
7.6 Boundary Lubrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
7.6.1 Adsorbed Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
7.6.2 Reaction Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
7.7 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
7.7.1 Water-Based Lubricant . . . . . . . . . . . . . . . . . . . . . . . . 228
7.7.2 Semi-Solid Lubricant . . . . . . . . . . . . . . . . . . . . . . . . . 229
7.7.3 Paper Friction Material (Wet Clutch) . . . . . . . . . . . . . . 231
Contents xi

7.7.4 Wire Drawing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

7.8 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
8 Micro-scale Real-Time Wear Dynamics Investigated
by Synchroton Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
M. Belin, Jean-Michel Martin, J. Schou, I. L. Rasmussen,
R. Feidenhans’l, T. Straasø and N. J. Mikkelsen
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243

Part III In-Situ Microscopy and Topography Measurements

9 Electrochemical Friction Force Microscopy . . . . . . . . . . . . . . . . . . 247
Florian Hausen
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
9.2 Electrochemical Foundations . . . . . . . . . . . . . . . . . . . . . . . . . . 249
9.2.1 Electrochemical Cell Design . . . . . . . . . . . . . . . . . . . . 253
9.3 Historical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
9.4 Friction Force Microscopy Under Electrochemical
Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
9.4.1 Aqueous Electrolytes . . . . . . . . . . . . . . . . . . . . . . . . . 256
9.4.2 Ionic Liquids as Electrolytes . . . . . . . . . . . . . . . . . . . . 263
9.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
10 In Situ Friction Tests in a Transmission Electron Microscope . . . . 273
Fabrice Dassenoy
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
10.2 Example 1: In Situ Compression Test on a Single MoS2
Nanoparticle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
10.3 Example 2: In Situ Compression Test on a Single WS2
Nanoparticle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
10.4 Example 3: In Situ Sliding Test on a Single MoS2
Nanoparticle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
10.5 Example 4: Friction Coefficient Measured on a Single
WS2 Nanoparticle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
10.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
11 In Situ Digital Holography for 3D Topography Analysis
of Tribological Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Martin Dienwiebel and Pantcho Stoyanov
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
11.1.1 Optical Roughness Measurements . . . . . . . . . . . . . . . . 290
11.1.2 Digital Holographic Microscopy (DHM) . . . . . . . . . . . 291
xii Contents

11.2 Integration of DHM into a Tribometer . . . . . . . . . . . . . . . . . . . 292

11.2.1 Analysis of Plowing Friction . . . . . . . . . . . . . . . . . . . . 293
11.3 Topography and Wear Evolution in Metal/Ceramic
Tribosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
11.3.1 Tungsten Sliding Against Tungsten Carbide . . . . . . . . . 298
11.3.2 Tungsten Sliding Against Diamond-like Carbon . . . . . . 299
11.3.3 Aluminum Sliding Against Sapphire . . . . . . . . . . . . . . 302
11.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304

Part IV Computer Simulations

12 Understanding the Tribochemistry of Lubricant Additives
by Ab initio Calculations: The Case of Phosphites . . . . . . . . . . . . . 309
M. Clelia Righi
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.2 Combining First Principles Calculations and XPS Experiments
to Study the Adsorption of Lubricant Additives . . . . . . . . . . . . 312
12.3 Effects of Mechanical Stresses on Additive Decomposition
at the Tribological Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
12.4 Functionality of Phosphorous in Reducing Adhesion
and Friction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
12.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Contributors

Maria-Isabel De Barros Bouchet Laboratoire de Tribologie et Dynamique des

Systèmes, Ecole Centrale de Lyon, CNRS, UMR5513, Université de Lyon I, Lyon,
Ecully, France
M. Belin Laboratoire de Tribologie et Dynamique des Systèmes, CNRS UMR
5512, Ecole Centrale de Lyon, Ecully, France
M. Clelia Righi Department of Physics, Informatics and Mathematics, University
of Modena and Reggio Emilia, Modena, Italy
Fabrice Dassenoy Laboratoire de Tribologie et Dynamique des Systèmes, CNRS,
UMR 5512, Ecole Centrale de Lyon, Ecully, France
Martin Dienwiebel Institute for Applied Materials, Microtribology Center µTC,
Karlsruhe Institute of Technology KIT, Karlsruhe, Germany; Fraunhofer-Institute
for Mechanics of Materials, Freiburg, Germany
A. Dorgham School of Mechanical Engineering, Institute of Functional Surfaces,
University of Leeds, Leeds, UK
Wlodzimierz Dudzinski Faculty of Technology and Engineering, Wroclaw
University of Science and Technology, Wroclaw, Poland
R. Feidenhans’l Niels Bohr Institute, University of Copenhagen, Copenhagen,
Denmark
Alfons Fischer Materials Science and Engineering, University of Duisburg-Essen,
Duisburg, Germany; CeNIDE-ICAN, Interdisciplinary Center for Analytics on the
Nanoscale, University of Duisburg-Essen, Duisburg, Germany
Birgit Gleising CeNIDE-ICAN, Interdisciplinary Center for Analytics on the
Nanoscale, University of Duisburg-Essen, Duisburg, Germany

xiii
xiv Contributors

Florian Hausen Forschungszentrum Jülich, Institute of Energy and Climate

Research, Jülich, Germany; Institute of Physical Chemistry, RWTH Aachen
University, Aachen, Germany; Jülich-Aachen Research Alliance,
Section JARA-Energy, Jülich, Germany
Paul G. Kotula Sandia National Laboratories, Albuquerque, NM, USA
Filippo Mangolini Department of Mechanical Engineering, Materials Science and
Engineering Program, The University of Texas at Austin, Austin, TX, USA
Jean-Michel Martin Laboratoire de Tribologie et Dynamique des Systèmes,
Ecole Centrale de Lyon, CNRS, UMR5513, Université de Lyon I, Lyon, Ecully,
France
J. Brandon McClimon Department of Materials Science and Engineering,
University of Pennsylvania, Philadelphia, PA, USA
Joseph R. Michael Sandia National Laboratories, Albuquerque, NM, USA
N. J. Mikkelsen CemeCon AS, Åbyhøj, Denmark
Shigeyuki Mori Department of Chemical Engineering, Faculty of Engineering,
Iwate University, Morioka, Iwate, Japan
A. Morina School of Mechanical Engineering, Institute of Functional Surfaces,
University of Leeds, Leeds, UK
A. Neville School of Mechanical Engineering, Institute of Functional Surfaces,
University of Leeds, Leeds, UK
Somuri V. Prasad Sandia National Laboratories, Albuquerque, NM, USA
I. L. Rasmussen DTU Fotonik, Technical University of Denmark, Roskilde,
Denmark
J. Schou DTU Fotonik, Technical University of Denmark, Roskilde, Denmark
Priska Stemmer Materials Science and Engineering, University of
Duisburg-Essen, Duisburg, Germany; CeNIDE-ICAN, Interdisciplinary Center for
Analytics on the Nanoscale, University of Duisburg-Essen, Duisburg, Germany
Pantcho Stoyanov Institute for Applied Materials, Microtribology Center µTC,
Karlsruhe Institute of Technology KIT, Karlsruhe, Germany; Fraunhofer-Institute
for Mechanics of Materials, Freiburg, Germany
T. Straasø DTU Fotonik, Technical University of Denmark, Roskilde, Denmark;
Niels Bohr Institute, University of Copenhagen, Copenhagen, Denmark
Wilfred T. Tysoe Department of Chemistry and Biochemistry, University of
Wisconsin Milwaukee, Milwaukee, WI, USA
 Part I
Microstructural Characterization
Chapter 1
Electron Microscopy and Microanalysis
for Wear Surface Characterization
Focused Ion-Beam, Electron Backscattered Diffraction
and Transmission Electron Microscopy

Joseph R. Michael, Paul G. Kotula and Somuri V. Prasad

Abstract Wear surface analysis, specifically in the subsurface regions, is essential

for understanding the fundamental mechanisms of friction and material removal dur-
ing wear. The advent of focused ion-beam systems (combined scanning electron and
focused ion-beam microscopes) has made it possible to prepare site-specific cross
sections of wear surfaces for scanning electron microscopy, electron backscattered
diffraction, transmission electron microscopy, and X-ray microanalysis. During the
past decade, these state-of-the-art electron microscopy techniques are being increas-
ingly used to characterize wear-induced changes to microstructures and crystallo-
graphic textures.

1.1 Introduction

Wear damage is typically initiated in the subsusurface regions of the contacting slid-
ing surfaces before it is fully manifested on the top surface, see Fig. 1.1. In metallic
alloys, sliding contact results in plastic deformation, which often leads to recrystal-
lization and development of crystallographic texture [1]. In the case of brittle mate-
rials, microscopic fracture events in the subsurface regions are the principal source

J. R. Michael · P. G. Kotula · S. V. Prasad (B)

Sandia National Laboratories, Albuquerque, NM 87185-0886, USA
e-mail: svprasa@sandia.gov
J. R. Michael
e-mail: jrmicha@sandia.gov
P. G. Kotula
e-mail: pgkotul@sandia.gov
© Springer Nature Switzerland AG 2018 3
M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_1
4 J. R. Michael et al.

Load

Wear Sub-Surface Debris Transfer

Pin

A
Disk

Fig. 1.1 Schematic illustration of the cross-section of a typical pin-on-disk wear system showing
the subsurfaces

for wear damage [2]. Amorphous to crystalline transitions in the subsurface regions
of transition metal dichalcogenide solid lubricating thin films (e.g., MoS2 , WS2 ),
with basal planes aligning in the sliding direction, and formation of transfer films on
the counterface have been well documented [3]. Now it is widely acknowledged that
a fundamental understanding of the friction-induced microstructural changes in the
subsurface regions is essential for analyzing the wear mechanisms.
Up until the 1960s, wear surface analysis has primarily relied on visualization
of wear damage from optical microscopy of wear surfaces. Metallographic cross-
sections were sometimes prepared by cutting the sample obliquely to enhance the
vertical magnification relative to the horizontal one. Like in the case of focused
ion beam milling, which will be described in the later sections, it is necessary to
protect the surface from damage from metallographic sample preparation. This is
typically done by electroplating the surface with a metal of similar hardness. Using
this technique, Moore [4] resolved the features indicative of plastic deformation and
work hardening underneath the finely scratched, machined and lapped surfaces of
metals (see also Bowden and Tabor [5]).
The advent of scanning and transmission electron microscopes in the 1960s, pro-
vided a powerful tool to the tribologists with high resolution imaging for visualizing
the surface features, orders of magnitude higher than the optical microscopy [6].
Simultaneously, elemental and chemical analyses of the wear surfaces were made
possible by X-ray and electron spectroscopy [7], while electron diffraction in the
transmission electron microcopy enabled the determination of the crystal structures.
However, specimen preparation for cross-sectional analyses remained a challeng-
ing task. SEM examination of subsurfaces continued to rely on dicing followed
by conventional metallographic sample preparation techniques. Earlier research on
cross-sectional transmission electron microscopy (TEM) of wear surfaces did not
gain widespread application due to the difficulty in preparing the TEM samples by
conventional techniques, which involved core drilling or dicing by diamond saw,
1 Electron Microscopy and Microanalysis for Wear Surface … 5

grinding, electropolishing and dimpling [8, 9]. These steps are not only cumbersome
and time consuming, but also are inadequate in locating the substructures in specific
locations on the wear surface. This chapter will first highlight the application of
modern focused ion beam (FIB) microscopy techniques for visualization of friction-
induced deformation in the subsurfaces as well as the preparation of TEM samples
in site-specific locations on a wear surface. This will be followed by specific exam-
ples from our recent research on the application of electron backscatter diffraction
(EBSD) and high-resolution TEM for analyzing the friction and wear mechanisms
in nanocrystalline metals and transition metal dichalcogenides.

1.2 Focused Ion Beam Preparation and Scanning Electron

Microscopy

1.2.1 Focused Ion Beam (FIB) Preparation of Samples

The development of the combined focused ion beam/scanning electron microscope

platforms (referred to as FIB/SEM) is possibly one of the most important develop-
ments that has enabled an improved understanding of sub-surface microstructural
changes due to wear [11, 12]. The FIB/SEM at the simplest level is the combination
of a high-resolution imaging tool (SEM) with a precision machining tool (FIB). This
combination has turned out to be extremely powerful and has enabled the production
of site specific samples for SEM imaging and other analytical techniques, such as
transmission electron microscopy (TEM), surface analysis, micromechanical testing
and many other analytical methods.
There are now two main ion sources for FIB tools, the liquid metal ion source
(LMIS) and the plasma source. The most common source is a LMIS that typically
produces a Ga+ ion beam that is used to micromachine via sputtering of selected
regions of the sample. The LMIS equipped FIB allows small ion beams to be formed
that are very useful for the preparation of samples that are on the order of 20 μm by
about 10 μm in size. The plasma source ion column can deliver much higher beam
currents to the sample at the expense of the size of the area of the beam. Thus, plasma
sourced FIB columns are quite useful for preparing much larger samples (as large as
1 mm in width) than the LMIS FIB. The beam size of the plasma FIB is not as small
as those that can be produced by LMIS sourced FIB tools. Selection of the plasma
FIB over the LMIS sourced FIB depends on the end use. If larger cross sections of
surfaces are to be routinely made then the plasma FIB may be the best choice. If
TEM samples are more important, then the LMIS FIB may be the better choice. One
also needs to keep in mind that the LMIS produces a beam of Ga+ ions that may
react chemically with the sample. Plasma sources generally utilize inert gases (most
commonly Xe) so chemical reactivity is not a concern. FIB instrumentation has been
covered in detail elsewhere [11, 12].
6 J. R. Michael et al.

To properly utilize any FIB tool, it is important for the user to have some under-
standing of the interactions that occur when an energetic ion interacts with the atoms
in a sample. There are many events that occur when an energetic ion interacts with the
atoms in a solid, but for the case of SEM and TEM sample preparation and ion imag-
ing, we are mainly interested in sputtering (the removal of atoms from the surface of
the sample), secondary electron production and damage to the sample in terms of ion
implantation and loss of crystalline structure. Sputtering is the process that removes
atoms from the target. Secondary electron production is important as images formed
with secondary electrons induced by ions have some important advantages over sec-
ondary electron imaging. In principle, it is impossible to have an ion beam interact
with a sample without causing some form of surface damage during ion irradiation.
Understanding the interaction of ions with the target material is helpful in reducing
the amount of damage to acceptable or tolerable levels through appropriate sample
preparation techniques. Energetic ions interact with a target in many ways. Once
the ion enters the sample its path can be deflected by interactions with the atomic
nuclei and the electron charges. As the ion moves through the sample it has sufficient
energy to knock other atoms off their respective lattice positions. The target atoms
that are knocked off their atomic positions can have enough energy to knock other
target atoms off their atomic positions. This series of moving atoms and ions within
the sample is referred to as a collision cascade. Some of the atoms that have been
knocked from their atomic positions may reoccupy a lattice position or may end up
in interstitial sites. There can also be lattice sites that are not reoccupied by target
atoms and these are vacancies. Both interstitials and vacancies are considered damage
to the crystalline structure of the sample. Most of the time, the original beam ion
will end up coming to rest within the sample. This is termed ion implantation. Ion
implantation results in the detection of the ion beam species in the sample and thus
it is desired to minimize this as much as possible. Many of the collision cascades
will eventually reach the surface of the sample. When the ion leaves the sample, it
may have sufficient energy to knock an atom from the surface into the vacuum. This
process is called sputtering and results in a net loss of material from the sample,
usually from areas very close to the beam impact point. At the same time when the
ion is either entering or leaving the sample, secondary electrons are generated that
are useful for producing images of the sample surface scanned by the ion beam. It is
important to remember that scanning an energetic ion beam over the surface of the
sample will always result in some damage to the sample. The details of this process
have been covered in other texts [10, 12].
It is most common to use the FIB to remove material to produce site-specific
specimens via sputtering, but ions also produce secondary electrons that can be
collected and imaged just like secondary electrons in the SEM. Ions generate many
more secondary electrons than do electrons, thus ion induced secondary electron
imaging can provide quality low-noise images. In addition, ions tend to channel
more strongly than electrons and thus can produce striking grain contrast images. It
is important to remember that during ion imaging of a sample, the continuous process
of ion damage is occurring that will eventually (given sufficient time and ion dose)
alter the sample to the point that it may no longer represent the original material.
1 Electron Microscopy and Microanalysis for Wear Surface … 7

Fig. 1.2 Ion induced

secondary electron image of
a FIB prepared cross section
of a wear scar (100 gm load
for 1000 cycles) in Ni
collected at 30 kV with an
ion beam current of 47 pA.
The dark layer on top of the
Ni is a solid lubricant thin
film

Figure 1.2 is a FIB prepared cross section of a wear scar produced by unidirectional
sliding on electrodeposited Ni. This is an ion induced secondary electron (ISE) image
obtained with a Ga+ ion beam using an accelerating voltage of 30 kV and a beam
current of 47 pA. Note the high grain contrast produced by the ion beam and it is
apparent that the stylus had moved from the right to the left-hand side of the image.
The excellent grain contrast provided by ISE imaging provides useful insights into
the wear process.
One common application of FIB is the preparation of cross sections of specific
features. An example, as shown in Fig. 1.2, is that of a cross section normal to a wear
surface that exposes sub-surface microstructural details. Cross sections prepared by
FIB can vary from a few 10’s of μm wide by a maximum of about 10 μm deep due
to the time required to prepare the cross section. However, new developments in the
use of plasma ion sources that can deliver much higher beam currents and thus faster
milling rates have allowed sections to be milled as large as 1 mm wide by 10’s of
μm deep.
This text is not intended to describe in detail all of the steps required to produce
quality cross sections as there are many texts already available that do this [11–13].
Cross sections of surfaces are accomplished in the FIB by careful control of the ion
beam current. The first step of the process is to deposit a protective layer over the
area of interest using electron or ion assisted deposition. Electron or ion assisted
deposition is a process where a precursor gas is injected into the FIB through a
needle that is very close to the sample surface. The electrons or ions in the beam
disassociate the precursor resulting in a very impure Pt layer or a C layer to be formed
only where the ion or electron beams are scanned during gas injections. The purpose
of this protective layer is to prevent the energetic ions from interacting with the area
of interest on the sample and causing damage as discussed previously. Once the
8 J. R. Michael et al.

protective layer is applied, the ion beam is then used to remove material to expose a
cross section of the area of interest. As the process continues lower ion beam currents
are used to allow more precise ion milling of the area and to produce very smooth
surfaces that are representative of the actual sample.
Another common application of FIB is the production of thin samples from spe-
cific sites that are suitable for transmission experiments in the transmission electron
microscope (TEM) or in the SEM (scanning transmission electron microscopy in the
SEM also called STEM in SEM or just STEM.). One other application that requires
thin samples is transmission Kikuchi diffraction (TKD) [14, 15]. This technique is
done in the SEM and provides high resolution (2–10 nm) orientation mapping of thin
samples. No matter which application STEM or TKD is to be used, sample prepa-
ration requirements are very similar. First, the sample must be free (or as close as
possible) of artifacts induced by the FIB milling process. Second, the samples must
be very thin, usually less than 100 nm in thickness to allow the electrons to pene-
trate the sample. Modern FIB milling techniques have been developed that reduce
or eliminate artifacts and allow extremely thin samples to be prepared and mounted
so that the sample can be introduced into the STEM or SEM for imaging. This text
is not intended to discuss all the methods of sample preparation as there are many
complete texts on this subject [11, 12]. A brief discussion of the steps to produce
thin samples with the FIB will be given.
Just as with cross sections made for SEM imaging the first step is to locate the area
of interest and then protect this area with a layer of Pt or C deposited by either ion
beam assisted or electron beam assisted deposition. Once this is done then trenches
are placed on either side of the area of interest as the final thin sample will come
from the material left in between these trenches. The sample thickness is carefully
reduced by using decreasing beam current ion beams in the FIB until the desired
thickness has been reached. At some point during the process (usually before the
sample is at the final thickness) the ion beam is used to cut the sample free from
the surrounding material so that it can be lifted out from the trench and placed on
a suitable support for subsequent imaging. It is often advantageous to use a lower
energy ion beam as a final step to remove the ion beam damage caused during
the sample milling at higher energies. This step is extremely important when high
resolution imaging in the TEM or STEM is to be done. Figure 1.3 is an example of a
completed thin sample prepared by FIB milling. Here the sample has been thinned to
the final dimensions and is ready to be manipulated to a suitable support for imaging.
The sample is manipulated outside the FIB tool to the support and this is called ex
situ lift-out [16]. Figure 1.4 shows the completed TEM specimen after manipulation
to a carbon-coated TEM grid.
Thin samples may also be produced by in situ lift-out in the FIB system using
an internal manipulator [17]. The general procedure adds a couple of steps to those
described above. The first step is to excise a chunk of material from the bulk sample
using the ion beam and the in situ manipulator. The chunk of material is then attached
to a suitable grid or support structure and held in place with ion beam assisted
deposition of Pt. The sample is then thinned by the removal of material from both
sides of the sample until the desired thickness is reached. The advantage to in situ
1 Electron Microscopy and Microanalysis for Wear Surface … 9

Fig. 1.3 Thin sample

prepared by FIB that is ready
for manipulation onto a
suitable support for
transmission imaging

Fig. 1.4 FIB prepared thin

sample manipulated onto a
thin carbon film support

lift-out samples is that there is no support film to interfere with subsequent analysis
and the sample is now firmly attached to the support structure and may be handled
more easily.
The study of the evolution of sub-surface microstructures during wear requires
that site-specific samples be produced. FIB tools can provide site-specific samples
that are suitable for SEM imaging and analysis, TEM and STEM analysis. Many
different types of samples can be produced in this manner that enhance the study of
wear.
10 J. R. Michael et al.

1.2.2 Electron Backscatter Diffraction and Transmission

Kikuchi Diffraction

An understanding of the microstructure and crystallography of wear scars is required

to fully describe the structure property relationships that control the tribological
response of crystalline materials. By linking the microstructure to the crystallography
of the sample, a full picture of the sample can be developed. Electron backscattering
diffraction (EBSD) and the related technique of transmission Kikuchi diffraction
(TKD) provide new insights into the microstructural changes that occur beneath a
wear scar.
EBSD is accomplished in the SEM by the addition of a suitable camera to image
the diffraction patterns and software to analyze the patterns and determine the crys-
tallographic orientation of the sample. Most often, EBSD is conducted on polished
bulk samples or bulk samples prepared in the FIB tool. Improved resolution over
EBSD of bulk samples has been achieved by utilizing thin samples and TKD. TKD
allows many of the fine details in wear microstructures to be observed. Both EBSD
and TKD will be introduced in this section, and the reader is directed to many of the
excellent texts available for more details [12, 18, 19].
EBSD patterns are obtained by illuminating a tilted sample surface with a sta-
tionary electron beam. The beam electrons interact with the sample and are initially
inelastically scattered. The scattered electrons are diffracted by the crystalline planes
in the sample to form the EBSD pattern that is then observed using a suitable camera.
In the case of TKD, the thin sample is held normal or near-normal to the electron
beam and the electrons transmit through the sample and are diffracted to form the
TKD pattern. Note that EBSD and TKD patterns appear visually to be quite similar
and can be indexed in the same manner (Fig. 1.5).
The EBSD pattern consists of sets of parallel lines. These are Kikuchi bands
and the distance between the parallel pairs of lines is inversely related to the lattice
spacing of the specific direction and the angles between the parallel sets of lines are
the angles between the reflecting planes in the crystal. Thus, the angles between the
planes can be compared to a table of angles prepared from known crystal structure
data to determine the orientation of the patterns obtained from each pixel in an image.
EBSD or TKD orientation imaging is conducted by arranging the EBSD or TKD
camera and the sample geometry in the SEM. The sample must be prepared so
that the surface is free of preparation induced artifacts. Typically with bulk samples
metallographic sample preparation is adequate and for TKD either FIB prepared
surfaces or electropolished surfaces must be used. Figure 1.6 shows the arrangement
of the sample and the EBSD or TKD cameras in the SEM. Note that a standard
EBSD camera may also be used for TKD. The electron beam is scanned pixel-by-
pixel over the areas of interest and patterns are collected at each pixel. The patterns
are indexed and the orientation of the crystal is determined for each pixel. This
procedure is repeated at each pixel within the image. Modern EBSD systems can now
do this reliably at over 3000 patterns/sec allowing large areas to be efficiently imaged.
Figure 1.7 shows inverse pole figure maps (IPF) of a wear scar made with a 10 g load
1 Electron Microscopy and Microanalysis for Wear Surface … 11

Fig. 1.5 EBSD pattern

obtained from Mo at 20 kV

on a <111> single crystal of Ni. The wear scar was made in the <211> direction. The
sample for these maps was made by FIB prepared thick lift-outs from the center of
the wear scar. Figure 1.7a is the IPF map with respect to the surface normal and this
is the <111> direction as can be seen by comparing the color of the substrate (blue)
with the color key shown in Fig. 1.7c where the <111> directions on the stereogram
is colored blue. Figure 1.7b is colored with respect to the wear direction and one can
see that the <211> direction on the stereogram is between the blue of the <111> and
the red color of the <100>. Thus by inspection it is relatively easy to understand the
sample crystallography. In these images it is also apparent that the surface of the
sample has been transformed from a single crystal to a layer of finer grained Ni with
more random orientations.
An example of TKD of a wear scar in Ta is shown in Fig. 1.8. In this case a Co pin
was worn against a Ta counterface. Note that this was obtained from a thin sample
that was prepared for STEM imaging. The TKD maps were obtained using a 30 kV
electron beam energy and a step size of 10 nm. Figure 1.8a is formed by a measure
of the pattern’s sharpness at each pixel and then plotting this as a grey scale image.
The result accurately represents the microstructure of the sample. Figure 1.8b is an
orientation map with respect to the wear direction (horizontal). Note that the upper
part of Fig. 1.8b represents very fine grained wear debris that contains some very
small grains that are just resolved.
EBSD and TKD are important characterization tools for wear phenomena in
crystalline materials. TKD has the advantage of higher resolution while standard
EBSD can provide data over larger areas at a reduced resolution.
12 J. R. Michael et al.

Fig. 1.6 Arrangement of the

sample and the EBSD or
TKD detectors in the SEM. a
EBSD b TKD

1.3 Transmission Electron Microscopy and Microanalysis

Transmission electron microscopy (TEM) encompasses a few powerful techniques

for characterizing wear surfaces at length scales from 100 pm to tens of μm.
Since electrons interact strongly with matter we need to use high-energy electrons
(100–300 kV) and create thin specimens (~100 nm in transmission). Before the
advent of the FIB, the challenge was making a sample in the desired location and
an appropriate thickness in projection. As discussed above the FIB now allows for
routine creation of site-specific TEM samples which in many cases would be diffi-
cult or impossible to do by many using conventional sample preparation methods.
In this section we will discuss the basic imaging methods of TEM, diffraction, and
microanalysis and related data analysis methods.
In this section we give a general overview of TEM and associated techniques. The
reader is directed to the text by Williams and Carter [7] for more detail on TEM and to
Carter and Williams [20] for advanced discussion of TEM techniques including this
1 Electron Microscopy and Microanalysis for Wear Surface … 13

Fig. 1.7 Inverse pole figure maps obtained from wear scars on <111> single crystal Ni substrates.
a IPF with respect to sample normal, b IPF with respect to wear direction, c color key for IPF maps

section’s discussion of microanalysis (spectral imaging and data analysis methods).

The TEM is analogous to a light microscope. We have an electron instead of light
source, electromagnetic instead of glass lenses, condenser lenses which structure
the electron illumination on the specimen, an objective lens, projection lenses and
detectors to record the images. Since electrons interact strongly with matter we
need to use both high-energy electrons, typically 100–300 kV, and thin specimens
typically 100 nm thick or less in the projection direction. As with light microscopes
we can think of the resolving power in terms of the wavelength of the illuminating
light or in the case of TEMs the electrons which are 3.7 pm at 100 kV and 2 pm at
300 kV (at these energies the electrons are relativistic). One of these requirements,
electron energy, leads to additional characterization capabilities to TEMs allowing
us to analyze crystallography of our materials. In particular, the wavelength of the
electrons (which are relativistic and thus act as both particles and waves) is smaller
than the crystallographic plane spacings in materials which are typically 1 nm and
14 J. R. Michael et al.

Fig. 1.8 TKD of Ta worn with a Co pin collected with a 10 nm step size. a A band contrast image
formed by a measure of the collected patterns sharpness at each point in the image. b Orientation
map with respect to the wear direction. Note that the colors represent orientations as shown in
Fig. 1.7c

smaller, so that we get diffraction. Additionally, given the samples must be thin and
analyzed in transmission we can improve the resolution of microanalysis versus a
SEM. This will be discussed in the following sections.

1.3.1 Transmission Electron Microscopy

TEM is both a general term describing an instrument and technique, namely passing
electrons through a specimen, and a specific term referring to the use of broad electron
illumination on a specimen. In this sense, it is analogous to the basic visible-light
microscope. However, because electrons interact strongly with matter, the electrons
can be scattered or diffracted multiple times while passing through a specimen. This
can be used to great advantage for imaging defects like dislocations and stacking
faults. Additionally, twins, precipitates, cracks, and amorphous regions can be imaged
1 Electron Microscopy and Microanalysis for Wear Surface … 15

in the conventional TEM modes. The types of data typically acquired for this include
bright-field (BF) and dark-field (DF) images which are distinguished by the use of a
small objective aperture (placed in the back-focal plane of the objective lens where a
diffraction pattern is located) centered on the direct or a diffracted beam respectively.
In TEM (specific definition) we have broad illumination on the sample (created by
the condenser lenses). The sample sits within the objective lens which has two parts,
an upper lens and a lower lens. The upper objective lens is used for focusing the
illumination to a point, and this is important for scanning-TEM (STEM). The lower
objective lens is the image-forming lens. Depending on how the projector lenses are
set we can form an image of our specimen, or a diffraction pattern on the detectors.
The former setting is used to collect a BF, DF or other image. The latter setting
allows us to collect a diffraction pattern. Furthermore, we can place an aperture at a
conjugate image plane within the projector lens system. This aperture can be moved
to capture a small image region from the specimen. By then switching to diffraction
mode, we can see a selected-area diffraction (SAD) pattern. BF, DF and SAD are
powerful methods for characterizing and surveying the coarse-scale structure of a
specimen.
In order to take advantage of the resolving power of the TEM we must operate
in a different mode, high-resolution TEM (HR-TEM). In this operational mode, we
use a larger objective aperture which allows the phase of several diffracted beams to
interfere. In this way can resolve lattice planes from a specimen provided they are
oriented parallel to the electron beam and of a sufficient spacing to be resolved by the
microscope. Recent advances in electron optics, namely the commercial availability
of spherical aberration correctors, have allowed for routine resolution of sub-100 pm
lattice spacings [21]. HR-TEM is useful for imaging basal planes in graphite and
MoS2 especially when aligned parallel to the electron beam as in looking at the
cross-section of a wear surface with those solid lubricants.

1.3.2 Scanning Transmission Electron Microscopy

STEM is analogous to SEM in that we focus the electrons to a point on the specimen
and then scan it across our specimen. We then get contrast from the specimen with
one of several possible detectors. These detectors sit below the projector lenses at
a conjugate diffraction plane and the projector optics can be changed to magnify or
de-magnify the diffraction pattern on the various detectors. By reciprocity [7] we can
form images equivalent to conventional TEM BF and DF images in STEM mode. We
can also form images with electrons scattered out to higher angles. These are called
high-angle annular dark-field (HAADF) images and contrast from the specimen
(presuming constant sample thickness in projection) is proportional to average atomic
number. STEM is a versatile imaging mode for looking at both coarse structure (tens
of microns) and atomic or near-atomic resolution 0.3 nm (lattice in some materials is
resolvable) in a conventional STEM. As with TEM, spherical aberration correctors
[22] have been developed for STEMs and have pushed direct resolution to 0.08 nm
16 J. R. Michael et al.

or better. It’s important to provide a caveat for such resolution numbers, namely that
these numbers are for a crystalline specimen oriented with lattice planes parallel to
the electron beam and are dependent on electron channeling. For a specimen that is
amorphous or not in a special orientation, the beam will broaden as it passes through
lowering the actual resolving power.

1.3.3 Microanalysis

When a high-energy electron hits a specimen, it’s possible for electrons to be knocked
out of atoms in the specimen. As a result of the subsequent electronic relaxation, one
possibility is that X-rays will be emitted from the excited atoms. The energy of the
emitted X-rays is characteristic of the elements present in the specimen. Therefore
we can use our electron probe to perform X-ray microanalysis to understand the
local chemistry of our specimen with sensitivity (varies by element) to the entire
periodic table above Li. As with SEMs, we can add an X-ray detector to detect
the X-rays emitted from the specimen. In the SEM, the resolution of this signal is
complicated and dependent upon both the elements in the specimen and the energy
of the incident electrons. The volume from which X-rays are emitted is known as
the interaction volume. In some cases, this can be a micron or more. This provides
a distinct advantage in resolution for TEM specimens where the bulk of material
has been removed, leaving only a thin lamella of material from which X-rays can
be generated and subsequently detected. The result of this is that fewer X-rays are
generated compared with a typical SEM specimen. The benefit is that the X-rays
that are generated come from a region comparable to the electron probe plus the
amount that the beam broadens as it passes through the specimen. The electron beam
will broaden more in a high atomic number material than in a lower one of similar
thickness. Thinning the specimen can improve resolution but at the expense of less
material which can generate an X-ray signal. Modern analytical electron microscopes
(AEM) are STEMs with X-ray and other chemically sensitive detectors (e.g., electron
energy-loss spectrometers). To make up for the smaller number of X-rays generated
by thin samples, X-ray detectors with larger solid-angles of collection have been
developed so that 0.5 steradians to 1 steradian systems are commercially available
[23]. For comparison, older AEMs typically had solid-angles of collection of 0.06
steradians.
There are a number of ways we can interrogate our specimen with X-ray micro-
analysis. The simplest is to manually move the electron beam to a feature and collect a
single high-signal spectrum. This can be repeated on additional features but entails a
significant amount of subjectivity with respect to appropriate sampling. Did we sam-
ple enough regions? And what if regions with important chemical differences show
the same image contrast? Will we sample those? X-ray microanalysis has undergone
a number of advances in the past 25 or so years. The most important was the advent
of spectral imaging systems for commercial X-ray detectors on SEMs [24] and later
STEMs. A spectral image combines imaging with full spectrum acquisition. A spec-
1 Electron Microscopy and Microanalysis for Wear Surface … 17

tral image consists of a full X-ray spectrum at each pixel in an image. We then are
statistically sampling many different points in a regular array. The next most impor-
tant development was the development of the silicon-drift detector (SDD) [25]. In
its modern form [26, 27] the SDD has completely replaced the older liquid nitrogen
cooled Si–Li technology [28]. SDDs are cooled to −20 °C allowing for window-
less operation improving light-element sensitivity. They are also able to operate at
higher X-ray fluxes (over 1 MHz is possible but practical only for an SEM geome-
try) at near theoretical resolution. Since the cooling requirements are less stringent
compared with older detectors, novel geometries have been implemented [29, 30]
improving solid-angle of collection in the AEM by more than ten times [23].
Spectral imaging solved the problem of sampling but created a new one, namely
how to analyze all the data comprehensively. In the next section, we discuss conven-
tional and advanced data analysis methods.

1.3.4 Spectral Image Analysis

In order to understand the structure and chemistry of our specimen with an AEM, we
collect electron images and spectral images. We can make direct measurements from
the images (e.g., film thickness, crack length, phase fraction, etc.) but the spectral
images are more complex, consisting of tens of thousands to hundreds of thousands
of spectra (pixels) each with 2000 energy channels typically. The challenge is how
to comprehensively analyze the sheer quantity of data we can quickly generate on
a modern AEM. The spectral image lends itself to retrospective analysis in that we
can collect the data on the AEM and then analyze and re-analyze it offline.
Conventional analysis of a spectral image is an iterative process [20, 31] starting
with looking at the sum of all the spectra in the spectral image. We can then form
region-of-interest maps of the elements identified from that spectrum. From the
maps, we can identify regions in the image and look at sum spectra from identified
regions. This forms the basis to confirm the analytical results. In X-ray spectroscopy,
there are a number of artifacts in maps. These include Bremsstrahlung or breaking
radiation generated as electrons in the specimen are decelerated. The result is a
continuum X-ray background which is stronger for heavy elements than for light
elements. Looking only at a map for a light element like oxygen, in a specimen with
a heavy element like gold will illustrate the artifact. Regions with gold will produce
a large background signal. In the oxygen map, regions with gold but not oxygen will
show elevated intensity. This can be visualized by looking at spectra from the gold
regions and verifying that no oxygen is present. More importantly we can see that
oxygen is present by looking at spectra from the other distinct regions in the oxygen
map to verify the presence of oxygen. An additional artifact in X-ray spectroscopy
comes from the fact that different elements and have X-ray lines at the same energy.
For example, the silicon K line is at 1.739 kV while the tantalum M-α line is at
1.709 kV. The result is that the silicon map will show intensity everywhere there is
Ta. Also, depending on which tantalum line we use to form the Ta map we might see
18 J. R. Michael et al.

silicon where it is not. Using the iterative analysis approach, we can help solve such
pathological overlap problem cases.
Conventional analysis can be time consuming and subjective. We can also miss
important features due to the artifacts mentioned above as well as just miss a weak
but potentially important feature. For these reasons, we discuss the use of statistical
analysis methods to quickly and comprehensively analyze X-ray spectral images
with no preconceptions of what is or is not present.

1.3.5 Multivariate Statistical Analysis

Multivariate statistical analysis (MSA) encompasses a family of data analysis meth-

ods that are used to visualized data sets subject to various constraints. For X-ray
spectral images the number of variables is typically the number of spectral channels,
perhaps 1000–2000. Understanding how collections of these co-vary can be used to
understand the chemistry of a specimen. One can think of the channels containing a
particular spectral peak or families of peaks as co-varying when the element respon-
sible for the peaks is present in regions in the specimen. Similarly, elements found
together in the same amounts in regions of the specimen would produce channels
that co-vary.
Before we discuss different MSA methods, it’s important to discuss scaling the raw
X-ray spectral image data for non-uniform noise [32, 33] and estimating the rank of
the data. When we discuss X-ray spectroscopy we have X-ray counts where the noise
is proportional to the signal. If we fail to account for the noise in our MSA methods
then we risk fitting noise when it is greater in magnitude than weaker but chemically
relevant signals in the spectral image. If we have a channel in our data where the
intensity is 100 counts, the associated noise for that datum is the square root of 100
or ±10. Variance-based MSA methods would preferentially fit such a variation over
a chemical signal smaller than that. Noise scaling, for the purpose of the examples
described in this chapter, amounts to dividing the spectral image by the square root
of the mean spectrum in the spectral domain and the square root of the mean image
in the image domain [32, 33]. Once we have performed the noise normalization,
we can then calculate the eigenvalues from the scaled spectral image data. The
sorted eigenvalues on a semi-log plot are referred to as a scree plot, an example of
which is given in Fig. 1.9. Here we examine just the 64 largest eigenvalues. Sandia’s
Automated eXpert Spectral Image Analysis (AXSIA) software [34–36] performs an
automated rank estimation by fitting a straight line to eigenvalues indexed 25 through
64, a region assumed to represent the noise baseline. When an eigenvalue rises above
the fit to the line by greater than 25 σ it is deemed non-noise and usually represents
chemical information. The index of this eigenvalue is the estimated rank, or number
of components needed to describe the data minus the noise.
Following noise normalization and rank estimation we then factor the data into a
form we can readily interpret. The reader may be familiar with one MSA variant called
Principal Components Analysis (PCA) [37]. PCA seeks to serially maximize the
1 Electron Microscopy and Microanalysis for Wear Surface … 19

Fig. 1.9 Scree plot showing the results of an Eigenanalysis of a spectral image weighted for non-
uniform noise. The automatically determined rank for the model of the data is four meaning that
only four factors are needed to describe the chemical information in the spectral image consisting
of hundreds of thousands of pixels and thousands of spectral channels

variance in components which are mutually orthogonal. The first principal component
is the mean. The second principal component describes the second most variance
in the data while being orthogonal to the first and so on. The problem with PCA is
that the constraints (serial maximization of variance and orthogonality) result in non-
physical components consisting of a component image and corresponding spectral
shape which both consist of positive and negative intensities. For this reason we add
an additional assumption to the factor analysis that none of the component images or
spectral shapes has negative (non-physical) intensities. The non-negativity constraint
relaxes the orthogonality imposed by PCA such that the resulting factors now are
directly interpretable. AXSIA achieves this via application of a MSA technique called
multivariate curve resolution (MCR) implemented by an alternating least squares
(ALS) approach [36]. The rotational ambiguity, the fact that there are an infinite
number of solutions to fit the data equally well, is addressed by performing a Varimax
rotation [38] on the data to make it simple in either the spectral [34–36] or spatial
[39] domain. Simple implies high contrast or zeros in many parts of the components.
The end result for both spectral and spatially simple representations is a compact and
physically interpretable model for the data. The factors making up this model contain
spatial components and corresponding spectral shapes. The component images tell
us where and how much of the elements are present, and the spectral shapes tell us
the identities of the elements.

1.4 Tribology Case Studies

This section describes a few specific examples on the application of these novel
electron microscopy techniques in tribology and tribological coatings.
20 J. R. Michael et al.

1.4.1 Diamond-Like Carbon Coating Characterization

The most widespread application of FIB in tribology is in the characterization of

the coatings on miniature parts, specifically those on the sidewalls of microelec-
tromechanical systems (MEMS). With miniaturization, the surface interactions play
a predominant role in determining the performance and reliability, necessitating the
need for solid lubricating coatings. Most of the commercial coatings are deposited
by physical vapor deposition processes, which are inherently line-of-sight. FIB has
become an invaluable tool in the characterization of thin films and coatings on the
sidewalls of miniature parts and on buried interfaces. The example shown in Fig. 1.10
is from our earlier study on the application on diamond-like carbon on Ni Microsys-
tems part fabricated by LIGA (German acronym for lithography, electroforming, and
molding). The DLC coating was applied by a commercial plasma enhanced chemical
vapor deposition technique. TEM analysis was used to assure that the DLC coating
coverage on the sidewalls, where it is most needed, is dense, conformal and without
any interfacial defects [40].

1.4.2 Wear-Induced Recrystallization

The first reported application of EBSD for the study of wear-induced recrystallization
in metals was on electroplated Ni surfaces [41]. Figure 1.11a shows the location of
the FIB cut on the wear surface created by a Si3N4 ball at an initial Hertzian contact
stress of 315 MPa for 1000 cycles of sliding in unidirectional mode. The direction
of sliding was from left to right, as shown by the arrow. The electroplated Ni had a
predominantly <001> fiber texture. There also appears to be a fine-grained region near
the top, i.e. right underneath the wear track. Figure 1.11b is the orientation map of the
substructure underneath the wear track. The colors represent the orientations normal
to the sample surface based on the color key shown in the inset. This is an interesting
area as there is a region that has a <110> fiber texture (designated by green) in the
predominantly <001> fiber textured material. The thick and thin black lines represent
high- and low-angle grain boundaries respectively. Figure 1.11b clearly reveals two
characteristic zones, each with its own unique features which differ significantly
from the microstructure in the bulk undisturbed material. A few microns below the
wear track, the bending of columnar grains in the direction of sliding is observed,
which is referred to as “Zone 1”. As we approach the wear surface, a lineup of thin
black lines appears in the microstructure indicating the formation of substructures
within the deformed zone. Right underneath the wear track, the columnar structure
broke down into more equiaxed submicron-size grain structure, which is referred to
as “Zone 2”. The depth of this zone extends to 1–2 μm. Zone 1 and Zone 2 are also
referred to as “plastically deformed” and “highly deformed” zones respectively. It
is also interesting to note that Zone 2 is thicker in the <110> fiber textured region
than in the <001> fiber textured region. The pole figures corresponding to <001>
1 Electron Microscopy and Microanalysis for Wear Surface … 21

(b)

(a)

FIB Cut

(c)

Conductive coating for FIB

DLC Coating

Ni

Fig. 1.10 a SEM micrograph of a Diamond-like carbon coated Ni-based microsystems part fabri-
cated by LIGA, b FIB cut on the sidewall to prepare a TEM sample, c Bright-filed TEM image of
the FIB cut showing the DLC coating on the sidewall. Reproduced by permission of IEEE Xplore
[41, Figs. 3 and 8]

and <110> textured grains are shown in Fig. 1.11c. The spread in orientation of pole
figures (Fig. 1.11c) can be used to judge the extent of wear-induced deformation
in the subsurfaces. The <110> textured grains have wider orientation spread in the
sliding direction (18°–44°) than the <001> textured grains (10°–17°), which is in
agreement with the microstructural findings in Fig. 1.11b.
22 J. R. Michael et al.

(a)

(b) Sliding DirecƟon

Zo
Zo

(c)

Fig. 1.11 EBSD analysis on the subsurface of a wear scar on electroplated Ni. a Location of the FIB
cut on the wear scar; b orientation map with respect to the surface normal (the arrow represents the
sliding direction); the heavy black lines represent orientation changes >10° and thin lines represent
orientation changes of 1° or less, c pole figures of the region underneath the wear scar showing
<001> and <110> fiber textured material (sliding direction is Y0). Inset at the center of Figure c
is the stereographic triangle with color key for (b). Reproduced by permission of Pergamon Press
[40, Fig. 3]

This very first study on the application of EBSD for friction-induced wear phe-
nomena has demonstrated the unique role of focused ion beam techniques in prepar-
ing cross sections of narrow wear tracks generated under very light loads for elec-
tron backscatter diffraction studies. By suitably thinning the samples further, this
technique can be easily extended to prepare cross sections of wear tracks for TEM
1 Electron Microscopy and Microanalysis for Wear Surface … 23

analysis. Unlike in conventional specimen preparation techniques, the FIB enabled

specimens are free of artifacts introduced during dicing, grinding, electropolishing,
dimpling, etc., and sections can be precisely cut at a specific location on the wear
track.

1.4.3 Tribology of Nanocrystalline Metals

While the deformation behavior in metals with relatively large grain sizes is predom-
inantly dictated by dislocation dynamics, gain boundary processes dominate in the
nanocrystalline regimes, when the grain sizes are below 20 nm [42]. Analogous to
the grain size dependency on the mechanical behavior, recent studies seem to confirm
that friction behavior of nanocrystalline metals could be significantly altered when
the friction-induced deformation results in the formation of stable ultrananocrys-
talline structures in the subsurfaces. This was first highlighted by Prasad, Battaile
and Kotula [42] from their study on friction behavior of nanocrstalline nickel, which
exhibited two distinct friction behaviors, μ ~ 0.30–0.35 or μ ~ 0.6–0.7, depending
upon the contact stress and sliding speed (Fig. 1.12a).
Figure 1.12b is an annular dark field STEM image showing detailed features of
the subsurface for the case which exhibited low friction behavior (green circles in
Fig. 1.12a). The arrow indicates the direction of sliding. The image shows three
zones (labeled in Fig. 1.12b). The zones observed include an ultrafine grained region
directly at the wear surface (Zone 1), a region with large, textured, flattened grains
(Zone 2), and a gradual transition of grain size approaching that of the bulk material
(Zone 3). The bulk has an average grain size of 20–100 nm. During frictional contact
at stresses and sliding speeds corresponding to the green circles in Fig. 1.12a, ultrafine
nanocrystalline gains with a size of 2–10 nm formed right underneath the wear surface
(Zone 1). The authors attributed the experimentally observed transition from high
friction to low friction regime to the formation of this ultrafine nanocrystlline struc-
ture underneath the wear surface elucidated by FIB-TEM. The authors also reported
that in Zone 1 microstructure was absent in cases that did not exhibit a friction tran-
sition. Zone 2 consisted of platelet shaped grains larger than 100 nm, elongated in
the sliding direction, in between Zone 1 and the bulk microstructure of Zone 3. The
grains in Zone 2 represent a layer approximately 150 nm thick that showed signif-
icant grain growth and the development of crystallographic texture. Additionally,
the transition between Zones 2 and 3 is not sharp but rather graded as the grain size
returns to bulk dimensions over a distance of a few hundred nanometers. The study is
a clear demonstration of the application of modern electron microscopy techniques
for the fundamental understanding of the mechanisms of friction in nanocrystalline
metals.
24 J. R. Michael et al.

(a)

(b)

Wear Voids
debris
Zone 1

Zone 2

Zone 3 (transition/bulk)

100 nm

Fig. 1.12 a Friction transitions in nanocrystalline Ni [42]. The data was generated over a wide range
of contact pressures and sliding speeds. The green dots represent the low friction regime while the red
dots correspond to the high friction regime. The dotted line depicts the friction transition. b Annular
dark field STEM image of the subsurface frim the low friction regime, showing ultrananocrystlline
grain structure at the top. Reproduced by permission of Pergamon Press [42, Figs. 2 and 3b]
1 Electron Microscopy and Microanalysis for Wear Surface … 25

1.4.4 Transition Metal Dichalcogenides

Transition Metal Dichalcogenides (TMD), most notably, molybdenum disulfide

(MoS2 ) and tungsten disulfide (WS2 ) are well known for their solid lubricating behav-
ior and are widely used in several applications. MoS2 crystallizes in the hexagonal
structure where a sheet of molybdenum atoms is sandwiched between two hexag-
onally packed sulfur layers with a high c/a ratio (c  12.29 Å, a  3.16 Å). The
bonding within the S–Mo–S sandwich is covalent, while weak Van der Waals forces
hold the sandwich together resulting in interlamellar mechanical weakness. It has
been widely reported [43] that under a shearing force the basal planes slide over one
another by intracrystalline slip, similar to a ‘deck-of-cards’, resulting in the formation
of a transfer film on the rubbing counterface (e.g., the ball surface in a ball-on-disk
friction and wear test), shown schematically in Fig. 1.1. According to the widely
accepted hypotheses, the main mechanisms for imparting low interfacial shear in
TMD are: creation of (0002) basal planes and subsequent (re)orientation parallel to
the sliding direction, and the development of a transfer film on the counterface to
accommodate interfacial sliding. Until the advent of FIB, preparing TEM samples of
these thin layers transferred onto the ball surfaces from ball-on-disk tribology tests
of MoS2 films that are suitable for high resolution lattice imaging has not been very
successful. A recent study by Scharf, Kotula and Prasad [43] demonstrated the appli-
cation of FIB to prepare TEM samples of transfer films form a molybdenum disulfide
film doped with antimony trioxide and gold (MoS2 –Sb2 O3 –Au). This film, which has
been successfully deployed in several space mechanisms, lacks any long-range crys-
tallinity in the as-deposited condition. Using FIB-TEM, these authors have reported
the presence of a thin layer at the top of the cross-section of the wear surface that
has crystalline 2H–MoS2 basal (0002) planes aligned parallel to the sliding direc-
tion (Fig. 1.13b). The arrow indicates the sliding direction, while the region above
the arrow corresponds to the carbon film that was applied as part of the TEM FIB
sample preparation. Just below this crystalline 2H–MoS2 layer, the TEM image in
Fig. 1.13b shows the presence of ~4–6 nm Au particles that are larger than the 2–3 nm
size nanoparticles in the as-deposited coating seen in the bulk. This Au nanoparticle
coarsening is clearly a result of the sliding process. The AXSIA map in Fig. 1.13d
(where red is MoS2 /Sb2 O3 /Au and green is Au) also confirms Au particle coarsen-
ing due to frictional contact. A typical HRTEM image of a FIB cross-section of the
transfer film (Fig. 1.13a) showed that it was comprised of predominantly crystalline
MoS2 through its entire thickness of ~1 μm. However, in the top layer, the basal
planes are aligned parallel to the sliding direction. This layer is about 6 nm thick,
which approximates to nine basal planes of MoS2 . A SAD pattern obtained from
this top most region of the cross-section (Fig. 1.13c) showed that the interlamellar
26 J. R. Michael et al.

(a)
5 nm

MoS c-axis ┴(0002)

2

Transfer film on ball

(c)
Wear
(0002)
MoS c-axis ┴(0002)
2 Au
nanoparƟcle
clustering

Amorphous
MoS /Sb O /Au
2 2 3

(b)
(d) 20

Fig. 1.13 a and b Cross-sectional HRTEM montage images of wear track on MoS2 /Sb2 O3 /Au
solid lubricant film with its transfer film on the ball surface. c SAED pattern of transfer film with
interlamellar d-spacing between the (0002) 2H2 –MoS2 planes measured at ~6 Å, which compares
well with indexed 6.1 Å lattice spacing of MoS2. d AXSIA results of the wear surface where red is
MoS2 /Sb2 O3 /Au and green is Au (with some Sb/O). Reproduced by permission of Pergamon press
[43, Fig. 9]

d-spacing between the (0002) 2H–MoS2 planes was ~6 Å, which compared well with
indexed (PDF#01-087-2416) 6.1 Å lattice spacing for MoS2 lattice. Once again, the
application of modern electron microscopy and microanalysis techniques has been
able to experimentally validate the widely held concept of self-mated ‘basal plane-
on-basal plane’ sliding as the fundamental mechanism of lubrication in transition
metal dichalcogenide thin films.
1 Electron Microscopy and Microanalysis for Wear Surface … 27

Acknowledgements Sandia National Laboratories is a multi-mission laboratory managed and

operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned
subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear
Security Administration under contract DE-NA0003525. The authors gratefully acknowledge Pro-
fessor Thomas W. Scharf, University of North Texas, for many years of collaboration with the
authors on this subject, and Dr. Corbett Battaile for critically reviewing the manuscript. Special
thanks to Michael Rye for the FIB work and to Bonnie McKenzie for the SEM, EBSD and TKD
studies.

References

1. D.A. Rigney, J.P. Hirth, Plastic-deformation and sliding friction of metals. Wear 53, 345–370
(1979)
2. S.V. Prasad, Wear, in Concise Encyclopedia of Advanced Ceramic Materials, ed. by R.J. Brook
(Pergamon Press, Oxford, 1991), pp. 511–519
3. S.V. Prasad, T.W. Scharf, J. Mater. Sci. 48, 511–531 (2013)
4. A.J.W. Moore, Proc. Roy. Soc. A 195, 231 (1948)
5. F.P. Bowden, D. Tabor D, The Friction and Lubrication of Solids (Clarendon Press, Oxford,
1986), pp. 112, 120
6. R.F. Pease, W.C. Nixon, J. Sci. Instrum. 42, 81–85 (1965)
7. D.B. Williams, C.B. Carter, Transmission electron microscopy: A Textbook for Materials Sci-
ence, 2nd edn. (Springer, 2013) ISBN-13: 978–0387765006
8. P. Heilmann, D.A. Rigney, Metall. Trans. 12A, 686 (1981)
9. W.M. Rainforth, Wear 245, 162 (2000)
10. M. Nastasi, J. Mayer, J.K. Hirvonen, Ion-Solid Interactions: Fundamentals and Applications
(Cambridge University Press, New York, 1996)
11. L. Giannuzzi, Introduction to Focused Ion Beams: Instrumentation, Theory, Techniques and
Practice (Springer, New York, 2006)
12. J.I. Goldstein, D.E. Newbury, J.R. Michael, N.W.M. Ritchie, J.H. Scott, D.C. Joy, Scanning
Electron Microscopy and X-Ray Microanalysis (Springer, New York, 2017)
13. J. Mayer, L.A. Giannuzzi, T. Kamino, J. Michael, TEM sample preparation and FIB-induced
damage. MRS Bull. 32(5), 400–407 (2007)
14. P.W. Trimby, Orientation mapping of nanostructured materials using transmission Kikuchi
diffraction in the scanning electron microscope. Ultramicroscopy 16–24 (2012)
15. R.R. Keller, R.H. Geiss, Transmission EBSD from 10 nm domains in a scanning electron
microscope. J. Microsc. 245–251 (2012)
16. L.A. Giannuzzi, Z. Yu, D. Yin, M.P. Harmer, Q. Xu, N.S. Smith, L. Chan, J. Hiller, D. Hess, T.
Clark, Theory and new applications of ex situ lift out. Microsc. Microanal. 21(4), 1034–1048
(2015)
17. N. Bassim, K. Scott, L.A. Giannuzzi, Recent advances in focused ion beam technology and
applications. MRS Bull. 39(4), 317–325 (2014)
18. V. Randle, Microtexture Determination and its Applications (Maney Publishing, London, 2003)
19. A.J. Schwartz, M. Kumar, B.L. Adams, D.P. Field (eds.), Electron Backscatter Diffraction in
Materials Science (Springer, New York, 2009)
20. C.B. Carter, D.B. Williams (eds.), Transmission Electron Microscopy (Springer International
Publishing Switzerland, 2016), Chapter 16, https://doi.org/10.1007/978-3-319-26651-0_16
21. M. Haider, S. Uhlemann, E. Schwan, H. Rose, B. Kabius, K. Urban, Electron microscopy
image enhanced. Nature 392, 768–769 (1998). https://doi.org/10.1038/33823
22. P.E. Batson, N. Dellby, O.L. Krivanek, Sub-ångstrom resolution using aberration corrected
electron optics. Nature 418, 617–620 (2002). https://doi.org/10.1038/nature00972
28 J. R. Michael et al.

23. H.S. von Harrach, P. Dona, B. Freitag, H. Soltau, A. Niculae, M. Rohde, J. Phys.: Conf. Ser.
241, 012015 (2010)
24. R.B. Mott, C.G. Waldman, R. Batcheler, J.J. Friel, Position tagged spectrometry: A new
approach for EDS spectrum imaging. in Proceeding of Microscopy Microanalysis, ed. by G.W.
Bailey, M.H. Ellisman, R.A. Hennigar, N.J. Zaluzec. Jones and Begell Publishing, New York,
pp. 592–593
25. E. Gatti, P. Rehak, Nuc. Inst. Meth. Phys. Res. 225, 608–614 (1984)
26. L. Strüder, N. Meidinger, D. Stotter, J. Kemmer, P. Lechner, P. Leutenegger, H. Soltau, F. Eggert,
M. Rohde, T. Schulein, High-resolution X-ray spectroscopy at close to room temperature.
Microsc. Microanal. 4, 622–631 (1998)
27. A. Niculae, P. Lechner, H. Soltau, G. Lutz, L. Strüder, C. Fiorini, A. Longoni, Optimized
readout methods of silicon drift detectors for high-resolution X-ray spectroscopy. Nucl. Instrum.
Methods. Phys. Res. A 568, 336–342 (2006)
28. R. Fitzgerald, K. Keil, K.F.J. Heinrich, Science 159, 528–530 (1968)
29. B.L. Doyle, D.S. Walsh, P.G. Kotula, P. Rossi, T. Schulein, M. Rohde, X-Ray Spectrom. 34,
279–284 (2005)
30. P.G. Kotula, J.R. Michael, M. Rohde, Microsc. Microanal. Suppl. 2(14), 116–117 (2008)
31. P.G. Kotula, M.R. Keenan, Application of multivariate statistical analysis to STEM X-ray
spectral images: interfacial analysis in microelectronics. Microsc. Microanal. 12(6), 538–544
(2006)
32. M.R. Keenan, P.G. Kotula, Accounting noise in the multivariate statistical analysis of TOF-
SIMS data. Surf. Interf. Anal. 36, 203–212 (2004)
33. M.R. Keenan, P.G. Kotula, Optimal scaling of ToF-SIMS spectrum-images prior to multivariate
statistical analysis. Appl. Surf. Sci. 231–232, 240–244 (2004)
34. M.R. Keenan, P.G. Kotula, Apparatus and System for Multivariate Spectral Analysis. US Patent
# 6,584,413, 2003
35. M.R. Keenan, P.G. Kotula, Method of Multivariate Spectral Analysis. US Patent # 6,675,106,
2004
36. M.R. Keenan, Multivariate analysis of spectral images composed of count data. in Techniques
and Applications of Hyperspectral Image Analysis, ed. by H. Grahn, P. Geladi (Wiley & Sons,
Chinchester, 2007)
37. I.T. Jollife, Principal Component Analysis, 2nd edn. (Springer, New York, 2002)
38. H.F. Kaiser, The varimax criterion for analytic rotation in factor analysis. Psychometrika 23(3),
187–200 (1958)
39. M.R. Keenan, Exploiting spatial-domain simplicity in spectral image analysis. Surf. Int. Anal.
41, 79–87 (2009)
40. S.V. Prasad, J.R. Michael, T.R. Christenson, EBSD studies on wear-induced subsurface regions
in LIGA nickel. Scripta Mater. 48, 255–260 (2003)
41. S.V. Prasad, T.W. Scharf, P.G. Kotula, J.R. Michael, T.R. Christenson, J Microelectromech.
Syst. 18, 695 (2009)
42. S.V. Prasad, C.C. Battaile, P.G. Kotula, Scripta Mater. 64, 729–732 (2011)
43. T.W. Scharf, P.G. Kotula, S.V. Prasad, Acta Materiala 58, 4100–4109 (2010)
Chapter 2
Analyzing Mild- and Ultra-Mild Sliding
Wear of Metallic Materials
by Transmission Electron Microscopy

Alfons Fischer, Wlodzimierz Dudzinski, Birgit Gleising and Priska Stemmer

Abstract Any understanding of tribological behavior is connected to sound analy-

ses of the wear appearances, which render insight into the acting wear mechanisms
and their sub-mechanisms. Today one important method to analyze wear appear-
ances at high resolution is transmission electron microscopy (TEM) being invented
by Knoll and Ruska in the early 1930th in Berlin (Knoll in Z fuer Phys 78:318–339,
1932 [1]).

Abbreviations and Terms

AES Auger atom emission spectroscopy

AFM Atomic force microscopy
ECC Electron channeling contrast
EDS Energy dispersive X-ray spectroscopy
EELS Electron energy loss spectroscopy
EFTEM Energy filtered transmission electron microscopy
FIB Focussed ion beam
KAM Kernel average missorientation

A. Fischer (B) · P. Stemmer

Materials Science and Engineering, University of Duisburg-Essen, Duisburg, Germany
e-mail: alfons.fischer@uni-due.de
P. Stemmer
e-mail: priska.stemmer@uni-due.de
W. Dudzinski
Faculty of Technology and Engineering, Wroclaw University of Science and Technology,
Wroclaw, Poland
e-mail: wlodzimierz.dudzinski@pwr.edu.pl
A. Fischer · B. Gleising · P. Stemmer
CeNIDE-ICAN, Interdisciplinary Center for Analytics on the Nanoscale, University of
Duisburg-Essen, Duisburg, Germany
e-mail: birgit.gleising@uni-due.de

© Springer Nature Switzerland AG 2018 29

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_2
30 A. Fischer et al.

MD Molecular dynamics computer simulation

RS Raman-spectroscopy
SEM Scanning electron microscopy
TEM Transmission electron microscopy
BF Bright field image
DF Dark field image
DP Diffraction pattern
XPS X-ray photoelectron spectroscopy
nc Nano-crystalline <100 nm
ufc Ultrafine-crystalline 100–500 nm
μc Micro-crystalline 500 nm

2.1 Introduction

Any understanding of tribological behavior is connected to sound analyses of the

wear appearances, which render insight into the acting wear mechanisms and their
sub-mechanisms. Today one important method to analyze wear appearances at high
resolution is transmission electron microscopy (TEM) being invented by Knoll and
Ruska in the early 1930th in Berlin [1]. The history of TEM and tribology started
about in the 1950 by work about solid lubrication [2]. Certainly wear and its mech-
anisms were a crucial part of research already at that time [3], therefore, TEM was
introduced into it in the 1960s by Bill Glaeser et al. [4] and wear mechanisms, sub-
mechanisms and the underlying elementary processes have been investigated for a
long time since then. Investigating metals at the desired small wear rates required
high resolution analyses at and just below the worn surfaces either because of the
materials properties of interest and their microstructural alterations under tribolog-
ical loading [5–8] or under certain processing parameters [9]. Last-but-not least it
is crucial to understand the influence of the interfacial media like lubricants [10]
and wear debris [11] as well as any causal materials science aspects or elementary
processes [12–14]. Thus the number of papers on TEM and wear related phenom-
ena as well as conference proceedings, letters and short notices increased steadily
(Fig. 2.1).
On the basis of the early investigations friction, wear, and lubrication were seen as
interconnected parts of an area being named tribology in 1966. It became clear that
any property is not just related to a certain material but it is connected to the entire
tribological system and its specific structure and acting wear mechanisms (Fig. 2.2)
[15, 16]. This somewhat rough scheme—and even though not all sub-mechanisms
are known or have yet been sufficiently investigated—has been proven to allow for
targeted countermeasures beyond just trial-and-error (or the common hardness vs.
wear) approaches [17]. Due to the fact that in the current English literature the mech-
anisms and sub-mechanisms are described or designated by similar terms we will use
quotes for the mechanisms and sub-mechanisms. In general we distinguish between
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 31

Table 2.1 Main wear mechanisms and their submechanisms

Main wear Tribochemical Surface fatigue Abrasion [18] Adhesion [19]
mechanisms reactions
Submechanisms Tribooxidation Delamination (by Microploughing Material transfer
[20] predominantly
elastic interaction
in solid contacts
[21])
Microcutting
Tribocorrosion Microcracking
[22]
Mechanical Indentation (by Microfatigue
mixing [24] predominantly
plastic interaction
in solid contacts
[23])

four main wear mechanisms: “Abrasion”, “Adhesion”, “Surface Fatigue” and “Tri-
bochemical Reactions” [18] which are related to and can be defined by certain wear
appearances. Still depending on the structure of the tribosystem and the properties of
the involved materials there are alterations of the surfaces as to topography, chem-
istry, and microstructure as well as changes of the interfacial and surrounding media.
This did lead to the definition of so-called sub-mechanisms. Table 2.1 sums about up
the current state of sub-mechanisms known to the authors from references of the last
50 years. The references given in Table 2.1 are far from being complete and should
just point on some substantial contributions to this approach.
It also has been shown by [18] and others that “Adhesion” might lead to wear
rates that are nine or more orders of magnitude larger than those caused by “Tri-
bochemical Reactions”. At the desired very small wear rates in the range of less

Fig. 2.1 Number of

reviewed papers on
transmission electron
microscopy and wear
(Scopus Search on Oct. 10th
2017)
32 A. Fischer et al.

Fig. 2.2 The main wear mechanisms, known related submechanisms and examples for primary
elementary mechanisms and processes

than e.g. <3 nm/h1 one would in general expect sub-mechanisms of “Tribochemi-
cal Reactions” and “Surface Fatigue”. The authors exemplified that a combination
of “Mechanical Mixing” and “Indentation” a sub-mechanisms of “Surface Fatigue”
characterized by predominantly plastic interaction generated by rotating sub-μm
size wear particles may lead to relatively small wear rates [17, 23]. This is the more
pronounced the better the mechanically mixed (or in other systems tribooxidized)
layer at the surface is supported by the subsurface microstructure [25]. Another pre-
requisite is that such layer must not be overloaded by too large contact stresses [26].
Thus if “Indentation” as well as frictional contact shear forces trigger “Mechanical
Mixing” and, therefore, generate a tribofilm and nano-size wear particles the total
wear rates might become very small and even reach the so-called ultra-mild wear
regime [26–28].
In order to properly analyze such near- or sub-surface structures one has to dis-
tinguish between either amorphous, nanocrystalline and/or chaotic surface layer and
the underlying strain gradient. The latter is also called “transition zone” (here: sub-
surface or strain gradient) in many papers and is generated by the accumulation of
cyclic plastic strains [25]. In contrast the uppermost layer (here: near-surface or tri-

1 This number is a criterion for piston ring wear that would allow for a minimum of 200,000 km
of a passenger car engine without substantial wear. It is chosen absolutely voluntarily in order
to distinguish between ultra-mild wear and mild wear. While most laboratory tribometers for time
reasons run under mild-wear conditions. Still most parts in application would require ultra-mild wear
for a sufficient life time. In sliding wear mild and severe can be distinguished by the characteristics
of contact mechanics; e.g. the plasticity index (s.a. Johnson (1985) Contact Mechanics, Cambridge
University Press, Cambridge, England). If the contact is predominantly elastic it might lead to mild
wear, while a predominantly plastic interaction brings about severe wear. Still all these terms are
not standardized and the regimes might overlap.
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 33

Fig. 2.3 Scheme of surface and subsurface structures generated under tribological stresses based
on [16, 30, 31]

bomaterial) appears like a shear band [13] caused by monotonic stresses and strains
similar to what is known from severe plastic deformation [29] (Fig. 2.3).
Despite all of the research of the last 70 years any direct relation or even quan-
titative connection to the so-called elementary processes (Fig. 2.2) is still missing.
Hence further research is necessary comprising low and high resolution analyses in
combination with the appropriate computer simulation method.
In this section we cannot give an overview over all the work that has been done
by TEM in tribology. We would just like to show how one can get information on
tribological characteristics of the elementary processes and mechanisms by TEM
analyses and—at least—qualitatively relate it to the tribological behavior of exem-
plary selected metallic materials.

2.1.1 TEM Analytics of the Cyclic Strain Gradient

Accumulation of cyclic strains are known from fatigue of metals (also called cyclic-
creep or ratchetting), but here the stress state is multiaxial. Hence no direct quanti-
tative relation between classical fatigue and wear is possible, but elementary mech-
anisms and processes are similar like e.g. phase transformations or the influence of
long- and short-range order effects. Finnie et al. pointed already in 1984 on the fact
that the strain gradient is generated by accumulated cyclic strains [32] they found
the conflicting requirement of high cyclic strength in combination with high ductil-
ity (ability to accumulate cyclic strains without crack nucleation) for wear resistant
metals. In fatigue and, therefore, as in wear this is governed by the characteristics of
either planar or wavy cyclic slip of metals. On the basis of TEM analyses of MgO
and pure Ni, Rigney and Glaeser did also show earlier that such differences in slip
behavior might govern sliding wear [5]. Glaeser pointed on such mechanisms already
in 1977 by investigating wear particles detached from copper showing dislocation
cells [33]. He hypothesized that crack nucleation must have taken place at such cell
34 A. Fischer et al.

walls. In 1983 Saleski and Ritchie proposed a detailed model for plain steel showing
that cracks are initiated at dislocation cells below a sliding surface, which then led
to the formation of wear particles [34]. These early papers already depict towards
the necessity to analyze the strain gradient under worn surfaces by TEM in order to
understand the nature of cyclic sliding mechanisms under such multiaxial stress and
strain fields.

2.1.1.1 The Strain Gradient in Austenitic Materials

By comparing Austenitic Materials with known planar, wavy, and mixed sliding
characteristics Fischer et al. were able to show, that planar sliding improves the wear
behavior distinctly (Table 2.2) [25].
Due to the fact that under planar sliding the lattice defects (dislocations, twins,
stacking faults, ε-martensite, Fig. 2.4a) stay mobile larger cyclic strains can be accu-
mulated over a bigger number of cycles. Thus crack initiation as well as propaga-
tion are hindered. By wavy slip cells are formed blocking dislocations (cell walls,
Fig. 2.4b) which bring about subsurface crack initiation. It is also known that wavy
slip promotes faster crack propagation [35]. If wear particles are generated by crack
initiation and propagation inside the strain gradient their size is likely within the
μm-range. Such elementary process can be attributed to “surface fatigue” and its
submechanism “delamination”. As a results μm-large wear particles deteriorate
the contact situation e.g. by bringing about “microploughing” and destroying also
any existing nanocrystalline tribomaterial (Fig. 2.5) or by “microcutting” generating
instantly further μm-size chips and increase the wear rate (Table 2.2).
Table 2.2 also shows by the wear rates—as has been reported also by others
[11, 36, 37] earlier—that the stacking fault energy as well as strain induced phase
transformations play an important role. From fatigue tests it is known that short-
range order effects are as important for the ability to accumulate cyclic strains as a
so-called low damage accumulation rate [38].

Table 2.2 Sliding characteristics of austenitic materials and their wear rates
Material Sliding characteristics Wear rate in m/m under dry
sliding after 90,000 cycles
X5CrNiMo17-13-2, Wavy 2 × 10−8
ISO5832-1
X6CrNiMoN22-10-4-3, Mixed 1.8 × 10−8
ISO5832-9
X13CrMnMoN18-14-3, Planar 7 × 10−10
ASTM F2229-02
CoCr29Mo6, ISO5832-6 Planar, very low stacking fault 2 × 10−10
energy
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 35

(a) Planar Sliding Structure of an Austenit- (b) Wavy Sliding Structure of an Austenitic
ic High Nitrogen CrMnMoN-Steel with Dis- CrNiMoN-Steel with Dislocations Cells and
locations, Twins, Stacking Faults and ε- Twins
Martensite

Fig. 2.4 TEM micrographs of defect structures within the subsurface strain gradient after dry
sliding wear. a Planar and b Wavy sliding austenitic steels from [25]

(a) Austenitic High Nitrogen CrMnMoN- (b) Austenitic CrNiMoN-Steel

Steel

Fig. 2.5 TEM micrographs of nanocrystalline tribomaterial at the worn surfaces of a planar and
b wavy sliding austenitic steels from [25]. The arrows are perpendicular to the contact surfaces
marked by the dashed lines. It should be mentioned here that even though such tribomaterial is
nanocrystalline it may show traces of a directional texture as can be seen by the oriented brightness
of the diffraction pattern in this figure

2.1.1.2 Strain Gradient in Martensitic Materials

Now most steels used in mechanical engineering are not austenitic but ferritic or
martensitic and are known for wavy slip as well. But in contrast to austenitic metals
their strength is much higher, while for martensites the ductility is much smaller.
Thus it is interesting to understand whether high strength martensitic steels would
generate a strain gradient by cyclic multiaxial stresses as well. Figure 2.6 shows
36 A. Fischer et al.

Fig. 2.6 TEM bright field image and diffraction pattern of a strain gradient of a cold work tool
steel 56NiCrMoV7 (AISI L6) under boundary lubricated sliding wear at a normal load of 150 N
from [39]. The diffraction pattern was rotated in order to parallelize its directional brightness to the
lattice orientation of the crystals inside the red circle

an overview of the subsurface region of a worn cold work tool steel depicting a
zone of distinct subsurface grain refinement parallel to the surface. The red circle
focusses into such refined zone with a grain size of about 200 nm, while the bulk
depicts martensite lathes of some μm. It is worth mentioning that within the strain
gradient the grain size stays above 100 nm also known as the ultra-fine (100–500 nm)
crystalline range.
Similar outcomes can also be found with carburized martensitic steels after 2 ×
106 cycles of ultra-mild sliding wear under boundary lubrication in gear oil at 80 °C
[40]. The bulk within the carburized zone shows μm-large martensite lathes while
the strain gradient depicts again smaller grains as can be derived from the streaks
around the diffraction reflexes (Fig. 2.7).
Now it is extremely important to notice that such strain gradient can already
exist before the wear tests have started. Figure 2.8 shows a SEM cross section of a
carburized and ground 18CrNiMo7-6. Both render similar diffraction patterns that
are characteristic for an ultrafine grain size, while the obvious tendency to become
circular can qualitatively be seen for about 100–200 nm grain size.
Even though the loading during grinding is completely different from those under
ultra-mild wear the internal structure of the about 1 μm thick strain gradient (empha-
sized by the arrows) is very similar to that after wear (compare Figs. 2.7b and 2.9)
even though the grains are smaller directly after grinding.
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 37

Fig. 2.7 TEM bright field micrographs and diffraction patterns of a the bulk and b the strain
gradient of a carburized steel 18CrNiMo7-6 (SAE 18NCD6) after 2 × 106 cycles under boundary
lubricated sliding wear at 80 °C in gear oil

Fig. 2.8 SEM cross section

of a carburized steel
18CrNiMo7-6 after grinding.
The strain gradient is
emphasized by white arrows

Hence it is absolutely mandatory, that any TEM analyses of near-surface and

subsurface structures must be put into relation to those, that were there before the
wear test (e.g. after machining, grinding, polishing, etc.). Moreover, any TEM anal-
ysis requires the knowledge of the absolute position of the final samples inside and
underneath the wear track in relation to the incipient surface before wear. Otherwise
neither the predeformation characterized by the lattice defect state (and properties)
of that specific area is known nor their development into the investigated state after
the wear test.
38 A. Fischer et al.

Fig. 2.9 TEM figure and diffraction pattern of the strain gradient of a carburized steel 18CrNiMo7-
6 after grinding. The numbers represent indices of martensitic lattice planes

2.1.2 TEM Analyses of the Tribomaterial

The investigation of the uppermost layer by TEM and based on replicas started in
the 1950s under the headline of transfer layer under mild-wear conditions [41] and
was steadily improved. In the beginning the main target was to understand “material
transfer” a submechanism of “adhesion” under more severe-wear conditions. But it
could also be shown quite early that under mild-wear such layers consist in parts or in
total of oxidized wear particles of a few nm in size. It was found later by SEM-EDS
that such layers could be a mixture of materials from the body, the counterbody,
the interfacial media, and the environment leading to an alteration of the internal
structure but also of the chemical composition. In 1983 Heilmann et al. concluded
on the basis of TEM analyses of near-surface and subsurface areas that such layer
is a common appearance of wear, bares some nanostructured, nanocrystalline com-
posite that represents the surface in contact (s.a. 3rd bodies [42]), is generated very
early in the wear process (s.a. [43, 44]), mostly matches the wear particles, and is
influenced by body, counter-body, interfacial medium (e.g. lubricant) as well as the
environment [7]. Today these early findings can be related to different mechanisms
and submechanisms of ultra-mild wear e.g. by “microploughing” (of “abrasion”),
“indentations” (of “surface fatigue”), as well as “tribooxidation”, “tribocorrosion”,
and “mechanical mixing” (of “tribochemical reactions”).

2.1.2.1 The Tribomaterial in Austenitic Material

The generation of such nanocrystalline metal-base composite—from a former micro-

crystalline bulk that has been predeformed inside the ultrafine-crystalline strain gra-
dient—appears like a shear band (Fig. 2.10) [45]. It is important to notice that within
such shear bands the chemical composition does not alter but the grain size does over
three orders of magnitude from 20 μm to about 20 nm. The elementary mechanisms
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 39

Fig. 2.10 TEM micrographs of a a shear-banded region and b nanocrystalline tribomaterial with
incorporated metal-oxides marked by White Arrows from [45]

of such formation can be related to those known from severe plastic deformation
[29].
Due to the fact that tribomaterial differs chemically from the bulk material Rigney
proposed a model, that allows for such chemical changes just by the rotation of atoms
or clusters of atoms and called it “mechanical mixing” [46]. As for atoms in a MD
model this can also take place for nanocrystals which bring about plastic deformation
by the rotation of grains [29, 44]. The marked difference between SPD and tribol-
ogy is, that the stress field in tribology is not uniaxial. The shear band is supposingly
brought about e.g. by rotating compact wear particles (or any other debris) within the
interface generating “indentations”. Besides shallow particles that slide would lead
to “microploughing”. Both submechanisms generate mainly a very localized plastic
deformation while “indentations” is of cyclic nature and, therefore, a submecha-
nism of “surface fatigue” whereas “microploughing” is more of monotonic kind
and a submechanism of “abrasion”. Yet such rotation of nanocrystals at the surface
incorporates any interfacial media consisting e.g. of lubricant, wear particles, con-
taminants, reaction products, and further debris into such layer and a metal-based
nano-composite of some sub-μm thickness is formed by “mechanical mixing2 ” a
submechanism of “tribochemical reactions”. At mild and ultra-mild wear this tribo-

2 Theterm tribomaterial was proposed by David Rigney on a tribochemistry meeting in Hagi,

Japan in 2011 in order to summarize the large number of already existing terms like Beilby layer,
40 A. Fischer et al.

Fig. 2.11 BF micrographs of nanocrystalline tribomaterial of a a CoCrMo-retrieved cup after eight

years in vivo from [47] and b a simulator tested head after 10 × 106 cycles from [48]. The EELS-
maps of b show an extensive amount of C at the surface, which represents denatured proteins, while
the metal is characterized by Co and Cr

material persists, separates body and counterbody, and allows for the extreme shear
rates. Because of the very small size of its constituents far below 100 nm TEM
with EDS or EELS is the only microscopical method that renders a combined local
information about structure, defect state, and chemistry.
Büscher et al. [47] as well as Pourzal et al. [48] showed that even at body tem-
perature and at relatively small Hertzian contact pressures of about 50 MPa a high
strength Co-Base alloy generates such tribomaterial inside the primary articulating
surface of an artificial hip joint with thicknesses up to 500 nm (Fig. 2.11a, b). It con-
sists of CoCrMo nanocrystals, human pseudosynovia (or bovine calf serum if tested
in laboratory), and denatured-proteins, while the latter can be analyzed by EDS or
EELS for their high Carbon content (Bright-Field and C-Map, Fig. 2.11b). It was
found that such metal-organic composite is the main reason for the longevity of such
selfmating metal-metal sliding contacts of austenitic alloys.
The metal part is a CoCrMo solid solution with a grain size ranging from 10
to 70 nm3 and supported by a strain gradient that is stabilized by twins, stacking
faults, and ε-martensite lathes forming rhomboid domains that become the smaller
the closer to the surface. Such tribomaterial is a chaotic structure and might consist

white layer, fragmented layer, transfer layer, glaze layer, mixed layer, 3rd bodies, highly deformed
layer,….
3 It is important to notice that even though the tribomaterial appears in different thicknesses and

mixtures nearly at any analyzed position while its grain size range was always very similar and
independent of source (cast, wrought, new or retrieved hip joint), origin (1960s–2000s), make (heat
treatment, low- or high Carbon content, manufacturer) of the CoCrMo alloy, and loading (simulator,
one or twenty-two years in vivo).
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 41

Fig. 2.12 BF Images of tribomaterial of a a simulator tested CoCrMo Head after 5 × 106 cycles
depicting a nanocrystalline zone with denatured proteins on top while b shows a denatured protein
film with embedded metal nanoparticles in a different position of the same sample from [48]. The
dotted line marks the interface within the tribomaterial (compare to the scheme in Fig. 2.3) between
nanocrystalline CoCrMo and the mixed zone of denatured proteins and CoCrMo nanosize particles

of a mixture of metal nanoparticles with a face-centered-cubic (austenite) and/or

hexagonally-closed-packed (ε-martensite) lattice topped by and intermixed with
denatured proteins (Fig. 2.12a) or denatured proteins with embedded metallic or
metal-oxide nanoparticles (Fig. 2.12b).
The structure of the denatured protein constituent in such tribomaterial has been
investigated by TEM as well. It is quite surprising to find graphitic structures which
stem from denatured proteins like e.g. albumin but any artefacts during prepara-
tion and electron-beam as well as photon-based analyses has been experimentally
ruled out [27]. Still the elementary mechanism of this transformation is not clear yet
(Fig. 2.13).

2.1.2.2 The Tribomaterial of Ferritic and Martensitic Steels

Ferritic or perlitic steels are used in many tribosystems demonstrating their high wear
resistance e.g. in wheel/rail contacts or in cylinders of combustion engines. The latter
might be coated with a ferritic steel by plasma-based processes for the need of higher
ignition pressures and/or because the crank case is made out of AlSi-based castings.
Principally plasma-spray processes allow for a very high solidification rate and lead
to quite chaotic microstructure of μm-size so-called splat grains, which internally
consist of a very fine microstructure (Fig. 2.14) [49].
42 A. Fischer et al.

Fig. 2.13 HRTEM image of a Tribomaterial with nanocrystals and an embedded graphitic particle
(white arrow); b graphitic structure at higher magnification from [27] and supporting online material

Fig. 2.14 TEM micrograph of the microstructure of a plasma-transferred-wire-arc coating (PTWA)

of plain steel with 0, 1 weight-% carbon depicting a BF; a microcrystalline ferrite, and b BF and c
DF nanocrystalline Fe3 C and FeO precipitates from [49]

Thus the incipient structure inside the splat grains appears like a precipitation
hardened ferrite, while the alloying elements C and O stem either from the alloy
by 0.1% C or from contamination by O during atmospheric spraying under just
an inert-gas shield. Laboratory, engine, and field tests did reveal that such coating
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 43

Fig. 2.15 a Bright field and b HRTEM micrographs of a PTWA coating surface in the combustion
chamber after a field test of 60,000 km in a passenger car’s diesel engine showing an amorphous
layered structure with incorporated nanocrystalline particles from [49]

provides a sufficient resistance against wear and lower friction against Cr-coated
piston rings. Beside the classical analyses of such tribosystems the TEM analyses
revealed some new findings at that time. A cylinder-piston ring system consists of four
different tribosystems as to the loading characteristics (top dead center, bottom dead
center, combustion chamber above top dead center, and stroke) resulting in different
microstructural alterations of the coating into tribomaterial and strain gradient. Thus
any alterations must be related to the specific position inside the cylinder. Despite
the fact that inside the combustion chamber above the top dead center no piston ring
contact takes place the tribomaterial shows some amorphous layer with embedded
nanocrystals of about 10 nm in size (Fig. 2.15a), while it ranges between 100 and
300 nm within the underlying strain gradient (Fig. 2.15b).
By checking the surface below the bottom dead center it was found that after
honing neither a strain gradient nor tribomaterial could be observed. The incipient
microcrystalline structure must have been refined about 1000 fold just by the cyclic
impacts of the ignition at higher temperatures. An EFETM analysis revealed that the
amorphous layer consists mainly of C, the nanosize particles contain higher amounts
of O and Fe, while the base material contains Fe. From this on can derive that combus-
tion products form an amorphous layer on top of the oxidized steel surface, which
is mechanically and thermally loaded by the ignition leading to grain refinement.
Whether the oxide particles stem from the precipitates or represent oxidized wear
particles was not differentiated (Fig. 2.16).
At the top dead center the piston ring is sliding over the surface but generates
cyclic impacts during ignition. Still the microstructural and chemical alterations
are not very different from the combustion chamber, but now measurable wear of
some μm took place at this position (Fig. 2.17). The tribomaterial appears like a
nanostructured multilayer coating while the strain gradient shows ultra-file grains.
44 A. Fischer et al.

Fig. 2.16 The EFTEM analysis of the amorphous layer with embedded nanocrystals in the com-
bustion chamber of Fig. 2.15; a bright field, b C-map, c O-map, d Fe-map

Fig. 2.17 HRTEM bright field micrographs of a PTWA coating surface at top dead center showing
a a nanostructured tribomaterial supported by b an ultrafine crystalline strain gradient from [49].
The white arrow represents the relative movement of the Piston Ring

It is quite interesting that the latter appears like a former subgrain or dislocation cell
structure generated by cyclic loading that is more or less immediately recrystallized
by elevated subsurface temperatures during the ignition phase. This was supported
by the fact that at the much colder bottom dead center the dislocation cells were still
visible [49].
An EELS analysis reveals that this layered structure is mainly brought about by O
and C, while interestingly enough O is not always combined with but also exchanged
by Fe in this nanostructure. From this one can conclude that this layer is a mixture
of combustion products and near-surface Fe-base coating. Still any constituents of
the lubricant are missing (Fig. 2.18).
Thus, from such TEM micrographs one might get a very good impression about
the chaotic nature of tribomaterial but in order to demystify it additional analyses
e.g. by XPS, AES, or RS are necessary.
Martensitic steels have a much higher strength—mostly characterized by hard-
ness—than ferritic ones but they might generate tribomaterial as well (Fig. 2.19a,
b). Now it is important to notice that such tribomaterial must not necessarily be a
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 45

Fig. 2.18 The EELS analysis of the nanostructured tribomaterial at the top dead center of Fig. 2.17;
a bright field, b C-map, c O-map, d Fe-map from [49]

Fig. 2.19 Nanocrystalline tribomaterial after boundary lubricated sliding wear of the polished cold
work tool steel 56NiCrMoV7 of Fig. 2.6 but now worn under a normal load of 400 N from [39]

nanocrystalline layer of a certain thickness (Fig. 2.19c, blue arrow), but it can also
appear as a certain near-surface volume of different thicknesses depending on the
loading history of that specific location (Fig. 2.19d, red arrow). Still the interface
between the strain gradient and the tribomaterial is a sharp line.
In order to get a 1st glimpse—still not a full insight—on the loading history of
a specific location for TEM analyses on can combine the depth of the wear scar in
comparison to the incipient surface with EBSD analyses of cross sections as shown
in Fig. 2.20. Due to the limited resolution of EBSD analyses even at 30 nm step
size any nanocrystalline tribomaterial is characterized by no-counts (black areas
in Figs. 2.20a, b). But from such a figure one can read that the nanocrystalline
surface zone from milling (Fig. 2.20a, c), which represents the secondary-shear zone
generated during chip formation, has completely been worn away and a totally new
one has been generated by further sliding wear (Fig. 2.20b, d). By this procedure it
is unequivocally shown the strain gradient in Fig. 2.7 as well as the tribomaterial in
Fig. 2.20d are no remains of the last production step but generated by wear.
46 A. Fischer et al.

Fig. 2.20 Cross sections depicting the microcrystalline martensitic structure of a carburized
17CrNiMo7-6. a EBSD after milling, c SEM after milling, b EBSD after boundary lubricated
sliding wear, d TEM BF and DP after Wear. Compare the BFs and DPs of (d) with those of 7a
(bulk) and 7b (strain gradient)

At ultra-mild wear this becomes quite tricky, because at extremely small wear
rates e.g. after milling and finishing the surface levels before wear (Fig. 2.21a) and
after wear (Fig. 2.21b) are the same. Thus one cannot distinguish from such EBSD
analyses alone whether there were any tribologically driven alterations of the near-
surface nanocrystalline microstructure. Now after milling and finishing there is a
nanocrystalline zone on the surface (Fig. 2.21c, d) which also looked absolutely
the same after wear. It was quite interesting to see that the polished microcrystalline
surface of the self-mating counterbody became nanocrystalline during wear and gen-
erated a distinct strain gradient similar to that in Fig. 2.20b, d. Thus, one can conclude
that the frictional work was dissipated mostly—if not solely—by the counterbody
which also corresponded to the measurable differences of the amounts of wear of
both bodies.
Thus the near-surface structures and the wear behavior after machining in compar-
ison to machining and finishing leads to the conclusion that not any nanocrystalline
layer generated by manufacturing processes is a good precursor of tribomaterial and
diminishes the wear rate. It strongly depends on the existence and structure of a
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 47

Fig. 2.21 Cross sections depicting the microcrystalline martensitic structure of a carburized
17CrNiMo7-6. a EBSD after milling and finishing, b EBSD after boundary lubricated sliding
wear, c TEM BF image and d DP after milling and finishing

supporting sub-surface strain gradient as well as it is for that of the counterbody. It is

absolutely necessary, therefore, to investigate both body and counterbody by means
of similar methods in order to understand wear of a tribosystem.

2.1.3 Wear Particle Analyses by TEM

In the past wear particles were often believed being representatives for a character-
istic segment of the contacting surfaces before they failed [33]. Thus wear particle
analyses were always a crucial part of exploring and explaining sub-mechanisms like
e.g. “triboxidation” [20, 34]. But by time it was understood that such particles might
distinctly change their structure and chemistry after they detach from surfaces [50].
Today the preparation of wear particles from e.g. the lubricant play an extremely
important role in order to avoid artefacts, that would mislead the analyses. In gen-
eral it is known that depending on its size and nature such particles would oxidize
immediately after detachment. It is not clear in any case, whether they might stem
from an oxide film that detached or whether they detached as metallic particle and
reacted with the interfacial medium because of their pyrophoricity. Either way wear
48 A. Fischer et al.

Fig. 2.22 TEM micrographs and corresponding diffraction patterns of wear particles extracted from
the lubricant after a self-mating wear test of carburized 17CrNiMo7-6 under boundary lubrication

particles contain information about the tribological behavior of the specific system
and should always be analyzed in combination with the surfaces they stem from [25].
E.g. nanometer small particles detaching from carburized 17CrNiMo7-6 under
boundary lubrication exhibit a compact granular shape and a highly deformed internal
nanostructure (Fig. 2.22).
This resembles that of the tribomaterial from which one can conclude that the wear
particles stem from the near-surface region. By diffraction patterns any oxidation
could be ruled out as it was found under dry sliding wear conditions of such steels.
While such nanosize particles tend to agglomerate (Fig. 2.23) either inside the wear
tracks or outside during preparation for TEM analyses any evaluation of agglomerates
needs further analyses beyond TEM in order to be related to a certain mechanism or
submechanism.
In biomedical engineering wear particles are of major concern, because most of
them are in the nm-range, likely causing adverse tissue reactions, and finally leading
to failure of artificial hip joints [51–53]. Doorn et al. analyzed such particles by
TEM isolated from periprosthetic tissue which stem from both the CoCrMo-head of
such hip joints as well as the TiAl6V4-metal backing [54]. Before his work particle
sizes were reported somewhat between 10 nm and 50 μm, while his particles were
sub-μm size from 10 nm to about 400 nm, while the agglomerates of them were
in the μm-range. He could show that most of the wear particles had a metallic
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 49

Fig. 2.23 SEM micrographs of agglomerated wear Debris a in full and b in detail extracted from
the lubricant after a self-mating wear test of carburized 17CrNiMo7-6 under boundary lubrication

core while some were of Cr- or Ti-oxides. Catelas et al. compared wear particles
taken from hip simulator tests and periprosthetic tissue by TEM and EDS [55]. They
concluded that most particles from the hip simulator resemble those in vivo and—if
not agglomerated—are smaller than 150 nm. According to the EDS analyses they
were either Cr-oxides, Cr-carbides, or they stem from the CoCrMo metal matrix. The
latter appeared needle-shaped and were later related to strain-induced ε-martensite
lathes by Büscher et al. [56]. Billi et al. offered a different protocol for particle
preparation and found additional ultrafine particles of about 200–500 nm in size that
mostly contained Mo and Cr [57]. Acc. to Stemmer et al. such particles originate
from intermetallic σ-phases that might precipitate in CoCrMo alloys together with
Cr-carbides of the same size forming “so-called” mixed hard phases [58]. Pourzal
et al. used the Catelas-protocol and compared wear particles by TEM, EDS and
EFTEM [59]. Three different groups of wear particles—named I, II, and III—could
be distinguished by size, chemical composition, and lattice structure.
All type I and II wear particles contain oxygen and chromium (Fig. 2.24). While
type II are Cr-oxides the type I might also contain some cobalt. None showed any
distinct amount of carbon, which would be characteristic for carbides. Surprisingly
some type I particles depicted an amorphous structure with a crystalline Co-rich core
(Fig. 2.25) while type III particles were 5–15 nm small fragments of type II fully
oxidized into Cr2 O3 .
All these particles would originate from the tribomaterial and can be related to
“mechanical mixing”. The oxides could either be parts of the passive film generated
by “tribocorrosion” that detaches under load or by nanosize particles that oxidize
because of their pyrophoricity. A possible hypothesis would be that single grains
within the nc-zone are transported due to the plastic flow of entire grains to the
mechanically mixed zone where carbon-rich denatured organic material as well as
phosphates settle at the grain boundaries. During ongoing plastic flow by “indenta-
50 A. Fischer et al.

Fig. 2.24 TEM micrograph of a type I and type II wear particles and b corresponding EFTEM
mapping from [59]

Fig. 2.25 a Diffraction pattern and b HRTEM micrograph of the core of a type I wear particle
from [59]. The region marked by the white arrow is a Co-rich nanocrystal being embedded into an
amorphous matrix

tion” and/or “microploughing”, such grains could be transported to the uppermost

surface where they would stick in a film of denatured proteins. Eventually, the grains
would be pulled out of the tribomaterial and form single particles. Interestingly
enough Pourzal showed by AFM-measurements that at this point, the particles would
still exhibit a relatively high amount of cobalt as well as chromium, and in addition
be protected by a shell of organic material [60]. Thus, it must be mentioned here that
this organic shell is removed by the current preparation protocols of wear particles
and was never found by TEM.
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 51

2.2 Conclusions and Outlook

TEM proved to be a suitable tool to understand the pathways of frictional energy

dissipation into metallic materials. It can render information on the uppermost tri-
bomaterial, its structure, and chemical composition from which the sequence of its
generation can be modelled. Further TEM shows, whether a wear particle is gen-
erated inside the tribomaterial or inside the strain gradient. It allows for structural
and chemical analyses of nm-size particles and brings about information of their
composition and the interaction inside and outside of the contact interface.
Elementary mechanisms and processes can only be analyzed by TEM—even
in situ [61]—rendering a combination of structural and chemical information at a
particular position at high resolution. It is one of the essential tools to understand
metallurgical features and to allow for a well-aimed development of more wear
resistant metallic materials.
Still TEM is not a stand-alone method even though today it is combined with EDS,
EFTM, EELS, and EBSD. In order to understand the relevance of any appearance
and whether it contributes to the characteristics of a tribosystem further techniques
can and should be involved like e.g. SEM, AFM, XPS, AES, or RS.
Future work should be based on what has been investigated in the past and continue
in connecting mechanisms and submechanisms to the elementary processes. Due to
the fact that most papers focus on steady-state the predominantly plastic interaction
during run-in should be analyzed in more detail. It has been proven that a combination
of TEM and MD might help to understand peculiar and unexpected tribological
behavior [62]. Even though with TEM we can get information at high resolution as
to location in a wear track we still lack of that high-resolution information as to the
time-line.

Attachments

Preparation of Samples

This chapter exemplifies processes to prepare samples for TEM analyses of near-
surface and subsurface areas of worn specimens and parts of metallic alloys.

Specimens from the Strain Gradient of Austenitic or Ferritic Metal

Alloys

For the preparation of cross sections the method developed by Büscher et al. [63] or
Hahn [64] can be used. Here, two corresponding segments of body and counterbody
were taken and glued onto each other’s articulating surface by means of a suitable
adhesive (Epoxy G1, Gatan, Munich, Germany). The sample was fixed by a slotted
52 A. Fischer et al.

pipe with a diameter of 2.5 mm and positioned in a brass tube of 3 mm diameter. The
tube was cut to slices of 400 μm thickness. The cross section of one slice precisely
exposed the segments of cups and head lying on each other separated by a small
gap of hardened epoxide adhesive. Using grinding, dimple grinding (Model 656,
Gatan, Munich, Germany) and ion milling (PIPS 691, Gatan, Munich, Germany)
the sample was thinned to the desired thickness of about 40 nm. Afterward the
specimens were investigated by means of transmission electron microscopy (TEM
400 Phillips, Eindhoven, Netherlands). In order to observe chemical changes in the
uppermost surface layers, a high resolution TEM (Tecnai F20 Phillips, Eindhoven,
Netherlands) with EDS and electron energy loss spectroscopy (EELS) was used.

Specimens from the Worn Surface of Martensitic Steels by FIB [39]

Sample preparation by means of focused ion beam (FIB), was performed using a
dual beam FIB/SEM system (Helios NanoLab 600, FEI, Eindhoven, Netherlands).
The cross-section procedure contained several steps. First, a protective Pt-layer was
deposited at the area of interest on the sample surface. After this, bulk material was
removed in a wedge-shaped trench by Ga+ ions, on one side of the area. Using
decreasing energies of the ion beam, the sidewall was polished at a small glancing
angle in subsequent polishing steps. The processing parameters are given in Table 2.3.
Micrographs of the cross-sections were obtained in SE mode at 2 kV accelerating
voltage. Microstructural features emerge due to electron channeling contrast (ECC).

Specimens from the Worn Surface of Martensitic Steels by Ion Milling

TEM cross-section samples of the wear tracks were prepared using an ion polishing
system (EM-09100IS, Jeol, Akishima, Japan). Therefore, small samples parallel to
the sliding direction were cut, as pictured in Fig. 2.4 (1–4) and a silicon wafer was

Table 2.3 FIB cross sectioning parameters

Acceleration Ion current (nA) Stage tilt angle Shape
voltage (kV) (°)
Fast milling 30 6.5 52 Regular cross
section
Milling 30 2.8 52 Cleaning cross
section
Polishing 30 0.46 53 Cleaning cross
section
Polishing 30 93 × 10−3 54 Cleaning cross
section
Cleaning 5 47 × 10−3 55 Rectangle
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 53

Table 2.4 Ion slicer parameter

Acceleration Gun tilt angle (°) Interval Time (h)
voltage (kV)
Milling 5 1 77 5
Cleaning 2 1.5 77 0.25

Fig. 2.26 Cross section preparation out of a wear track. 1 Cutting, 2 Glueing a Si-Wafer on top
of the wear track 3 grinding down to a thickness of about 100 μm, 4 ion polishing by IS, 5 final
thinning by PIPS from [65]

applied to the surface as described in 4.2. Afterwards, the samples were ground with
1200 grit SiC paper to a thickness of 100 μm. Using an ion slicer, the cross-sections
were then polished on both sides with Ar+ ions, while a thin ridge of the sample
was masked by a metal foil. The process parameters are shown in Table 2.4. In
order to gain electron transparency, an additional ion-milling process was necessary
(Fig. 2.26). Therefore, the specimens were successively thinned on both sides using
an ion-mill (PIPSII Model No. 695, Gatan, USA). The ion-mill was operated at
accelerating voltages between 5–0.5 kV and gun angles between 5° and 10°.
54 A. Fischer et al.

Wear Particle Preparation from Lubricant (Gear Oil)

Particle Isolation
(1) Centrifuge lubricant samples and carefully remove supernatants (centrifuga-
tion time depends on the viscosity of the medium). Do not touch the pellet at
the bottom of the tubes.
(2) Resuspend and wash particles with cyclohexane, use shaker (Vortex-Genie-2,
Scientific Industries, Bohemia, NY, USA) and ultrasonic cleaner.
(3) Centrifuge for 15 min at 20,000 g.
(4) Carefully discard supernatants. Do not touch the pellet at the bottom of the
tubes!
(5) Resuspend and wash particles with acetone, use shaker and ultrasonic cleaner.
(6) Centrifuge for 15 min at 20,000 × g.
(7) Carefully discard supernatants. Do not touch the pellet at the bottom of the
tubes!
(8) Resuspend and wash particles with isopropanol, use shaker and ultrasonic
cleaner.
(9) Centrifuge for 15 min at 20,000 × g.
(10) Carefully discard supernatants. Do not touch the pellet at the bottom of the
tubes!
if necessary repeat steps 2–8.
(11) Store particles in isopropanol or ethanol at 4 °C.
Cyclohexane and acetone are used in order to dissolve the lubricant. These solvents
(particularly cyclohexane) should not be stored in the tubes for too long since they
can damage the tubes! They would also damage the carbon film on the Cu-grids.
Therefore, it is necessary to suspend the particles in a less aggressive solvent, such
as isopropanol or ethanol, before applying them on Cu-grids.
Particle embedding
1st day:
(1) Centrifuge for 15 min at 20,000 × g.
(2) Prepare acetone/epoxy mixture
Mix epoxy and hardener at a ratio of 100:50 (Epoxy 3000 Quick, Cloeren
Technology GmbH, Wegberg, Germany)
Add 100% acetone at a ratio of 1:1.
(3) Carefully discard supernatants. Do not touch the pellet at the bottom of the
tubes! Add 0.5 ml of the acetone/epoxy mixture.
(4) Place tubes in a rotator (Thermomixer compact, Eppendorf, Hamburg, Ger-
many) for 24 h.
2nd day:
(1) Centrifuge for 15 min at 20,000 × g.
(2) Place tubes, with open lids, under vacuum for 1 h in order to remove the acetone.
(3) Place tubes, with open lids, at 80 °C for 24 h for polymerization.
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 55

3rd day:
Remove tubes to obtain solid pieces of resin with particles embedded at the bottom.
Section with diamond blade to a thickness of approx. 90 nm.

Wear Particle Preparation from Biomedical Lubricant (Bovine Calf

Serum)

The protein rich testing fluid (bovine serum) from both tests containing the wear
particles was stored at −20 °C. Proteins disturb the image contrast, increase the
degree of agglomeration and inhibit the identification of single particles in SEM and
TEM. Compared to rather inert polyethylene particles, the isolation of metallic wear
particles requires a more sophisticated protocol using enzymatic digestion to remove
organic components, especially proteins, without damaging the particles. Particles in
the present study were isolated following the protocol developed by Catelas et al. [55,
66] but with minor changes. This protocol was shown to minimize particle damage.
It consists of several washing steps as well as incubation steps with two different
enzymes, papain and proteinase K. Papain is a cysteine protease enzyme. Its main
mechanism of action is destruction of peptide bond. Proteinase K is a broad-spectrum
serine protease. This enzyme is known for its protein denaturing abilities, especially
in presence of reagents like etylenediaminetetraacetic acid (EDTA). The amount of
testing fluid retrieved from the laboratory tests ranged from 60 ml (hip simulator
tests) to 300 ml (sliding wear test rig). Fluid samples were aliquoted into 40 ml
tubes and centrifuged at 18,000 × g for 15 min using a Sorvall RC-5B superspeed
centrifuge (particles from the different aliquots of the same testing fluid sample were
recombined after the first overnight incubation). Supernatants were discarded except
1.5 ml in order to resuspend the pellets and transfer them in microcentrifuge tubes.
All centrifugations were performed for 10 min at 18,000 × g. First, the pellets were
resuspended in 2.5% sodium dodecyl sulfate (SDS) (v/v distilled water) and boiled
for 10 min. This was followed by one wash in 80 % acetone and three washes in
1 ml of 250 mMol sodium phosphate buffer solution (PBS) containing 25 mMol
EDTA at pH 7.4. After 30 s of sonification, papain (3.2 Units in 1 ml PBS/EDTA)
was added. The samples were then incubated overnight on an Eppendorf thermo-
mixer at 65 °C under slight motion. After the incubation, pellets from aliquots of the
same initial fluid sample were combined in microcentrifuge tubes. Pellets were then
resuspended in 2.5% SDS, boiled again for 10 min and then washed twice in 1 ml
of 50 mM Tris-HCl pH 7.6. Before adding proteinase K (3 Units in 1 ml Tris-HCl),
the pellets were sonicated for 30 s. A second overnight incubation was conducted at
55 °C under slight motion. After the incubation, the samples were boiled again in
2.5% SDS. The pellets were then washed once in 1 ml of 50 mMol Tris-HCl, once in
3 % SDS (v/v 80% acetone) and once in distilled, deionized water. At the end of the
isolation protocol, the pelleted particles were stored in 100% ethanol at 4 °C. In a
few cases, the pellets still contained denatured organic components. For those cases,
the second overnight incubation was repeated. This may have been due to the use of
40 ml initial aliquots instead of 15 ml. Particles were then embedded in epoxy resin
56 A. Fischer et al.

for TEM analysis. Prior to embedding, the particles were centrifuged for 20 min at
18,000 × g. The supernatant was removed and the particles were resuspended in
a solution of 0.5 ml acetone and 0.5 ml liquid epoxy resin (Epoxy 3000, Cloeren
Technology GmbH, Wegberg, Germany). Microcentrifuge tubes were then placed
on a thermomixer overnight under slight motion at room temperature to allow pellet
infiltration by the resin. The tubes were further centrifuged for 20 min at 18,000 × g
and degassed for 1 h to evaporate the acetone. One ml of epoxy resin was added and
the tubes were placed in a vacuum chamber for 1 h to remove potential air bubbles
and trace of remaining acetone. The samples were then placed in an oven at 60 °C
for one hour to harden the epoxy resin. Once hardened, the transparent epoxy cone
at the tip of the tubes showed the embedded particles. Approximately 100 nm thick
sections were cut using a diamond knife and placed on carbon film coated copper
grid nets (formvar carbon-film S162-4, Plano GmbH, Wetzlar, Germany) for TEM
analysis.

References

1. M. Knoll, E. Ruska, Das Elektronenmikroskop. Z. fuer Phys. 78, 318–339 (1932)

2. R.F. Deacon, J.F. Goodman, Lubrication by lamellar solids. Proc. R. Soc. Lond. Math. Phys.
Eng. Sci. 243, 464–482 (1958)
3. F.P. Bowden, A.J.W. Moore, D. Tabor, The ploughing and adhesion of sliding metals. J. Appl.
Phys. 14, 80–91 (1943)
4. R.S. Plumb, W.A. Glaeser, Wear prevention in molds used to mold boron filled elastomers.
Wear 46, 219–229 (1966)
5. D.A. Rigney, W.A. Glaeser, The significance of near surface microstructure in the wear process.
Wear 46, 241–250 (1978)
6. P.A. Higham, F.H. Stott, B. Bethune, Mechanisms of wear of the metal surface during fretting
corrosion of steel on polymers. Corros. Sci. 18, 3–13 (1978)
7. P. Heilmann, J. Don, T.C. Sun, D.A. Rigney, W.A. Glaeser, Sliding wear and transfer. Wear
91, 171–190 (1983)
8. S.B. Newcomb, W.M. Stobbs, A transmission electron microscopy study of the white-etching
layer on a rail head. Mater. Sci. Eng. 66, 195–204 (1984)
9. N. Ohmae, T. Nakai, T. Tsukizoe, Prevention of fretting by ion plated film. Wear 30, 299–309
(1974)
10. J.M. Martin, J.L. Mansot, I. Berbezier, H. Dexpert, The nature and origin of wear particles
from boundary lubrication with a zinc dialkyl dithiophosphate. Wear 93, 117–126 (1984)
11. K.H.Z. Gahr, Formation of wear debris by the abrasion of ductile metals. Wear 74, 353–373
(1981)
12. J.J. Wert, F. Srygley, C.D. Warren, R.D. McReynolds, Influence of long-range order on defor-
mation induced by sliding wear. Wear 134, 115–148 (1989)
13. W.M. Rainforth, R. Stevens, J. Nutting, Deformation structures induced by sliding contact.
Philos. Mag. A Phys. Condens. Matter Struct. Defects Mech. Prop., 66, 621–641 (1992)
14. C. Greiner, Z. Liu, L. Strassberger, P. Gumbsch, Sequence of stages in the microstructure
evolution in copper under mild reciprocating tribological loading. ACS Appl. Mater. Interfaces
8, 15809–15819 (2016)
15. H. Czichos, D. Dowson, Tribology: a systems approach to the science and technology of
friction, lubrication and wear. Tribol. Int. 11, 259–260 (1978)
16. H. Czichos, Wear mechanisms in tribological systems, in Overview and Classification, ed. by
K.F. Ehmann (Publ by ASME, New Orleans, LA, USA, 1993), pp. 239–241
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 57

17. A. Fischer, Well-founded selection of materials for improved wear resistance. Wear 194,
238–245 (1996)
18. K.H. Zum Gahr, Microstructure and Wear of Materials (Elsevier Science Publishers, Amster-
dam, The Netherlands, 1987)
19. M.E. Sikorski, The adhesion of metals and factors that influence it. Wear 7, 144–162 (1964)
20. T.F.J. Quinn, NASA Interdisciplinary Collaboration in Tribology. A Review of Oxidational
Wear (Georgia Inst. Technol, 1983)
21. S. Jahanmir, N.P. Suh, E.P. Abrahamson II, The delamination theory of wear and the wear of
a composite surface. Wear 32, 33–49 (1975)
22. D. Landolt, S. Mischler, M. Stemp, Electrochemical methods in tribocorrosion: a critical
appraisal. Electrochim. Acta 46, 3913–3929 (2001)
23. M.A. Wimmer, J. Loos, M. Heitkemper, A. Fischer, The acting wear mechanisms on metal-
on-metal hip joint bearings—in-vitro results. Wear 250, 129–139 (2001)
24. D.A. Rigney, J.E. Hammerberg, Mechanical mixing and the development of nanocrystalline
material during the sliding of metals., in Advanced Materials in the 21st Century: The 1999
Julia R. Weertman Symposium, The Minerals, Metals & Materials Society, ed. by Y.W. Chung,
D.C. Dunand, P. Liaw, G.B. Olson (Warrendale, PA, USA, 1999), pp. 465–474
25. A. Fischer, S. Weiss, M.A. Wimmer, The tribological difference between biomedical steels and
CoCrMo-alloys. J. Mech. Behav. Biomed. Mater. 1, 50–62 (2012)
26. M.A. Wimmer, M.P. Laurent, M.T. Mathew, C. Nagelli, Y. Liao, L.D. Marks, J.J. Jacobs,
A. Fischer, The effect of contact load on CoCrMo wear and the formation and retention of
tribofilms. Wear 332–333, 643–649 (2015)
27. Y. Liao, R. Pourzal, M.A. Wimmer, J.J. Jacobs, A. Fischer, L.D. Marks, Graphitic tribological
layers in metal-on-metal hip replacements. Science 334, 1687–1690 (2011)
28. M.A. Wimmer, A. Fischer, R. Buscher, R. Pourzal, C. Sprecher, R. Hauert, J.J. Jacobs, Wear
mechanisms in metal-on-metal bearings: the importance of tribochemical reaction layers. J.
Orthop. Res. 28, 436–443 (2010)
29. R. Valiev, Nanostructuring of metals by severe plastic deformation for advanced properties.
Nat. Mater. 3, 511–516 (2004)
30. G. Schmaltz, Technische Oberflächenkunde; Feingestalt und Eigenschaften von Grenzflächen
technischer Körper, insbesondere der Maschinenteile, J. (Springer, Berlin, Germany, 1936)
31. R. Büscher, Gefügeumwandlungen und Partikelbildung in künstlichen Metall/Metall-
Hüftgelenken, Werkstofftechnik, PhD-Thesis, Universität Duisburg-Essen, Germany, s.a. VDI-
Fortschr.Ber., Reihe 17, Nr.256, (VDI-Verlag, Düsseldorf, Germany, 2005)
32. R. Glardon, S. Chavez, I. Finnie, Simuation of sliding wear by cyclic plastic deformation under
combined stresses. J. Eng. Mater. Technol. Trans. ASME 106, 248–252 (1984)
33. W.A. Glaeser, Transmission electron microscopy on wear debris from bronze bearings. Wear
43, 393–394 (1977)
34. W.J. Saleski, R.M. Fisher, R.O. Ritchie, G. Thomas, The nature and origin of sliding wear
debris from steels, in Wear of Materials ‘83 ed. by K.C. Ludema, (ASME, 345 East 47th
Street, New York, N.Y. 10017, USA, Reston, VA, USA, 1983), pp. 434–445
35. M. Schymura, R. Stegemann, A. Fischer, Crack propagation behavior of solution annealed
austenitic high interstitial steels. Int. J. Fatigue 79, 25–35 (2015)
36. N. Jost, I. Schmidt, Friction-induced martensitic transformation in austenitic manganese steels.
Wear 111, 377–389 (1986)
37. Z.M. He, Q.C. Jiang, S.B. Fu, J.P. Xie, Improved work-hardening ability and wear resistance
of austenitic manganese steel under non-severe impact-loading conditions. Wear 120, 305–319
(1987)
38. C.W. Shao, P. Zhang, R. Liu, Z.J. Zhang, J.C. Pang, Q.Q. Duan, Z.F. Zhang, A remarkable
improvement of low-cycle fatigue resistance of high-Mn austenitic TWIP alloys with similar
tensile properties: importance of slip mode. Acta Mater. 118, 196–212 (2016)
39. A. Brink, Einlaufverhalten von geschmierten Stahl-Stahl PAarungen unter Berücksichtigung
der Mikrostruktur., PhD-Thesis, Institut für Angewandte Materialien - Computational Materials
Science (Karlsruhe Institute of Technology, Karlsruhe, Germany, 2015)
58 A. Fischer et al.

40. D. Stickel, A. Fischer, The influence of topography on the specific dissipated friction power in
ultra-mild sliding wear: experiment and simulation. Tribol. Int. 91, 48–59 (2015)
41. J.F. Archard, W. Hirst, An examination of a mild wear process, Proc. R. Soc. Lond. Ser. A.
Math. Phys. Sci. 238, 515–530 (1957)
42. M. Godet, The third-body approach: a mechanical view of wear. Wear 100, 437–452 (1984)
43. A. Fischer, D. Stickel, C. Schoss, R. Bosman, M. Wimmer, The growth rate of tribomaterial in
bovine serum lubricated sliding contacts. Lubricants 4, 21 (2016)
44. N. Beckmann, P.A. Romero, D. Linsler, M. Dienwiebel, U. Stolz, M. Moseler, P. Gumbsch,
Origins of folding instabilities on polycrystalline metal surfaces. Phys. Rev. Appl. 2, 064004
(2014)
45. W.M. Rainforth, R. Stevens, J. Nutting, Deformation structures induced by sliding contact.
Philos. Mag. A 66, 621–641 (1992)
46. D.A. Rigney, J.E. Hammerberg, Unlubricated sliding behavior of metals. MRS Bull. 23, 32–36
(1998)
47. R. Büscher, A. Fischer, The pathways of dynamic recrystallization in all-metal hip joints. Wear
259, 887–897 (2005)
48. R. Pourzal, R. Theissmann, M. Morlock, A. Fischer, Micro-structural alterations within dif-
ferent areas of articulating surfaces of a metal-on-metal hip resurfacing system. Wear 267,
689–694 (2009)
49. M. Hahn, R. Theissmann, B. Gleising, W. Dudzinski, A. Fischer, Microstructural alterations
within thermal spray coatings during highly loaded diesel engine tests. Wear 267, 916–924
(2009)
50. D.A. Rigney, The role of characterization in understanding debris generation, in Tribology
Series; Wear Particles: Frorn the Cradle to the Grave, Proceedings of the 18th Leeds-Lyon
Symposium on Tribology ed. by D. Dowson, C.M. Taylor, T.H.C. Childs, M. Godet, G. Dalmaz
(Lyon, France, 1992), pp. 405–412
51. I. Catelas, J.J. Jacobs, Biologic activity of wear particles. Instr. Course Lect. 59, 3–16 (2010)
52. H.G. Willert, H. Bertram, G. Hans Buchhorn, Osteolysis in alloarthroplasty of the hip: the role
of ultra-high molecular weight polyethylene wear particles. Clin. Orthop. Relat. Res. 95–107
(1990)
53. H.G. Willert, G.H. Buchhorn, C.H. Lohmann, Hypersensitivity to CoCrMo-debris from
metal/metal hip endoprostheses, in Transactions—7th World Biomaterials Congress, Sydney,
(2004), p. 486
54. P.R. Doorn, P.A. Campbell, J. Worrall, P.D. Benya, H.A. McKellop, H.C. Amstutz, Metal wear
particle characterization from metal on metal V total hip replacements: transmission electron
microscopy study of periprosthetic tissues and isolated particles. J. Biomed. Mater. Res. 42,
103–111 (1998)
55. I. Catelas, J.B. Medley, P.A. Campbell, O.L. Huk, J.D. Bobyn, Comparison of in vitro with
in vivo characteristics of wear particles from metal-metal hip implants. J. Biomed. Mater. Res.
Part B Appl. Biomater. 70, 167–178 (2004)
56. R. Büscher, G. Täger, W. Dudzinski, B. Gleising, M.A. Wimmer, A. Fischer, Subsurface
microstructure of metal-on-metal hip joints and its relationship to wear particle generation.
J. Biomed. Mater. Res. Part B Appl. Biomater. 72, 206–214 (2005)
57. F. Billi, P. Campbell, Nanotoxicology of metal wear particles in total joint arthroplasty: a review
of current concepts. J. Appl. Biomater. Biomech. 8, 1–6 (2010)
58. P. Stemmer, R. Pourzal, Y. Liao, L. Marks, M. Morlock, J.J. Jacobs, M.A. Wimmer, A. Fischer,
Microstructure of retrievals made from standard cast HC-CoCrMo alloys, in ASTM STP 1560
Metal-on-Metal Total Hip Replacement Devices, ed. by S.M. Kurtz, A.S. Greenwald, W.M.
Mihalko, J.E. Clemson (ASTM, West Conshohocken, 2013), pp. 251–267
59. R. Pourzal, I. Catelas, R. Theissmann, C. Kaddick, A. Fischer, Characterization of wear particles
generated from CoCrMo alloy under sliding wear conditions. Wear 271, 1658–1666 (2011)
60. R. Pourzal, Possible pathways of particle formation in CoCrMo sliding wear. Ph.D.-thesis
University Duisburg-Essen, Duisburg, Germany, VDI Verlag, s.a. Fortschr Ber VDI Z, 17(285),
Düsseldorf, Germany (2011)
2 Analyzing Mild- and Ultra-Mild Sliding Wear of Metallic … 59

61. Y. Liao, L. Marks, Direct observation of layer-by-layer wear. Tribol. Lett. 59, 1–11 (2015)
62. P. Stoyanov, P. Stemmer, T.T. Järvi, R. Merz, P.A. Romero, M. Scherge, M. Kopnarski, M.
Moseler, A. Fischer, M. Dienwiebel, Friction and wear mechanisms of tungsten-carbon sys-
tems: a comparison of dry and lubricated conditions. ACS Appl. Mater. Interfaces 5, 6123–6135
(2013)
63. R. Büscher, B. Gleising, W. Dudzinski, A. Fischer, Transmission electron microscopy exami-
nations on explanted metal-on-metal hip joints. Prakt. Metallogr./Pract. Metallogr. 42, 15–34
(2005)
64. M. Hahn, Mikrostrukturelle Veränderungen in der Zylinderlaufbahn von PKW Dieselmotoren
aus Grauguss und mittels thermischer Spritzverfahren hergestellter Stahlschichten. Disser-
tation Universität Duisburg-Essen, 2013s.a. Fortschr.-Ber. VDI Reihe 5: Grund- und Werk-
stoffe/Kunststoffe, Nr. 750 (VDI-Verlag, Düsseldorf, Germany, 2013)
65. P. Stemmer, The divergent pathways and mechanisms of energy dissipation at the interfaces
of martensitic tribocouples. Ph.D.-thesis, Materials Science and Engineering, University of
Duisburg-Essen, Germany, DuEPublico ID: 42437 (2016)
66. I. Catelas, J. Dennis Bobyn, J.B. Medley, J.J. Krygier, D.J. Zukor, A. Petit, O.L. Huk, Effects
of digestion protocols on the isolation and characterization of metal-metal wear particles. I.
Analysis of particle size and shape. J. Biomed. Mater. Res. 55, 320–329 (2001)
 Part II
Tribochemical Characterization
Techniques
Chapter 3
Near Edge X-Ray Absorption Fine
Structure Spectroscopy: A Powerful Tool
for Investigating the Surface Structure
and Chemistry of Solid Lubricants

Filippo Mangolini and J. Brandon McClimon

Abstract Synchrotron-based spectroscopic techniques have been critical tools for

developing a better understanding of the structure and properties of materials and
material surfaces as well as their evolution in response to energetics inputs, such as
mechanical strains present in tribological contacts. Among these techniques, near
edge X-ray absorption fine structure (NEXAFS) spectroscopy is one of the most
powerful tools thanks to its elemental specificity, surface sensitivity, and ability to
provide important information about local bonding configurations, such as hybridiza-
tion, chemical states, and bond orientations. In addition, when coupled with imaging
methods like photoemission electron microscopy and magnetically-guided imag-
ing, NEXAFS spectroscopy enables chemical imaging of materials with high spatial
resolution. This capability can be critical when investigating materials after tribo-
logical experiments, where chemical changes and structural transformations occur
in the first few atomic layers and spatial inhomogeneities can be present across small
length scales. The present contribution first describes the principles of NEXAFS
spectroscopy, followed by experimental methods for the acquisition and processing
of NEXAFS data. Finally, the potential of this analytical method for fundamental
and applied research in tribology is demonstrated by discussing case studies in the
area of solid lubricating carbon-based thin films.

F. Mangolini (B)
Department of Mechanical Engineering, Materials Science
and Engineering Program, The University of Texas
at Austin, Austin, TX 78712, USA
e-mail: Filippo.Mangolini@austin.utexas.edu
J. B. McClimon
Department of Materials Science and Engineering, University of Pennsylvania,
Philadelphia, PA 19104, USA
e-mail: mcclimon@seas.upenn.edu

© Springer Nature Switzerland AG 2018 63

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_3
64 F. Mangolini and J. B. McClimon

3.1 Introduction

Fundamental research in tribology hinges on the development of novel experimental

methods capable of assessing, characterizing, and monitoring the phenomena occur-
ring at buried sliding interfaces. It is exactly thanks to instrumental developments
and progress in research that over the last four decades the state of knowledge in
tribology has been shifting from empirically-based to scientifically-based. In par-
ticular, advancements in surface analysis have allowed researchers in tribology to
systematically investigate the local structure, chemistry, and mechanics of materials
after tribological interactions take place, and correlate the resulting information with
the observed tribological behavior, thus providing new and exciting opportunities to
investigate the fundamentals of friction, wear, and lubrication [1–3].
Among the tools and techniques that allow for the exploration of the chemical,
structural, and electronic properties of materials, in the last four decades synchrotron-
based methods have been of utmost importance [4, 5]. In particular, the development
of new generations of synchrotron radiation facilities (e.g., third generation syn-
chrotrons), which are brilliance-optimized (i.e., they provide a beam with small
cross section and a high degree of collimation), have tremendously expanded the
material characterization toolbox and provided flexibility in the development of syn-
chrotron endstations [6]. This has effectively given researchers the possibility of
simultaneously accessing several advanced analytical methods in the same facility.
Among the several synchrotron-based techniques developed for the analysis of
the properties of materials, surfaces, and interfaces, X-ray absorption fine structure
(XAFS) spectroscopy is a unique tool for investigating, at the atomic and molecular
scale, the local structure around a certain element contained in a material [7–10].
Traditionally, the XAFS signal is distinguished in two regions [11]: (a) the near edge
region, which is within 30–50 eV above the absorption edge and usually referred
to as X-ray absorption near edge structure (XANES) or near edge X-ray absorption
fine structure (NEXAFS). The term XANES is more commonly employed for solids
and inorganic complexes, while the term NEXAFS is preferentially employed for
surface studies [5]; and (b) the fine structure region, which extends from the near
edge region up to typically one thousand eV above the absorption edge and is usually
referred to as extended X-ray absorption fine structure (EXAFS).
While the EXAFS signals are dominated by single scattering contributions that
can be used for gaining information about the local geometric structure surround-
ing a given atomic species [8, 10, 12, 13], the absorption features characterizing
the NEXAFS region are sensitive to changes in valence state, coordination chem-
istry, local structure, and ligand symmetry around the absorbing atom. One of the
most important applications of NEXAFS spectroscopy since its introduction in the
1980s is the study of low atomic number (low-Z, i.e., carbon, oxygen, nitrogen,
and fluorine) molecules and materials [5]. The capability of NEXAFS spectroscopy
to provide information about low-Z elements derives from the strong directionality
and short length of the covalent bonds between low-Z atoms, the strong dependence
of the bond length on its hybridization, and the large backscattering amplitude of
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 65

low energy electrons from low-Z atoms [5]. Because of this, resolvable, structure-
dependent resonances are usually detected in NEXAFS spectra of low-Z elements,
whose intensity is strongly affected by the orientation of the final state orbital with
respect to the electric field vector of the incident photon beam [5].
Thanks to this elemental specificity, ability to obtain important information about
local bonding environment (such as hybridization, chemical state, and bond orien-
tation [5]), as well as the possibility to obtain spectra with high spatial resolution
using imaging techniques such as photoemission electron microscopy [14–17] and
magnetically-guided imaging [18], NEXAFS spectroscopy has become an attrac-
tive analytical tool for several research fields, such as tribology [18–20], catalysis
[21–25], self-assembly at surfaces [26–29], nanomaterials [30–38], and polymer
science [39–44].
The present contribution describes the basics of NEXAFS spectroscopy and the
experimental methods for the acquisition and processing of NEXAFS data. Selected
examples of application of this analytical method for fundamental and applied
research in tribology is demonstrated by discussing case studies in the area of solid
lubricating thin films, with a particular focus on carbon-based coatings.

3.2 Principles of Near Edge X-Ray Absorption Fine

Structure (NEXAFS) Spectroscopy

To gain an understanding of the spectral features characterizing NEXAFS data, a

solid theoretical basis of the principles of this spectroscopic technique is a prereq-
uisite. In the following, a brief and simplified description of the basics of NEXAFS
spectroscopy is presented. For a more comprehensive overview, the reader can refer
to selected, focused monographs [4, 5].
When a collimated X-ray beam with flux Φ 0 (number of photons per unit time
and unit cross-section) passes through a sample of thickness t, the flux is reduced
according to Beer-Lambert’s Law:

Φ  Φ0 exp(−μ(ω)t) (3.1)

where Φ is the flux of the transmitted X-ray beam, and μ(ω) is the linear attenua-
tion coefficient, which depends on the energy of the X-ray photons and the sample
composition and density.
For X-ray photons with energy between 0.1 and 40 keV (normally employed
in XAFS experiments), two different mechanisms lead to X-ray attenuation, i.e,
photoelectric absorption and scattering (coherent and incoherent). However, since
photoelectric absorption dominates over other interactions (Fig. 3.1a) in the energy
range from 0.1 to 40 keV, the attenuation coefficient in (3.1) can be approximated
with the photoelectric absorption coefficient.
66 F. Mangolini and J. B. McClimon

(a) 105 (c)

Molybdenum Photoelectric Abs.
π∗
NEXAFS
Cross Section (cm2 g-1)
Coherent Scat.
Incoherent Scat.
103 Coh. + Incoh. Scat.
Total Cross Section
spectrum

Intensity
101 σ∗
X-ray
10-1 photon
Photon
10-3 2 3 4 56 2 3 4 56 Energy
1 10 100 Valence States
Photon Energy (keV) (MO’s)
(b) 105 1s 2s
1s edge
Binding Energy (eV)

104 2p3/2 edge

Continuum
States
3
10
3d5/2 edge
σ π π∗
102 σ∗
Rydberg Ionization
101 States
0 10 20 30 40 50 60 70 Potential
Atomic Number (-) Potential Energy

(d) Auger Electron (e) 1.0

1s edge
Fluorescence Yield

Vacuum 0.8
Fermi Level
Valence Band 0.6
Fluorescent
2p or LII,III photon 0.4

0.2 2p3/2 edge

2s or LI
Core
hole 0.0
1s or K 0 10 20 30 40 50 60 70
Atomic Number (-)

Fig. 3.1 a X-ray absorption cross-section (photoelectric absorption together with coherent and
incoherent scattering and their sums) for molybdenum (data taken from the NIST XCOM database).
The photoelectric absorption exhibits four edges: the 2p3/2 (LIII ) at 2.52 keV, the 2p1/2 (LII ) 2.63 keV,
the 2 s (LI ) at 2.87 keV, and the 1 s (K) at 19.99 keV; b binding energy of the levels 1s (K), 2p3/2 (LIII )
and 3d5/2 (MV ) as a function of atomic number; c absorption of an X-ray photon with excitation
of a core electron to an unoccupied molecular orbital; d de-excitation mechanisms following the
emission of a core electron, i.e., emission of an Auger electron or X-ray photon (fluorescence); e
average fluorescence yield as a function of the atomic number. Data taken from NIST X-ray Data
Booklet

Even though the photoelectric absorption coefficient is overall a smooth function

of the energy of the incident X-rays, sharp discontinuities (absorption edges) occur at
well-defined photon energies and correspond to the extraction of an electron from a
core level. For example, K edges, which are photoabsorption edges with the highest
energy for a given element, result from the emission of electrons from 1s levels.
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 67

Table 3.1 Relationships between X-ray notation, core level notation, and quantum numbers for K,
L, and M edges
X-ray Core level Quantum numbers
notation notation
n l s j  |l + s|
K 1s1/2 1 0 +1/2; −1/2 1/2
L1 2s1/2 2 0 +1/2; −1/2 1/2
L2 2p1/2 2 1 −1/2 1/2
L3 2p3/2 2 1 +1/2 3/2
M1 3s1/2 3 0 +1/2; −1/2 1/2
M2 3p1/2 3 1 −1/2 1/2
M3 3p3/2 3 1 +1/2 3/2
M4 3d3/2 3 2 −1/2 3/2
M5 3d5/2 3 2 +1/2 5/2
Quantum numbers: n principal quantum number; l orbital angular momentum; s spin angular
momentum; j total angular momentum (equal to |l + s|)

The relationship between the X-ray notation for high-energy edges, core electronic
levels, and quantum numbers is provided in Table 3.1. The fact that the edge energy
varies with the atomic number as shown in Fig. 3.1b makes NEXAFS spectroscopy
element-specific.
The absorption edge is the result of the excitation of a core electron to a con-
tinuum or quasi-continuum of final states. Since the edge energy is related to the
oxidation state of the absorbing atom (it raises by several eV per oxidation unit),
the edge position can be used to determine the valence state of the photoabsorber
[45]. Additionally, the shape of the edge can provide valuable information about the
chemical environment and ligand geometry, thus making it useful for fingerprinting
molecular species [46, 47]. Around the ionization potential (IP), which is usually
superposed on transitions into Rydberg states, resonant transitions, which originate
from the excitation of an electron from a core level to unoccupied molecular states,
are detected. These transitions are superposed on a step-like absorption edge and
only occur when the energy of the incoming X-ray photon matches the energy differ-
ence between the core level and the unoccupied molecular level. Empty molecular
orbitals can be labeled according to their symmetry, i.e., π ∗ or σ ∗ orbitals. In the case
of π -bonded diatomic subunits, the lowest unoccupied molecular orbital (LUMO)
is usually a π ∗ orbital, while σ ∗ orbitals are normally detected at higher photon
energies. While π ∗ resonances are observed below the ionization threshold because
of electron-hole Coulomb interactions [5], σ ∗ spectral features are most often found
above the vacuum level for the neutral molecule. A schematic showing the origin of
NEXAFS features for a diatomic molecule is provided in Fig. 3.1c together with the
effective electrostatic potential and the corresponding K-shell spectrum.
The width of the resonances detected in experimental NEXAFS data is a con-
volution of the natural width of the core hole (or lifetime of the excited state), the
68 F. Mangolini and J. B. McClimon

resolution of the instrument, as well as the vibrational motion of the molecule [48].
Compared to π ∗ resonances, transitions corresponding to orbitals with σ -symmetry
are broader because of their large overlap with continuum states, which increases the
decay probability of the electron into continuum states and decreases the lifetime of
the core hole. As a rule of thumb, the higher is the energy of a resonance, the larger
is its linewidth. σ ∗ resonances are often characterized by distinct high-energy tails,
which derive from molecular vibrations along the bond direction.
The energy of σ ∗ resonances can be used to obtain information about bond lengths
in organic molecules through the so-called bond length with a ruler method [5,
49–54], which relies on the determination of the core level IP together the position of
the σ ∗ transition relative to the IP. Even though there is empirical evidence supported
by theoretical predictions [54, 55] for the dependence of the position of σ ∗ resonances
on the bond length, the validity of the use of these spectral features for bond length
determination has been controversial [55–59].
Weak and sharp resonances are often detected between the characteristic π ∗ spec-
tral features and the IP. These resonances are assigned to Rydberg orbitals. For the
case of strongly chemisorbed molecules, pure Rydberg resonances are quenched
due to their large spatial extent (they have most of their density at the periphery of
the molecule). On the other hand, in the case of hydrocarbons or molecules with
carbon-hydrogen bonds, Rydberg orbitals mix with hydrogen-derived antibonding
orbitals with the same symmetry, thus leading to an increase of the intensity of the
corresponding absorption feature [60, 61].
Finally, multi-electron features (also called shake-up structures, analogous to
those observed in X-ray photoelectron spectroscopy, XPS) can also be detected in
NEXAFS spectra. While these features are observed at lower kinetic energies (or
higher binding energies) than the main photoelectron signal in XPS spectra, they are
detected at higher photon energy than the primary resonance in X-ray absorption
data.
The absorption process results in a photoelectron and a core hole. An atom with
a core hole is unstable, and thus spontaneously tends to relax by filling the core hole
with an electron from an upper level. The resulting relaxation energy can be released
in two ways: (a) through the emission of X-ray photons (fluorescence); and (b)
through the emission of an Auger electron (Fig. 3.1d). Both channels can be used for
measuring the formation of a core hole following the absorption of X-rays and, thus,
are a measure of the absorption cross section. The experimental methods used for
detecting either photoemitted Auger electrons or fluorescent photons are described
in Sect. 3.3.1.1. It is worth mentioning that these two de-excitation mechanisms are
in competition. Their relative strength is usually referred to as fluorescence yield ηi :
Xi
ηi  (3.2)
X i + Ai

where the subscript i represents a defined absorption edge (e.g., K or 1s), while X i
and Ai are the emission probabilities of a fluorescent photon and an Auger electron,
respectively. The fluorescence yield depends on the atomic number (Fig. 3.1e). In
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 69

particular, the Auger electron yield is much higher than the fluorescence yield for
low-Z elements (e.g., carbon).
Even though X-ray photons penetrate deeply into the sample (several microns),
electron detection provides much higher surface sensitivity compared to fluorescence
yield due to the short mean free path of electrons in condensed matter (typically less
than 2 nm for energies between 250 and 600 eV) [62, 63]. The information depth
for electron yield detection, which is the specimen thickness measured normal to the
surface from which a specified percentage of typically 95% of the detected signal
originates, is thus usually less than 5 nm [5]. A detailed description of the influence
of the detection mode on the resulting surface sensitivity is reported in Sect. 3.3.1.1.
Since synchrotron radiation is polarized, NEXAFS spectroscopy can be used to
gain information about the orientation of molecules and molecular fragments [5,
26–28, 30, 31, 64, 65]. Through the investigation of the dependency of the intensity
of a given resonance on the X-ray incidence angle (Θ) or, in other words, the direction
of the electric field vector of the impinging X-rays, insights into the spatial orientation
of the corresponding orbital (i.e, the direction of maximum amplitude of the excited
atomic species) can be obtained.
Under the assumption of linearly polarized radiation, the equation that relates the
initial (Ψ i ) and final (Ψ f ) states to the absorption cross section (σ x ) can be derived
from a quantum mechanical description of the excitation process for a single electron
in the dipole approximation (Fermi’s Golden Rule):
 2
σx  Ψ f |e · p|Ψi  ρ f (E) (3.3)

where e is the unit electric field, p the dipole transition operator, and ρ f (E) the density
of final states [5].
In the case of linearly polarized radiation, the matrix element can be simplified
           
 Ψ f e · p|Ψi 2  e Ψ f  p|Ψi 2   Ψ f  p|Ψi 2 , where Ψ f  p|Ψi  is known as
transition dipole moment (TDM). In the case of a 1 s (K) initial state and a directional
final state, the transition intensity becomes:
 2
I ∝ eΨ f | p|Ψ1s  ∝ |e · O|2 ∝ cos2 δ (3.4)

where δ is the angle between the electric vector (E) and the TDM direction. There-
fore, the resonance intensity is maximum when the electric field vector is along the
direction (O) of the final state orbital (or the TDM direction). Note that σ ∗ orbitals
have a maximum amplitude along the bond axis, while π ∗ orbitals have maximum
amplitude normal to the bond direction.
70 F. Mangolini and J. B. McClimon

Fig. 3.2 Schematic of a synchrotron beamline for NEXAFS spectroscopy

3.3 Experimental Methods in NEXAFS Spectroscopy

3.3.1 Data Acquisition

3.3.1.1 Synchrotron Beamline for NEXAFS Spectroscopy

Since NEXAFS spectroscopy requires monochromatic X-rays with tunable energy,

NEXAFS measurements are limited to synchrotron facilities. These facilities, which
generate electromagnetic radiation by accelerating electric charges (electrons) and
forcing them to move on a curved path, usually consist of a storage ring and a booster
(Fig. 3.2). The latter accelerates charged particles (electrons) close to the speed of
light before injecting them into the former, where they orbit. In the storage ring,
synchrotron radiation is generated in bending magnets, which keep electrons on a
closed path, and in wigglers and undulators placed in the linear sections [10, 66–68].
In the bending magnets and wigglers/undulators, the electrons are accelerated and
thus emit radiation. The emitted radiation covers a wide range of the electromag-
netic spectrum, with a broad peak in intensity at an energy that depends on the
energy of the electrons. This synchrotron light with a high photon flux is conveyed
to experimental chambers (endstations) through beamlines. The primary elements
of a beamline are the monochromator and optical mirrors. The former, together with
the exit slits, selects a single wavelength and thus provides a monochromatic X-ray
beam with high-brilliance to experimentalists. The latter collimates and focuses the
beam. Synchrotron X-ray sources have three main advantages compared to labora-
tory sources: (a) they provide monochromatic X-rays with tunable energy; and the
emitted radiation has (b) high brilliance and flux, as well as (c) a high degree of
polarization.
Before reaching the samples, the X-ray photon beam often passes through a gold
grid (with typical transmission around 85%). The electron yield signal from this grid
is used to monitor the beam intensity and becomes critical when processing NEXAFS
data (see Sect. 3.3.2).
Even though the most straightforward way to acquire X-ray absorption spectra
would be to measure the attenuation of a beam of X-rays transmitted through a
sample, this method would require thin samples since the penetration depth of X-rays
is only a few microns (depending on their energy). Additionally, this methodology
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 71

would not allow for assessing the structural, chemical, and electronic properties of
the near-surface regions of materials.
In light of this, the large majority of NEXAFS measurements is nowadays carried
out by detecting the electrons or fluorescent photons emitted from a bulk sample as
a consequence of the relaxation of the excited atom (see Sect. 3.3.2).
In the case of electron yield detection, three main modes are currently in use:
Auger yield, partial electron yield, and total electron yield. Auger yield detection
(AEY) uses an electron energy analyzer to count only those electrons with a defined
kinetic energy E a (corresponding to a specific Auger emission energy) chosen by
the experimenter. This leads to a very high signal to background ratio at the cost
of reduced signal intensity. However, during the emission from the sample surface,
some of the electrons will suffer an energy loss and emerge with a kinetic energy
lower than E a . Since the intensity of inelastically-scattered electrons follows the one
of elastically-scattered electrons, secondary electrons can be used to increase the
signal-to-noise. This is used in partial electron yield (PEY) detection mode, where
electrons of kinetic energy larger than a threshold energy E p are detected. This results
in boosting the signal intensity substantially, while also increasing the background.
PEY measurements are normally carried out by selecting the potential (also called
entrance grid bias, EGB) of the center grid of a standard three-grid high pass kinetic
energy filter for the channeltron electron yield detector [26]. Upon increasing the
values of the EGB voltage, only those Auger electrons that have suffered limited
energy loss during the emission from the sample and have sufficiently high kinetic
energy to reach the channeltron will be detected. Thus, the information depth from
which Auger electrons are detected can be adjusted by changing the EGB (the higher
the EGB, the smaller the information depth). Total electron yield (TEY) detection
is the simplest detection scheme, which relies on measuring the drain current from
an electrically isolated sample by means of a sensitive ammeter. Electrons of all
energies and from all X-ray penetrations depths are captured, so this scheme has
the lowest signal-to-background ratio and surface sensitivity. To summarize, the
surface sensitivity of electron yield detection modes increases in the following order:
AEY > PEY > TEY.
The fluorescence yield (FY) detection mode is not discussed in detail here because
of its limited surface sensitivity. The FY detection mode is, however, the method of
choice when investigating bulk samples or liquids, since the short mean free path
of electrons in these environments hinders the acquisition of data in electron yield
detection modes (either PEY or TEY).

3.3.1.2 Imaging Techniques

Photoemission Electron Microscopy (PEEM)

Photoemission electron microscopy (PEEM) is a full field imaging technique that

is commonly combined with X-ray synchrotron radiation to give a surface sensitive
spectroscopy with excellent spatial resolution (<50 nm [69], or even below 10 nm
72 F. Mangolini and J. B. McClimon

in aberration corrected instruments [70]) [14, 71–75]. Since the pioneering work of
Tonner and Harp [76], several instrumental development have been reported, which
have allowed for the use of PEEM in numerous fields, including tribology [14, 20,
77–82], thin film magnetism [83–85], catalysis [86], semiconductor [87], polymer
science [88–90], geology [91–93], and microbiology [94].
A typical PEEM microscope consists of a sample held at a high negative voltage
followed by a series of electron optics, and a detector (microchannel plane and phos-
phor screen). A schematic of this technique is shown in Fig. 3.3. Monochromatized
X-rays impinge on the sample surface at a grazing angle. This geometry is required
by the close proximity of the electron lens optics to the sample surface (a few mil-
limeters). The sample is held at a large negative bias (between −15 and −20 kV).
Because of this, specimens for PEEM measurements should be flat as any feature
on the sample surface can locally concentrate the electric field, lead to distortions of
the electron images, and, in the worst case, cause arcing between the sample and the
electron lens optics. The photoemitted electrons (primary, secondary, and Auger) are
accelerated by the applied field from the sample towards the electron optics column,
which can be either electrostatic or magnetic and often includes a series of deflectors
and stigmators. After the optics column, which accelerates, filters, and focuses the
electrons, they are detected using conventional channeltrons and phosphor screens.
Overall, the signal acquired in PEEM measurements is in most of the cases close
to NEXAFS data acquired in TEY detection mode, even though detection schemes
equivalent to PEY and AEY modes have been implemented by inserting retarding
grids or energy analyzers upstream of the image detector [95, 96]. Upon varying the
photon energy of the impinging X-rays in a similar manner to the approach used for
the acquisition of NEXAFS spectra, a stack of images is captured, where each pixel
represents a spatially-resolved NEXAFS spectrum.
Magnetically-Guided Imaging NEXAFS Spectroscopy
Magnetically-guided imaging NEXAFS spectroscopy combines NEXAFS spec-
troscopy, a full field parallel processing magnetic field electron yield optics detector,
and a large incident soft X-ray beam (Fig. 3.4a). This technique, which was pre-
viously described in [18, 97], relies on a rapid parallel processing magnetic field
electron yield optics detector, in which the emitted photoelectrons move along the
magnetic field lines towards the partial electron yield grid in front of the channel
plate electron detector. The detected electrons produce a series of two-dimensional
NEXAFS lateral images as the incident soft X-ray energy is scanned across the
selected absorption edge (Fig. 3.4b). The rejection of low energy electrons by the
partial electron yield detector allows for the mitigation of surface charging. The
image stack can be employed for the structural and chemical characterization of the
near-surface region of materials with a lateral resolution of 50 μm and a field of view
of 18 × 13 mm2 . Unlike photoelectron emission microscopy (PEEM), magnetically-
guided NEXAFS spectro-microscopy can handle insulators, conductors, and non-
planar samples without charging effects [18, 97]. In particular, the possibility of
analyzing curved surfaces provides significantly increased flexibility in the range of
samples that can be analyzed, while making this technique particularly attractive for
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 73

Fig. 3.3 Schematic of

electron optical path through Detector
a photoemission electron
microscope (PEEM).
Adapted from [73]

Projector Lens

Intermediate Lens

Stigmator/Deflector
X-rays Aperture

Objective Lens

Sample

the ex situ characterization of tribological samples, which often comprise a spherical

counterface.
As in the case of the acquisition of NEXAFS spectra in PEY detection mode,
changing the entrance grid bias (EGB) of the channel plate electron detector allows
magnetically-guided imaging NEXAFS measurements with different surface sensi-
tivity to be performed [26, 98, 99] (Fig. 3.4b).

3.3.1.3 Data Processing

The analysis of NEXAFS data is complicated by many experimental factors, includ-

ing photon energy calibration, intensity normalization, and artifacts due to beam
instabilities, signal offsets, and the beamline transmission function [5, 100–102]. In
particular, the presence of an adventitious carbon contamination on X-ray optics in
synchrotron beamlines makes the processing of carbon 1s NEXAFS spectra chal-
lenging, since the carbon contamination can result in artifacts in the as-collected
data, which can even dominate true NEXAFS spectra. The common approach to
remove major artifacts (including the beamline transmission function, flux monitor
contamination, beam stability and signal offsets) with the aim to deconvolute the
real NEXAFS spectrum for the sample under investigation relies upon normaliza-
tion techniques. These techniques are usually based on the use of reference spec-
tra acquired from metallic mesh materials (such as gold) placed in the beamline
upstream from the analysis chamber (see Sect. 3.3.1.1). Even though several meth-
ods of normalization, energy calibration, and artifact removal have been published in
74 F. Mangolini and J. B. McClimon

(a) Soft X-ray beam NEXAFS image

(15 x 20 mm 2) (3x magnification)

Channel plate,
phosphor,
CCD camera
22.5 μm/pixel

Sample (rot.)
1 Tesla shaped pole
High pass grid

(b)
Entrance grid bias (EGB) voltage
Surface sensitivity

Fig. 3.4 a Schematic of the parallel processing imaging NEXAFS system at the NIST/Dow endsta-
tion of beamline U7A at the National Synchrotron Light Source (Brookhaven National Laboratory,
Upton, NY, USA). A large incident soft X-ray beam (15 × 20 mm2 ) illuminates the sample surface.
The photoelectrons emitted from the specimen are magnetically guided onto a detector that forms a
two-dimensional NEXAFS image over a 13 × 18 mm2 region, while providing a spatial resolution
of 50 μm at all relevant X-ray incidence angles [18]; b principle of parallel processing imaging
NEXAFS measurements. The rapid parallel processing magnetic field electron yield optics detector
produces a series of two-dimensional NEXAFS lateral images as the incident soft X-ray energy
is scanned across an absorption edge (e.g., the carbon 1s absorption edge). Changing the entrance
grid bias (EGB) of the channeltron detector allows NEXAFS measurements with different surface
sensitivity to be performed [26, 98, 99]

the literature, they have been reported in a fragmented manner, thus hindering their
use and implementation.
Recently, Watts et al. reviewed and implemented the numerous methods for cali-
bration, photon flux normalization, and artifact removal for NEXAFS spectra (with
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 75

a particular focus upon carbon 1s spectra) reported in the literature [102]. Besides
providing criteria for assessing the most appropriate method for the photon flux nor-
malization, Watts et al. also presented the methodologies for the determination of
appropriate background functions along with a way for their removal from exper-
imental data. Additionally, the methods used for photon flux normalization were
discussed together with the criteria for assessing the most appropriate one for a
specific application.
While the approaches outlined by Watts et al. are effective for addressing the
issues commonly encountered in the evaluation of NEXAFS data [102], the corrected
spectra they yield represent the photo-absorption spectra of the specimens under the
assumption of structural and compositional homogeneity within the nanometer-scale
depth probed by NEXAFS spectroscopy (in the case of NEXAFS spectra of low-Z
materials acquired in electron yield mode, the information depth, which is the spec-
imen thickness measured normal to the surface from which a specified percentage
of typically 95% of the detected signal originates, is usually less than 5 nm [5]).
However, the assumption of chemical and structural homogeneity does not hold in
the vast majority of solid surfaces, since complex surface-bound species and layers,
e.g., natural oxide and contamination layers, are often present [103, 104]. Thus, this
assumption can lead to significant errors when analyzing elements that are simulta-
neously present in multiple layers. In the case of carbon-based materials previously
exposed to air, as a particular example, their carbon 1s NEXAFS spectra, even when
corrected using any of the approaches outlined by Watts et al. [102], are a convolution
of the spectrum of the sample of interest and the spectrum of the adventitious carbon
contamination on its surface since the thickness of the latter (typically <2 nm [105])
is smaller than the information depth at the carbon 1s.
The authors of this chapter recently developed a methodology for accurately
removing the contribution of thin overlayers (with thickness smaller than the infor-
mation depth) from NEXAFS spectra of two-layered systems to reveal the photo-
absorption NEXAFS spectrum of the substrate [99]. This method relies on the sub-
traction of the NEXAFS spectrum of the overlayer adsorbed on a reference surface
(e.g., gold) from the spectrum of the two-layer system under investigation, where
the thickness of the overlayer is independently determined by XPS. The approach
was applied to NEXAFS data acquired for one of the most challenging cases: air-
exposed hard carbon-based materials (namely ultrananocrystalline diamond (UNCD)
and hydrogenated amorphous carbon (a-C:H)) with adventitious carbon contamina-
tion from ambient exposure. The possibility of removing the contribution of the
adventitious carbon contamination from the as-acquired NEXAFS carbon 1s spectra
allowed the computation of the contamination-corrected photo-absorption NEXAFS
spectra for the materials under investigation (UNCD and a-C:H films) (Fig. 3.5),
which provided qualitatively distinct interpretations and quantitatively distinct values
regarding the sample’s composition and bonding compared to the as-acquired data.
In particular, the method, which also revealed that the adventitious contamination can
be described as a layer containing carbon and oxygen ([O]/[C]  0.11 ± 0.02) with a
thickness of 0.6 ± 0.2 nm and a fraction of sp2 -bonded carbon of 0.19 ± 0.03, signif-
icantly altered the calculated fraction of sp2 -hybridized carbon, thus demonstrating
76 F. Mangolini and J. B. McClimon

that the assumption of chemical and structural homogeneity could introduce large
errors in the computed carbon hybridization state (between 5 and 20%) when analyz-
ing carbon-based materials previously exposed to air. Additionally, upon removing
the contribution of the carbonaceous contamination from the as-acquired data, the
absorption features assigned to the C–O σ * antibonding orbital (at 288.5–289.3 eV)
[106] was greatly reduced (Fig. 3.5), which clearly indicated the importance of
accounting for nanometer-thick adventitious contamination when investigating the
surface chemistry of air-exposed carbon surfaces by NEXAFS spectroscopy.

3.4 Applications of NEXAFS Spectroscopy

3.4.1 NEXAFS Spectroscopic Investigation of the Surface

Structure and Chemistry of Thin Films for Tribological
Applications

As outlined above, NEXAFS spectroscopy can provide valuable information about

the chemical, structural, and electronic properties of materials and materials sur-
faces. Additionally, it can be effectively used for the determination of molecular
orientations. This wealth of detailed information NEXAFS yields has made it an
attractive analytical tool for fundamental and applied studies in tribology [18–20],
catalysis [21–25], self-assembly at surfaces [26–29], nanomaterials [30–38], and
polymer science [39–44].
One of the most important applications of NEXAFS spectroscopy is the study of
carbon-based materials. The strong directionality and short length of the covalent
bonds between carbon atoms together with the strong dependence of the bond length
on its hybridization and the large backscattering amplitude of low-energy electrons
from carbon atoms results in the detection of resolvable absorption resonances in
carbon 1s NEXAFS spectra. These absorption features are structure-dependent and
their intensity depends on the orientation of the final state orbital with respect to the
electric field vector of the impinging X-rays [5].
For carbon 1s NEXAFS spectroscopy, in particular, the resolvable energy differ-
ence between the resonant X-ray excitations of a core-level (1s) electron to unoc-
cupied molecular orbitals (either π * or σ * ) allows the identification of the bonding
configuration and hybridization state of carbon atoms in the near-surface region for
many materials, including diamond [18, 20, 33, 34, 36, 37, 77, 80, 82, 107–109],
diamond-like carbon [78, 110–128], graphene [38, 129, 130], carbon nanotubes [131,
132], and polymers [31, 39–44], as well as the determination of the surface molecular
orientation of nanomaterials [30, 31, 64, 65, 132] and adsorbates [26–28, 133].
As an example, the characteristic carbon 1s NEXAFS spectra (not corrected for
the contribution of the contamination layer following the method outlined in [99])
of two tribologically-relevant, as-deposited thin films, i.e., hydrogenated amorphous
carbon (a-C:H) and silicon- and oxygen-containing hydrogenated amorphous carbon
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 77
Non-corrected spectra
UNCD a-C:H (PECVD 180 V) a-C:H (PECVD 120 V) a-C:H (PECVD 60 V)

Corrected spectra
UNCD a-C:H (PECVD 180 V) a-C:H (PECVD 120 V) a-C:H (PECVD 60 V)

(a) 21 UNCD
(b) 3 UNCD
(c) 3 UNCD

C-O
18
15
12 2 2

C-H
9

C=C
6 1 1
3
0 0 0
21 a-C:H a-C:H a-C:H
18 PECVD 180 V 2 PECVD 180 V 2 PECVD 180 V
15
Partial Electron Yield (A.u.)

Partial Electron Yield (A.u.)

Partial Electron Yield (A.u.)

12
9 1 1
6
3
0 0 0
a-C:H
21 a-C:H PECVD 120 V a-C:H
18 PECVD 120 V 2 2 PECVD 120 V

15
12
9 1 1
6
3
0 0 0
a-C:H a-C:H
21 PECVD 60 V PECVD 60 V a-C:H
18 2 2 PECVD 60 V
15
12
9 1 1
6
3
0 0 0
280 290 300 310 320 280 290 300 310 320 280 285 290 295
Photon Energy (eV) Photon Energy (eV) Photon Energy (eV)

Fig. 3.5 Carbon 1s NEXAFS spectra (before and after the correction for the presence of a carbona-
ceous contamination layer) of UNCD and a-C:H films grown by PECVD with different acceleration
bias voltages (60, 120, and 180 V). a Pre-edge normalized spectra (whole photon energy scale).
The pre-edge normalization removes factors such as synchrotron ring current, beam dispersion
character, and monochromator profile. The spectra were then normalized to the absorption current
measured simultaneously from a gold mesh placed in the beamline upstream from the analysis
chamber. The pre-edge region was then subtracted to zero; b pre- and post-edge normalized spectra
(whole photon energy scale). The post-edge intensity depends on the number density of the absorb-
ing atoms. By normalizing the post-edge to unity, variations in spectral intensity only arise from
chemical changes and are independent of the total number density of absorbers; and c pre- and
post-edge normalized spectra (zoomed absorption edge region). Upon removing the contribution
from the carbonaceous contamination layer, the intensity of the C 1s→π * transition for disordered
carbon− carbon bonds [5] at 285.0 ± 0.1 eV slightly decreased in the case of UNCD, whereas it
increased for a-C:H films. Very significantly, for the a-C:H films, the broad absorption feature at
288.5–289.3 eV (assigned to the C–O σ * antibonding orbital [106]) was greatly reduced. The spec-
tral changes induced by the removal of the contribution of the carbonaceous contamination from
the as-acquired data substantially affected the computation of the carbon hybridization state. From
[99]

(a-C:H:Si:O), are shown in Fig. 3.6. More details about these thin films can be found
in [134], respectively. For comparison, the NEXAFS spectra of ultrananocrystalline
78 F. Mangolini and J. B. McClimon

diamond (UNCD) and highly ordered pyrolytic graphite (HOPG) are also displayed.
The NEXAFS spectrum of as-deposited a-C:H, a-C:H:Si:O, and UNCD exhibits an
absorption feature at 285.0 ± 0.1 eV, which is due to the C 1s→π * transition for
disordered carbon–carbon bonds [5, 118]. The width of this C 1s→π * absorption
feature provides information about the bond length distribution of sp2 carbon–carbon
bonds: the broader the peak, the larger the bond length distribution. It should be noted
that in the case of HOPG, the C 1s→π * feature is detected at 285.5 ± 0.1 eV. Thus,
the position of the C 1s→π * transition can be used to gain insights into the ordering
of sp2 -bonded carbon atoms in amorphous carbon materials: the higher the photon
energy of this spectral feature, the more ordered the sp2 -bonded carbon phase [112].
A broad hump between 288 and 310 eV, which is due to the C 1s→σ * transition for
disordered carbon–carbon σ bonds [5, 118], characterizes the NEXAFS spectrum
of as-deposited amorphous carbon materials (Fig. 3.6a), whereas the spectrum of
UNCD exhibits sharper C 1s→σ ∗ transitions that are characteristic of ordered sp3 -
hybridized carbon–carbon bonds, namely the edge jump at ~289 eV, the exciton
peak at ~289.3 eV, and the second band gap at 303 eV [20] (Fig. 3.6b). Similarly,
in the case of HOPG sharp C 1s→σ ∗ features are detected at 291.8 ± 0.2 eV and
292.8 ± 0.3 eV [135]. The presence of a significant amount of carbon–hydrogen
bonds in hydrogenated amorphous carbon materials and the hydrogen-termination of
UNCD result in the detection of a shoulder at ~287.0 eV (for a-C:H and a-C:H:Si:O)
and ~287.5 eV (for UNCD), which can be assigned to the C 1s→σ * transition for C–H
bonds [5, 80, 99]. The observed shift of the characteristic C–H absorption feature
between the spectra of a-C:H and the spectrum of UNCD may be due to the different
bonding states in these materials: while hydrogen terminates primarily sp3 -bonded
carbon atoms at the UNCD surface, it is present in a range of bonding environments
within the bulk of hydrogenated amorphous carbon films. Due to the sample exposure
to air before the NEXAFS analysis, a broad absorption feature is usually detected at
288.8 ± 0.2 eV and could be assigned to the C 1s→σ * transition for carbon–oxygen
bonds, as well as to the C 1s→π * transition for carboxyl groups [5, 106, 136]. In the
case of a-C:H:Si:O, the characteristic C 1s→σ * transition for carbon–silicon bonds
[136] contributes to the spectral feature at 288.8 ± 0.2 eV. The presence of carbonyl
groups on the surface of a-C:H and a-C:H:Si:O due their exposure to air also results
in a weak shoulder at 286.7 ± 0.2 eV (mainly due to the C 1s→π ∗ transition for
C=O bonds and with a small contribution from the C–O Rydberg orbitals [5, 106])
(Fig. 3.6b).
Besides providing valuable information about the surface chemistry of carbon-
based materials, NEXAFS spectroscopy can also be used for the quantitative eval-
uation of the hybridization state of carbon atoms [137]. The characterization of
the carbon bonding configuration in carbon-based materials, which is of paramount
importance in the case of carbon-based materials due to the strong dependency of the
tribological, mechanical, electrical, and optical properties on the carbon hybridiza-
tion state [138], is a particularly challenging materials science problem, which has
resulted in the use of some of the most powerful weapons in the materials character-
ization arsenal for the analysis of these materials, including NEXAFS spectroscopy
[99, 111–113, 119, 122, 126, 138–142], Raman spectroscopy [113, 138, 143–146],
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 79

(a) C1s→π C1s→σ (b)

3 UNCD 3 UNCD
2 2 C-H
1 1 C=C
0 0 C-O, C-Si, COOH
C=O C-H
Partial Electron Yield (A.u.)

Partial Electron Yield (A.u.)

2 a-C:H:Si:O 2 C=C

1 1
a-C:H:Si:O
0 0 C-O, COOH
C=O C-H
2 a-C:H 2 C=C

1 1
a-C:H
0 0 C=C
5 HOPG 5 HOPG
4 4
3 3
2 2
1 1
0 0
280 300 320 340 280 285 290 295 300 305
Photon Energy (eV) Photon Energy (eV)

Fig. 3.6 a Carbon 1s NEXAFS spectra of HOPG acquired at an X-ray incidence angle (relative
to the sample surface) of 40° together with carbon 1s NEXAFS spectra of a-C:H, a-C:H:Si:O, and
UNCD acquired at an X-ray incidence angle of 55°; b zoomed view of the absorption edge region
of the C 1s spectra. The spectra were not corrected for the contribution of the contamination layer
following the method outlined in [99]

XPS [99, 121, 134, 146–153], Auger electron spectroscopy (AES, including X-ray
induced AES, XAES) [134, 147, 149–151, 153–155], electron energy-loss spec-
troscopy (EELS) [145, 156–158], Fourier-transform infrared spectroscopy (FT-IR)
[159], nuclear magnetic resonance (NMR) spectroscopy [160, 161], and X-ray reflec-
tivity (XRR) [145]. Among the surface-sensitive techniques, electron spectroscopies
(XPS, AES, XAES, NEXAFS, and EELS) are the mostly widely used for the char-
acterization of carbon-based materials.
The characterization of the bonding configuration of carbon by XPS is usually car-
ried out through the acquisition and fitting of the C 1s spectrum [121, 134, 146–149,
154, 155]. However, the validity of the methodology for the quantitative evaluation
of the hybridization state of carbon on the basis of the C 1s signal, which relies on
fitting it with two distinct features for sp2 - and sp3 -bonded carbon [121, 146, 154],
has recently been questioned [134, 147], since the binding energy values of the C
1s transition for graphite (100% sp2 -bonded carbon), diamond (100% sp3 -bonded
carbon), and UNCD (94 ± 1% sp3 -bonded carbon [99]) are not significantly different
[134, 147]. However, insights into the carbon hybridization state can be gained by
XPS through the analysis of the plasmon band near the C 1s signal [134, 152], the
π –π * shake-up satellites [162, 163], or the X-ray induced C KVV Auger spectrum
[134, 147, 150, 151, 153, 155].
80 F. Mangolini and J. B. McClimon

Besides XPS, EELS and NEXAFS spectroscopy are effective methods for the
determination of the hybridization state of carbon atoms in carbon-based films [138].
Even though EELS allows the fraction of carbon atoms in sp2 - and sp3 -hybridization
state to be quantified through the analysis of the low loss region (from 0 to 40 eV) or
the high loss region at the C 1s [138, 145, 156–158], it requires the film to be removed
from the substrate or for a cross-section to be produced. Additionally, the acquisition
of the C 1s is usually carried out with limited energy resolution of approximately
0.5 eV [138, 164].
NEXAFS spectroscopy overcomes these two limitations of EELS since spectro-
scopic data are acquired with an energy resolution lower than 0.1 eV (depending
on the spectrometer) and no sample preparation is usually needed. Even though
NEXAFS spectroscopy has been extensively applied to quantitatively evaluate the
hybridization state of carbon atoms in the near-surface region, a critical assessment of
the methodology for the quantification of the local bonding configuration of carbon
on the basis of NEXAFS data was lacking.
The authors of the present contribution recently reviewed and critically assessed
the common methodology for quantitatively evaluating the carbon hybridization state
using carbon 1s NEXAFS measurements [137]. This method, which is based on the
approach developed for the determination of the carbon hybridization state from the
carbon core loss 1s edge in EELS [156, 165] and qualitatively supported by Car-
Parrinello ab initio molecular dynamics simulations [166], considers the relative
intensity of the C 1s→π * and C 1s→σ * absorption features in NEXAFS spectra of
carbon-based materials. Since no theory exists to predict the π * /σ * ratio, a reference
material (usually HOPG) with known sp2 content is required for quantitative analysis.
The equation used to compute the fraction of sp2 -bonded carbon ( f sp2 ) is:
∗
π
Isam /Isam (E)
f sp2  π∗ (3.5)
Iref /Iref (E)
∗ ∗
π π
where Isam and Iref are, respectively, the areas of the C 1s→π * peaks for the sample
and the reference (which arise exclusively from sp2 bonds), whereas Isam (E) and
Iref (E) are the areas under the NEXAFS spectrum between two integration limits
(x 1 and x 2 ) for the sample and the reference, respectively. A graphical representation
of the methodology is shown in Fig. 3.7.
A survey of the published literature revealed inconsistencies in applying this
method. First of all, a wide range of integration limit values (x 1 and x 2 ) has
been used for computing the area of the C 1s→σ * transition (289–295 eV [123],
294–301 eV [111, 122], 293–302 eV [139], 288.6–325 eV [80], 289–325 eV [20],
and 288.6–320 eV [99]). Secondly, HOPG reference spectra acquired at different
X-ray beam incidence angles have been used in the literature (for example, HOPG
spectra were acquired at 55° with respect to the sample surface [80], while others
were acquired at 45° [99, 139]). This is particularly critical in the case of HOPG due
to the strong dependence of the π * /σ * ratio on the orientation of the basal planes
with respect to the incoming X-ray beam [5] (Fig. 3.8a).
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 81

C1s→π C1s→σ

2.0

1.5

1.0 a-C:H

Partial Electron Yield (A.u.)

0.5

0.0

4.0

3.0

2.0
HOPG
1.0

0.0
270 280 290 300 310 320 330 340
x1 Photon Energy (eV) x2

Fig. 3.7 Carbon 1s NEXAFS spectra of a-C:H and HOPG. Spectra are pre- and post-edge nor-
malized. The NEXAFS spectrum of a-C:H was acquired at a photon incidence angle of 55° with
respect to the sample surface (the so-called “magic angle”) to suppress the effects related to the
X-ray polarization [5], whereas the NEXAFS spectrum of HOPG was collected with the X-ray
beam incident at an angle of 45° to the sample surface. The quantitative evaluation of the carbon
hybridization state in carbon-based materials is based on the relative intensity of the C 1s→π * and
C 1s→σ * absorption features (red- and green-shaded areas, respectively). A reference material with
a known fraction of sp2 -bonded carbon (e.g., HOPG) is required for quantitative analysis. The area
of the C 1s→π * transition is determined by fitting this feature with a Gaussian synthetic peak [80,
99], whereas the area of the C 1s→σ * absorption feature is computed by numerically integrating
the spectrum between two limits x 1 and x 2 . From [137]

The analysis of the dependence of total resonance intensity for the π * and σ *
orbital on the angle of incidence of the impinging X-ray beam allowed the authors
to derive the experimental conditions (i.e., the critical X-ray incidence angle, θ c )
to be used for the acquisition of HOPG reference spectra to which the π * and σ *
states contribute equally (i.e., the conditions under which a molecular orbital oriented
normal to the substrate surface contributes the same NEXAFS intensity as the orbital
oriented within the substrate plane).

−1 1
θc  cos (3.6)
2P

where P is the polarization factor in the plane of the electron beam orbit (P 
  2   2  ⊥ 2 
 E  /  E  +  E  ) [5], which is equal to 0.85 for the beamlines used in the
present work (NIST/Dow endstation of beamline U7A and at the Oak Ridge National
Laboratory endstation of beamline U12A at the National Synchrotron Light Source
(NSLS), Brookhaven National Laboratory, Upton, NY, USA) [26, 167]. In the case
82 F. Mangolini and J. B. McClimon

Fig. 3.8 a Carbon 1s (a)

NEXAFS spectra of HOPG X-ray incidence angle 6
6

Partial Electron Yield (A.u.)

acquired at different X-ray

PEY (A.u.)
30°
35° 4
incidence angles. Spectra are 5 40°
pre- and post-edge 45° 2
50°
normalized. Inset: zoomed 4 55° 0
view of the absorption edge 60° 280 285 290 295 300
3 70° Photon Energy (eV)
region of the C 1s spectra. 80°
90°
Spectra displayed without 2
any offset to allow for
comparisons; b ratio of the 1
integrated intensity of the C 0
1s→π * (computed by fitting 270 280 290 300 310 320 330 340
this signal with a Gaussian Photon Energy (eV)
synthetic peak) and C
1s→σ * (computed by
(b)
numerically integrating the 0.20
spectrum between 288.6 and
Ratio AC1s-π*/AC1s-σ*

320 eV) absorption features 0.15

as a function of the X-ray
incidence angle. Adapted
from [137] 0.10

0.05

0.00
30 40 50 60 70 80 90
X-ray Incidence Angle (º)

of linearly polarized X-rays (P  1), θ c is equal to 45°, while for elliptically polarized
X-rays with polarization factor P equal to 0.85 (as in the case of U7A and U12A),
θ c is equal to 40°. It should be noted that there are claims in the literature that this
angle should be the magic angle (i.e., 55° with respect to the sample surface). The
analysis reported in [137] demonstrated that this is not correct and provided a formula
for computing the X-ray incidence angles that should be used for acquiring HOPG
spectra to be employed as reference in the determination of the carbon local bonding
configuration from NEXAFS data (using (5)).
Since the π * /σ * ratio in HOPG spectra strongly depends on the X-ray incidence
angle (θ ) [5] (Fig. 3.8b), uncertainty in the angular position of the HOPG relative
to the impinging X-ray photons can lead to errors in the quantitative analysis. The
derivation of an analytical expression that describes the dependence of the π * /σ *
ratio on the X-ray incidence angle allowed for the calculation of the uncertainty
of in the fraction of sp2 -bonded carbon due to uncertainties in the X-ray incidence
angle (Fig. 3.9a, b). The uncertainty increases with both the fraction of sp2 -bonded
carbon (Fig. 3.9a) and the uncertainty in the angle of X-ray incidence (Fig. 3.9b). In
the case of the beamlines used in this work, the uncertainty in the X-ray incidence
angle was 1°, which translates into an uncertainty in the computed fraction of sp2 -
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 83

bonded carbon linearly increasing with the fraction of sp2 -bonded carbon up to 3.6%
for 100% sp2 -bonded carbon, which is modest. However, a larger uncertainty in X-
ray incidence angle of 5° produces a much more significant error of 8.9% for 50%
sp2 -bonded carbon, and 17.9% for 100% sp2 -bonded carbon. These findings clearly
demonstrated the importance of minimizing any uncertainty in the angular position
of HOPG relative to the impinging X-ray beam.
As pointed out above, a wide range of integration limits (x 1 and x 2 ) has been
used in the literature for computing the integrated intensity of the C 1s→σ * absorp-
tion feature in NEXAFS spectra. The investigation of the evolution of the computed
fraction of sp2 -bonded carbon for an a-C:H film as a function of the integration lim-
its used to calculate the area of the C 1s→σ * transition is shown in Fig. 3.9c, d.
For highlighting the influence of the integration limits on the results of the quan-
tification, the error in the calculated fraction of sp2 -bonded carbon relative to the
fraction obtained using the integration limits employed in [99] (x 1  288.6 eV; x 2 
320 eV) is also shown in Fig. 3.9c, d. The outcomes of this analysis indicated that
for the quantification of the carbon hybridization state using carbon 1s NEXAFS
spectroscopy, the low-photon-energy integration limit (x 1 ) can be chosen arbitrarily
between 286.6 and 295 eV, whereas the high-photon-energy integration limit (x 2 )
should be taken at photon energies above 310 eV. In fact, large variations, as large as
10%, in the computed fraction of sp2 -bonded carbon occurred upon decreasing the
high-photon-energy integration limit below 310 eV.

3.4.2 NEXAFS Spectroscopic Investigation

of the Thermally-Induced Structural Evolution of Solid
Lubricating Films

The thermally-induced structural evolution of amorphous carbon (a-C) thin films is a

subject of fundamental and technological interest [112, 168–170]. As one particular
example, the investigation of the effect of annealing temperature on the structure of
tetrahedral amorphous carbon (ta-C) has important implications regarding the use of
ta-C as an engineering material. In fact, thermal annealing at elevated temperatures
(~650 °C) for a period of minutes is necessary in the case of ta-C to relieve the high
compressive stress (2–8 GPa) typical for these films [112, 171, 172]. As another
example, knowledge of the temperature effects on the structure of a-C coatings,
which are used as overcoat material in hard-disk drives to protect the magnetic
medium against corrosion and wear [173], is fundamental for developing one of
the most promising solutions for increasing the magnetic storage density, namely
heat-assisted magnetic recording (HAMR) [174]. The intensive localized heating
employed in HAMR for data encoding raises concerns about the structural changes
occurring in a-C overcoats.
Grierson et al. investigated the thermally-induced evolution of the structure of ta-
C by NEXAFS spectroscopy [112]. The results provided evidence that sp3 -bonded
84 F. Mangolini and J. B. McClimon

(a) (b)
18 18
Uncertainty in Fraction of

Uncertainty in Fraction of
Uncertainty of the Nominal fraction

sp2-bonded Carbon (%)

sp2-bonded Carbon (%)

16 angle of X-ray incidence 16 of sp 2-bonded carbon
14 1º 14 1
12 2º 12 0.75
3º 0.5
10 4º 10 0.25
8 5º 8
6 6
4 4
2 2
0 0
0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5
Nominal Fraction Uncertainty of the X-ray
sp2-bonded Carbon Incidence Angle (°)

(c) (d)
x1=288.6 eV x2=320 eV
x1=288.6 eV (%)
Error Relative to

Error Relative to
10 10

x2=320 eV (%)
5 5
0 0
-5 -5
-10 -10
-15 -15
0.60 0.60
Fraction sp2-bonded C

Fraction sp2-bonded C

0.55 0.55

0.50 0.50

0.45 0.45
286 288 290 292 294 296 300 310 320 330
Integration Limit x 1 (eV) Integration Limit x 2 (eV)

Fig. 3.9 Uncertainty in the computed fraction of sp2 -bonded carbon due to uncertainties in the
X-ray incidence angle used for the acquisition of the HOPG reference spectrum as a function of:
a the nominal fraction of sp2 -bonded carbon; and b the uncertainty in the X-ray incidence angle. c
Influence of the lower-photon-energy integration limit (x 1 ) used to calculate the area of the C 1s→σ *
transition on the computed fraction of sp2 -bonded carbon for a hydrogenated amorphous carbon
film (the high-photon-energy limit x 2 was kept fixed at 320 eV). Error bars represent the standard
deviation calculated from multiple independent measurements; d influence of the high-photon-
energy integration limit (x 2 ) used to calculate the area of the C 1s→σ * transition on the computed
fraction of sp2 -bonded carbon for a hydrogenated amorphous carbon film (the low-photon-energy
limit x 1 was kept fixed at 288.6 eV). Error bars represent the standard deviation calculated from
multiple independent measurements. For highlighting the influence of the integration limits on the
results of the quantification, the error in the calculated fraction of sp2 -bonded carbon relative to the
fraction obtained using the integration limits employed in [99] (x 1  288.6 eV; x 2  320 eV) is
shown in the upper part of the graphs. Adapted from [137]

carbon–carbon bonds are directly converted to sp2 -hybridized carbon-carbon bonds

upon annealing. Additionally, the broadening of the C 1s→π * absorption feature
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 85

clearly indicated an increase in bond length distribution with temperature, while the
shift of this peak to higher photon energies suggested an increase in the degree of
ordering of the sp2 carbon atoms. The quantification of the fraction of carbon atoms
in sp2 -hybridization state as a function of temperature allowed Grierson et al. to apply
Sullivan et al.’s model [171] to determine an activation energy range for the sp3 -to-sp2
conversion of carbon hybridization. Sullivan’s model describes the transformation
of sp3 - to sp2 -bonded carbon as a series of first order chemical reaction. Because of
bond length and angle disorder in a-C materials, a distribution of activation energies
for sp3 -to-sp2 conversion of carbon hybridization is also included in the model. The
activation energy range calculated by Grierson et al. (3.5 ± 0.9 eV) was in agreement
with the activation energy for bulk sp3 -to-sp2 conversion of carbon hybridization in
ta-C derived from Raman spectroscopic measurements (3.3 eV) [172].
More recently, the authors of the present contribution investigated in situ the
thermally-induced structural evolution of a class of doped hydrogenated amorphous
carbon materials, namely silicon- and oxygen-containing hydrogenated amorphous
carbon (a-C:H:Si:O—also referred to as silicon oxide-doped diamond-like carbon
(SiOx -DLC) or diamond-like nanocomposite (DLN)) [141, 142, 175]. a-C:H:Si:O
is a promising class of multicomponent materials for several applications since the
incorporation of silicon and oxygen reduces their residual stress [176, 177], while not
significantly affecting the mechanical properties [177]. In addition, a-C:H:Si:O films
may present other interesting properties, such as enhanced thermal stability [178,
179] and good tribological behavior (i.e., low friction and wear) across a broader
range of conditions and environments compared to hydrogenated amorphous carbon
(a-C:H) films [177, 180, 181]. The superior thermal stability of a-C:H:Si:O com-
pared to a-C:H was suggested to arise from the fourfold coordination of silicon,
which stabilizes the carbon atoms in the sp3 hybridization state (thus inhibiting their
conversion into sp2 -bonded carbon at elevated temperatures) [182–185]. Yet, a fun-
damental understanding of the origin of the superior thermal and thermo-oxidative
stability of a-C:H:Si:O has not been achieved. This hampers the possibility to tailor
the deposition process with the aim of depositing coatings with improved properties.
The thermally-induced structural evolution of a-C:H:Si:O was investigated in situ
by NEXAFS and XPS spectroscopy under high vacuum conditions (pressure <1.0 ×
10−8 Torr). Carbon 1s NEXAFS spectra (corrected for the contribution of the contam-
ination layer following the method outlined in [99]) of a-C:H:Si:O before and after
annealing are displayed in Fig. 3.10a, b. The NEXAFS spectrum of as-deposited
a-C:H:Si:O exhibited an absorption feature at 285.0 ± 0.1 eV, which is due to the
C 1s→π * transition for disordered carbon-carbon bonds [5, 118]. A broad hump
between 288 and 310 eV, which is due to the C 1s→σ * transition for disordered car-
bon–carbon σ bonds [5, 118], characterized the NEXAFS spectrum of as-deposited a-
C:H:Si:O. The presence of a significant amount of carbon–hydrogen and carbon–sil-
icon bonds in a-C:H:Si:O also resulted in the detection of distinct absorption features
at 287.0 ± 0.1 eV (assigned to the C 1s→σ * transition for C–H bonds [5, 80, 99]) and
288.9 ± 0.1 eV (C 1s→σ * transition for C–Si bonds [136]). It has to be emphasized
that carbon–oxygen bonds are assumed to be only present in the contamination layer.
Under this assumption the composition calculated by XPS more closely agrees with
86 F. Mangolini and J. B. McClimon

(a) (b) (c)

C1s→π* C1s→σ* C=C C-H C-Si
Partial Electron Yield (A.u.)

Partial Electron Yield (A.u.)

2.5 2.5 0.8 Annealing at:

Fraction sp2 Carbon

2.0 2.0 a-C:H:Si:O
0.7
Annealing at: a-C:H
1.5 1.5
RT 0.6
1.0 Annealing at: 1.0 150°C
RT 250°C 0.5
0.5 150°C 350°C 0.5 350°C
0.0 250°C 450°C 0.0 450°C 0.4
270 280 290 300 310 320 330 280 285 290 295 300 100 200 300 400
Photon Energy (eV) Photon Energy (eV) Temperature (ºC)

Fig. 3.10 a Pre- and post-edge normalized carbon 1s NEXAFS spectra of a-C:H:Si:O acquired
before annealing, and after annealing at different temperatures in vacuum (pressure <1.0 × 10−8
Torr); b zoomed view of the absorption edge region of the carbon 1s spectra. Spectra displayed
without any offset to allow for comparisons; and c fraction of sp2 -hybridized carbon versus anneal-
ing temperature calculated from NEXAFS spectra following the approach outlined in [137]. For
comparison, the evolution of the fraction of sp2 -bonded carbon for a-C:H thin films annealed under
similar experimental conditions [134] is also displayed (note: the as-deposited a-C:H thin film has
a fraction of sp2 -bonded carbon comparable to the one of as-grown a-C:H:Si:O as well as a similar
hydrogen concentration to a-C:H:Si:O, i.e., 26 ± 3 at.% for a-C:H and 34 ± 3 at.% for a-C:H:Si:O)

the results of secondary ion mass spectrometry (SIMS) and Rutherford backscattering
spectrometry (RMS) (SIMS and RBS measurements performed by Evans Analyti-
cal Group, Sunnyvale, CA, USA). Thus, carbon–oxygen absorption features do not
contribute to the contamination-corrected NEXAFS spectrum of a-C:H:Si:O shown
in Fig. 3.10a, b, in contrast to as-acquired NEXAFS data (displayed in Fig. 3.6).
Upon annealing, the intensity of the peak at 285.0 ± 0.1 eV progressively
increased, suggesting an increase in the fraction of sp2 -bonded carbon in the near-
surface region. At the same time, the absorption feature at 288.9 ± 0.1 eV decreased
in intensity with the annealing temperature, thus indicating the breakage of car-
bon–silicon bonds. The absorption feature assigned at 287.0 ± 0.1 eV also changed
at elevated temperatures: its intensity first increased upon annealing at 150 °C, which
might be due to hydrogen diffusion to the near-surface region upon low temperature
annealing, and then progressively decreased, which suggests the scission of car-
bon–hydrogen bonds upon annealing. These spectral variations could be highlighted
by computing the difference in NEXAFS spectra between subsequent annealing
steps (not shown). The calculation of difference spectra also highlighted a progres-
sive increase in spectral intensity on the high photon energy side (at ~285.5 eV) of
the C 1s→π * transition, suggesting an increase in the degree of ordering of the sp2
carbon phase [5, 112, 118]. This finding was corroborated by XPS measurements
(not shown).
The fraction of sp2 -hybridized carbon, which was calculated following the
methodology reported in [137], is displayed as a function of annealing temperature
in Fig. 3.10c. For comparison, Fig. 3.10c also reports the evolution of the fraction
of sp2 -bonded carbon for an a-C:H film obtained from in situ XPS annealing experi-
ments performed under similar experimental conditions (the as-deposited a-C:H film
has a fraction of sp2 -bonded carbon comparable to the one of as-grown a-C:H:Si:O
as well as a similar hydrogen content to a-C:H:Si:O, i.e., 26 ± 3 at.% for a-C:H and
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 87

34 ± 3 at.% for a-C:H:Si:O) [134]. These results clearly indicate that doping a-C:H
with silicon and oxygen at the modest level of, respectively, 6 ± 1 at.% and 3 ± 1 at.%
slightly increases the thermal stability of a-C:H:Si:O compared to a-C:H in high vac-
uum [141, 142]. The analysis of the energetics of the thermally-induced structural
evolution of a-C:H:Si:O is reported in [141], where atomistic insights (from molec-
ular dynamics (MD) simulations) into the processes occurring in this material at
elevated temperatures are also presented.

3.4.3 NEXAFS Spectroscopic Investigation

of the Tribochemistry of Solid Lubricants

3.4.3.1 Photoemission Electron Microscopy (PEEM)

As outlined in Sect. 3.3.1.2.2, photoemission electron microscopy (PEEM) is a full

field imaging technique commonly combined with X-ray synchrotron radiation. The
high surface sensitive of this spectroscopic technique together with its excellent spa-
tial resolution (<50 nm [69], or even below 10 nm in aberration corrected instruments
[70]) [14, 71–75] has made it an increasingly attractive characterization technique
in several fields, including tribology [14, 20, 77–82], thin film magnetism [83–85],
catalysis [86], semiconductor [87], polymer science [88–90], geology [91–93], and
microbiology [94].
Anders et al. employed PEEM to study lubricated and unlubricated hard disks
after wear tests [78, 79]. The hard disk/slider interface involves mechanical and tri-
bochemical processes between the hard carbon-based overcoat on the disk, the lubri-
cant (a perfluoropolyether (ZDOL) fluid [173]), and the carbon coated or uncoated
slider surface. By exploiting the excellent lateral resolution of PEEM, Anders et al.
acquired NEXAFS spectra inside the wear track after the disks failed the tribological
tests. The spectroscopic results indicated the formation of various new carbon–oxy-
gen (mostly carboxylic) bonds in the wear tracks with a reduction of the amount
of fluorine and carbon. These findings provided clear evidence that the chemical
modifications inside the wear track were induced by the degradation of the lubricant,
which was not solely a mechano-chemical process, but it was also accompanied by
oxidation reactions.
Konicek et al. investigated by PEEM the surface-chemical phenomena governing
the environmental dependence of the tribological properties of UNCD [20]. This
material exhibits poor tribological performance in dry or vacuum conditions, which
is hypothesized to be due to the high strength of the carbon–carbon bonds that rapidly
form across the interface. This happens as a consequence of the removal of passivat-
ing surface species (like hydrogen) induced by the sliding process [186, 187]. The
reduction of friction and wear as molecular species such as hydrogen, oxygen or water
are present in the surrounding vapour is thought to be due to the dissociative adsorp-
tion of gaseous species, which saturate the dangling bonds formed on the surface by
88 F. Mangolini and J. B. McClimon

the applied stress [188–191]. These findings are consistent with molecular dynam-
ics (MD) simulations [192–194] and density functional theory (DFT) calculations
[195–197] as well as with the studies performed at the nanometer scale by atomic
force microscopy (AFM) [198]. However, others proposed that low friction and wear
are achieved in humid environments in the case of highly sp3 -bonded hydrogen-free
carbon by the rehybridization of carbon atoms to ordered (graphitic) sp2 bonding
[199–201]. The debate regarding the origin of the impressively low friction and wear
of highly sp3 -bonded hydrogen-free carbon-based films in humid environments may
have been settled thanks to the work of Konicek et al., who, using PEEM combined
with NEXAFS spectroscopy, provided a spectroscopic evidence of the passivation
hypothesis [20, 82], as outlined in the following.
Konicek et al. performed tribological experiments using UNCD-coated silicon
nitride (Si3 N4 ) spheres and silicon wafers at low (0.1 N, 300 MPa mean Hertzian
contact pressure) and high (1.0 N, 649 MPa) loads, in low (1%) and high (50%)
relative humidity (RH). The tests were referred to as HD, LD, HW, and LW for high
(H) or low (L) load and dry (D) or humid (W for wet) conditions. In all cases, friction
coefficients below 0.05 were achieved at steady state, even though in the case of the
experiments performed in dry environment at high load (HD), more sliding cycles
were required to achieve this low friction coefficient.
Figure 3.11a displays an example of PEEM image captured at a photon energy of
289 eV (corresponding to the edge jump in the NEXAFS spectrum of UNCD, see
Fig. 6) [20]. Carbon 1s NEXAFS spectra could be extracted from regions of interest
(ROI) and compared to the spectrum of unworn UNCD (Fig. 3.11b). The authors
observed increases in spectral intensity from 286 to 289 eV, an energy window in
which the characteristic signals of C=O (~286.4 eV [5, 139]) and C–H (~287.5 eV
[5]) are found. These changes in the intensity of NEXAFS spectra, which strongly
depended on the environment and applied load (Fig. 3.11b), were accompanied by
some conversion of sp3 -bonded carbon to sp2 -bonding, as evidenced by the increased
intensity at 285 eV (due to the C 1s→π * transition for disordered carbon–carbon
bonds [5, 118]. In the case of as-deposited UNCD, this absorption peak is due to
sp2 -bonded carbon present at grain boundaries, as surface contamination, and due
to the reconstruction of diamond surfaces). However, upon sliding none of the C
1s→π * absorption features exhibited the shift from 285 to 285.5 eV (characteristic
position for the C 1s→π * absorption peak for HOPG), thus indicating that graphitic
(ordered) carbon is not formed upon sliding. Based on these findings, the authors
concluded that the lubrication mechanism for diamond involving the formation of
substantial graphitic carbon layers does not occur for UNCD under the broad range
of conditions explored. Rather, the mechanism leading to low friction is passivation
of dangling bonds.
This conclusion was substantiated by a subsequent study by the same authors
[82], in which tribological experiments were performed using UNCD- and hydrogen-
free tetrahedral amorphous carbon (ta-C)-coated silicon nitride spheres and silicon
wafers. The experiments were carried out under different environmental conditions
(relative humidity, RH, from 1 to 50%) and applied normal pressure (initial mean
Hertzian contact pressure ranging from 240 to 649 MPa). For the ta-C, tracks created
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 89

Fig. 3.11 a Example of PEEM image (acquired at 289.0 eV photon energies) of the wear track
generated upon sliding a UNCD-coated silicon nitride (Si3 N4 ) pin on UNCD with an applied load
of 1.0 N and a relative humidity of 1.0% (HD); b carbon 1s NEXAFS spectra extracted from PEEM
images. Top spectrum (black line): reference taken far from the wear track; bottom spectrum (gray
line): spectrum extracted from the region drawn in (a). Inset: magnified plot (C1 s→π * region) of
the spectra shown in (b) together with a spectrum extracted from the track generated at 0.1 N and
50% RH (LW. PEEM image not shown). From [20]

at lower humidity and higher load experienced increased friction and wear, while for
the UNCD friction trends were unclear but lower humidity did lead to greater wear
rates.
Ex situ PEEM measurements were carried out to gain information about the
chemical changes and structural transformations induced by the sliding process.
Figure 3.12a displays a typical PEEM image (with a defined region of interest, ROI)
obtained at 289.0 eV for the case of the track formed on ta-C at 0.5 N and 1.0%
RH. This track was heavily worn. The extracted carbon 1s and oxygen 1s spectra
from the selected ROI are also displayed in Fig. 3.12a. Reference spectra taken from
the unworn region are reported together with the spectra obtained from the PEEM
analyses of the track created upon sliding at 0.5 N and 50% RH (this track was only
lightly worn). Two main differences could be noticed when comparing the carbon 1s
NEXAFS spectra of the heavily-worn (0.5 N, 1.0% RH) and unworn regions: first,
the C 1s→π * transition for disordered carbon–carbon bonds at 285.0 eV [5, 118]
90 F. Mangolini and J. B. McClimon

(a) (b)
C 1s 0.5 N, 1.0% RH
C 1s 1.0 N, 1.0% RH

Electron Intensity (A.u.)

Electron Intensity (A.u.)
0.5 N, 50.0% RH 1.0 N, 50.0% RH
C-O Unworn C1s→σ* Unworn

C=O
C1s →π*

C1s →π* C1s →π*

C1s →π*

281 283 285 287 289 281 283 285 287 289

280 290 300 310 320 280 290 300 310 320
Photon Energy (eV) Photon Energy (eV)
O 1s C-O 0.5 N, 1.0% RH
O 1s C-O 1.0 N, 1.0% RH

Electron Intensity (A.u.)

Electron Intensity (A.u.)

0.5 N, 50.0% RH 1.0 N, 50.0% RH

Unworn Unworn

20 μm 20 μm

O-H
O-H C=O
C=O

530 540 550 560 530 540 550 56 0

Photon Energy (eV) Photon Energy (eV)

Fig. 3.12 a Example of PEEM image (acquired at 289.0 eV photon energy) of the wear track
generated upon sliding a ta-C-coated silicon nitride (Si3 N4 ) pin on ta-C with an applied load of
0.5 N and a relative humidity of 1.0%. The carbon 1s and oxygen 1s spectra extracted from the
region of interest defined in PEEM images are shown. The spectra extracted from PEEM images
collected from a sample after tribological testing at a relative humidity of 50% are also displayed;
b example of PEEM image (acquired at 289.0 eV photon energies) of the wear track generated
upon sliding a UNCD-coated silicon nitride (Si3 N4 ) pin on UNCD with an applied load of 1.0 N
and a relative humidity of 1.0%. The carbon 1s and oxygen 1s spectra extracted from the region of
interest defined in PEEM images are shown. The spectra extracted from PEEM images collected
from a sample tribologically-stressed at a relative humidity of 50% are also displayed. From [82]

increased in intensity in the former, which indicates the conversion of carbon atoms
from sp3 - to sp2 -hybridization state upon sliding; secondly, a significant amount of
oxidation is present in the track, as suggested by the peaks at ~286.7 and 288.6 eV,
which was assigned to C–O Rydberg orbitals and C–O σ * orbitals, respectively [106].
As for the case of the track generated at 0.5 N and 50.0% RH, which was lightly
worn, these spectral differences occurred to a much lesser extent, indicating a small
amount of rehybridization and some traces of oxidation. The oxygen 1 s spectra
provided additional information. While the spectrum of the track formed at 0.5 N
and 50.0% RH was very similar to the one extracted from the unworn area, in the
case of the experiments performed at 0.5 N and 1.0% RH there was a substantially
higher oxygen concentration in the wear track. Additionally, the oxygen was more
σ bonded as opposed to π bonded, as demonstrated by the much larger edge jump at
538.0 eV and the C–O feature at ~541 eV. In the case of the tracks for at 0.5 N and
1.0% RH, the pre-edge peak shifted from 532.7 eV (characteristic to a π * transition
for double-bonded oxygen) to 533.4 eV, indicating the presence of O–H bonds [106].
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 91

The PEEM results obtained from the tribological tests carried out with UNCD
were similar to those from the experiments performed with ta-C. Figure 3.12b shows
a typical PEEM image (acquired at 289 eV) from a region entirely inside the UNCD
track created at 1.0 N and 1.0% RH. The carbon 1s and oxygen 1s spectra extracted
from the worn and unworn regions are reported, together with the spectra obtained
from the experiments ran at 1.0 N and 50.0% RH. For the experiments carried out at
1.0 N and 1.0% RH, the increase in intensity of the peak at 285.0 eV, which is assigned
to the C 1s→π * transition for disordered carbon-carbon bonds [5, 118], provided
evidence of sp3 -to-sp2 rehybridization. Additionally, the spectral features appearing
at ~286.4 and ~288.6 eV clearly indicated the oxidation of the near-surface region.
The oxygen 1s data acquired on UNCD are also similar to those collected on ta-C:
there is a weak oxygen signal in the unworn region, while this signal significantly
increased in intensity in the case of the tracks generated at 1.0 N and 1.0% RH.
While the spectrum extracted from the wear track generated at 1.0 N and 50.0% RH
has a similar intensity to the one of the unworn region, the pre-edge peak shifted
from 533.2 to 533.5 eV, indicating, as in the case of the ta-C experiments, more O–H
bonding.
These results supported the conclusion that the lubrication mechanism of ta-C
and UCND is related to the passivation of surface dangling bonds by dissociated
water vapor. Upon sliding, dangling bonds are generated, but the overall tribological
behavior is dominated by the competition between their passivation by the disso-
ciative adsorption of water and the formation of covalent bonds across the slid-
ing interface. Additionally, the experimental results provided clear evidence that no
ordered graphitic carbon is formed upon sliding, even though some rehybridization
of sp3 -bonded carbon to sp2 bonding occurred. All together, the outcomes of this
work indicated that the primary solid-lubrication mechanism for highly sp3 -bonded,
nearly hydrogen-free carbon materials (either amorphous or polycrystalline) is the
passivation of dangling bonds by OH and H from the dissociation of vapor-phase
water. This was supported by ab initio density functional theory (DFT) calculations
[195–197].

3.4.3.2 Magnetically-Guided Imaging NEXAFS Spectroscopy

As outlined in Sect. 3.3.1.2.2, magnetically-guided imaging NEXAFS spectroscopy

is a spatially-resolved techniques that allows for the acquisition of spectroscopic data
over a field of view of 18 × 13 mm2 with a lateral resolution of 50 μm.
Konicek et al. employed this surface-analytical method to perform chemical imag-
ing of both countersurfaces of a self-mated sphere-on-flat UNCD contact [18]. Two
wear tracks were created using a bidirectional linear reciprocating microtribometer
placed in a sealed environment (5% RH): the first track was generated using a new,
unconditioned sphere, whose surface was then conditioned during the first experi-
ment and used for the second test. Conditioning of the sphere led to faster run-in,
lower friction, and lower wear when producing the second wear track. By taking
advantage of the capabilities of the rapid parallel processing magnetic field electron
92 F. Mangolini and J. B. McClimon

yield optics detector of NIST/Dow endstation of beamline U7A at the National Syn-
chrotron Light Source (NSLS, Brookhaven National Laboratory, Upton, NY, USA),
spectromicroscopic images could be collected on both the spherical and flat surfaces.
The spectroscopic results acquired on the flat indicated that the carbon chemical state
undergoes less modification when friction and wear are low. Additionally, the NEX-
AFS spectra extracted from the imaging data acquired on the film deposited on the
sphere exhibited a C 1s→π * transition at a photon energy of 285.2 eV both before
and after sliding, thus suggesting the presence of both disordered and ordered car-
bon–carbon bonds [5, 118, 122, 139]. These results led the authors conclude that the
higher ordering of the sp2 -bonded carbon for the material present in the near-surface
region of the sphere compared to the material deposited on the flat was not responsi-
ble for low friction and short run-in. Rather, conditioning of the sphere, which could
likely remove asperities and passivate surface dangling bonds, was proposed to be
the reason leading to immediate run in when sliding on a new portion of the flat and
resulting in lower friction with less chemical modification of the substrate.
The authors of the present contribution recently employed magnetically-guided
imaging NEXAFS spectroscopy to investigate the tribological properties of silicon-
and oxygen-containing hydrogenated amorphous carbon (a-C:H:Si:O). a-C:H:Si:O
thin films are promising materials for several applications (e.g., microelectrome-
chanical systems, aerospace, overcoats for hard-disks), since the incorporation of
silicon and oxygen in the hydrogenated amorphous carbon matrix renders its tri-
bological behavior less dependent on the environmental conditions compared to
undoped carbon-based materials [177]. While the tribological properties of these
films have been studied quite extensively in different environments (from dry to
humid atmospheres) [177], only a few studies have focused on the mechanisms by
which their excellent tribological performance is achieved, particularly under low
pressure conditions.
Tribological tests were performed using a pin-on-flat configuration, where a steel
pin was slid on an a-C:H:Si:O surface. The results revealed a strong environmen-
tal dependence of the tribological performance of a-C:H:Si:O: under high vacuum
conditions, the friction coefficient quickly reached values above 1, but a signifi-
cant reduction in friction was observed when hydrogen or oxygen was leaked in the
experimental chamber at a pressure of at least 50 or 10 mbar, respectively [98]. Addi-
tionally, upon increasing the oxygen pressure in the chamber from 10 to 1000 mbar,
the coefficient of friction increased from 0.02 ± 0.01 to 0.06 ± 0.01, whereas upon
increasing the hydrogen pressure from 50 to 2000 mbar, the coefficient of friction
decreased from 0.08 ± 0.01 to 0.02 ± 0.01.
To investigate the structural transformations and chemical reactions occurring in
the near-surface region of a-C:H:Si:O upon sliding under different environmental
conditions, imaging NEXAFS measurements were performed. Figure 3.13a shows
a schematic of an a-C:H:Si:O sample on which multiple tribological experiments
were performed at different oxygen and hydrogen partial pressures. Using the same
experimental apparatus employed by Konicek et al. [18], a series of two-dimensional
NEXAFS images were acquired as the energy of the incident soft X-rays was scanned
across the carbon 1s edge. Figure 3.13b-e displays NEXAFS images collected at pho-
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 93

ton energies corresponding to the characteristic absorption features of the NEXAFS

spectrum of a-C:H:Si:O [98], i.e., 285.1 ± 0.2 eV, 286.7 ± 0.2 eV, 287.6 ± 0.2 eV,
and 288.8 ± 0.2 eV. These images were generated after pre- and post-edge normal-
izing the NEXAFS data to ensure that variations in spectral intensity only arise from
chemical changes and are independent of the total carbon content. Independently
of the sliding environment, the intensity of the signals at 285.1 ± 0.2 eV (assigned
to the C 1s→π * transition for disordered carbon–carbon sp2 bonds [5, 118]) and
287.6 ± 0.2 eV (assigned to the C1 s→σ * transition for C–H bonds [5, 80, 99])
increased inside the wear tracks, whereas the intensity of the spectral feature at
288.8 ± 0.2 eV (due to the C 1s→σ * transition for carbon–silicon and carbon–oxy-
gen bonds, as well as to the C 1s→π * transition for carboxyl groups [5, 106, 136])
decreased in the wear tracks. As for the signal at 286.7 ± 0.2 eV (mainly due to the C
1s→π * transition for C=O bonds and with a small contribution from the C–O Ryd-
berg orbitals [5, 106]), its intensity significantly increased in the tracks generated
upon sliding under oxygen environment.
NEXAFS spectra were extracted from the wear tracks generated at different hydro-
gen and oxygen partial pressure (Fig. 3.13f). Subtle changes in the intensity of the
characteristic absorption features of a-C:H:Si:O could be discerned by comparing
the spectra. To more easily investigate the structural transformations and chemical
changes occurring in the near-surface region of a-C:H:Si:O upon sliding at differ-
ent hydrogen and oxygen partial pressure, the difference between spectra obtained
from the worn and unworn a-C:H:Si:O regions was calculated (Fig. 3.13g). Com-
puting these difference spectra also allows the contribution of the carbonaceous
contamination layer present on the sample surface due to its exposure to air to be
eliminated. This is valid under the assumption that the adventitious contamination
layer is laterally homogeneous in composition, density, and thickness across both
worn and unworn region, i.e., is due to ambient exposure after the tribotesting. This
approach avoids the need for correcting the as-acquired spectra for the presence of an
adventitious carbonaceous contamination using a reference sample [99]. As for the
tests performed in hydrogen, the intensity of the C 1s→π * transition for disordered
carbon–carbon sp2 bonds at 285.1 ± 0.2 eV significantly changed upon sliding at
different gas pressure: it increased for low gas pressure (50 and 200 mbar), which
demonstrates a significant conversion from sp3 - to sp2 -bonded carbon, whereas it
decreased for the highest hydrogen pressure (2000 mbar). Furthermore, a decrease
in intensity of the signal at 288.8 ± 0.2 eV (assigned to the C 1s→σ * transition for
carbon–silicon and carbon–oxygen bonds, as well as to the C 1s→π * transition for
carboxyl groups [5, 106, 136]) was observed. Since the computation of difference
spectra allows the contribution of the carbonaceous contamination layer present on
the sample surface due to its exposure to air to be eliminated, the main contribution
to the signal at 288.8 ± 0.2 eV derives from carbon–silicon bonds (carbon–oxygen
bonds are assumed to be only present in the contamination layer. Under this assump-
tion the composition calculated by XPS more closely agrees with the results of
secondary ion mass spectrometry (SIMS) and Rutherford backscattering spectrom-
etry (RMS). SIMS and RBS measurements performed by Evans Analytical Group,
Sunnyvale, CA, USA). Finally, intense peaks were also observed in difference spec-
94 F. Mangolini and J. B. McClimon

(a) (b) PEY Intensity at 286.7±0.2 eV (c) PEY Intensity at 288.8±0.2 eV

r
C1s →π (C=O) and C-O Ryd. orb. C1s →σ (C-O, C-Si); C1s →π (COOH)

ba
ba

m
r

r
ba
m

ba

ba
r
ba
00
m
00

m
m
10
50
20

10
20

0
10
O
2
H
2
H
2

O
2
H
2

O
2
2 mm low high
(d) PEY Intensity at 285.1±0.2 eV (e) PEY Intensity at 287.6±0.2 eV Partial Electron
Yield (A.u.)
C1s →π 2
(sp carbon) C1s →σ (C-H)

(f) (g) C=O, C-O

sp2 C
C-H
C-O, C-Si, COOH
Partial Electron Yield (A.u.)

Partial Electron Yield (A.u.)

3.0 O2 1000 mbar
0.4
2.5 O2 100 mbar
0.3 O2 10 mbar
2.0 0.2
1.5 0.1
0.0
1.0 Non-contact
H2 2000 mbar O2 1000 mbar -0.1 H2 2000 mbar
0.5 H2 200 mbar O2 100 mbar H2 200 mbar
-0.2
H2 50 mbar O2 10 mbar H2 50 mbar
0.0 -0.3
280 290 300 310 320 330 280 285 290 295 300 305 310
Photon Energy (eV) Photon Energy (eV)

Fig. 3.13 a Schematic of an a-C:H:Si:O sample on which multiple tribological experiments were
performed at different oxygen and hydrogen partial pressure using a steel counterbody; b–e NEX-
AFS maps acquired at different photon energies, i.e., 285.1 ± 0.2 eV (assigned to the C 1s→π *
transition for disordered carbon-carbon sp2 bonds [5, 118]), 286.7 ± 0.2 eV (mainly due to the C
1s→π ∗ transition for C=O bonds and with a small contribution from the C–O Rydberg orbitals
[5, 106]), 287.6 ± 0.2 eV (assigned to the C 1s→σ * transition for C–H bonds [5, 80, 99]), and
288.8 ± 0.2 eV (due to the C 1s→σ * transition for carbon–silicon and carbon–oxygen bonds, as
well as to the C 1s→π * transition for carboxyl groups [5, 106, 136]) are also displayed; f carbon
1s NEXAFS spectra of a-C:H:Si:O extracted from the contact and non-contact regions of the NEX-
AFS images reported in (b–e); g difference NEXAFS spectra (contact minus non-contact spectra)
computed to highlight chemical differences between contact and non-contact regions

tra at 286.0 ± 0.2 eV and 287.8 ± 0.2 eV (respectively due to the C 1s→π * transition
for carbonyl groups [5, 106] and the C 1s→σ * transition for C–H bonds [5, 80, 99]).
The detection of a peak that can be assigned to carbon–oxygen bonds in the spectra
extracted from the tracks produced in hydrogen can be due to some residual oxygen
in the vacuum chamber. It is worth noticing that the relative intensity of these two
peaks decreased with the hydrogen pressure, indicating a progressive increase in the
concentration of carbon–hydrogen bonds in the near-surface region of a-C:H:Si:O
upon increasing the hydrogen partial pressure during tribological tests against steel.
As for the tribological tests performed in oxygen, the difference carbon 1s NEX-
AFS spectra revealed a positive peak at 285.1 ± 0.2 eV, thus indicating an increase
in the fraction of disordered carbon–carbon sp2 bonds. Additionally, the peak at
288.8 ± 0.2 eV (assigned to the C 1s→σ * transition for carbon–silicon and car-
bon–oxygen bonds, as well as to the C 1s→π * transition for carboxyl groups [5,
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 95

106, 136]) decreased in intensity with gas pressure, while a positive peak appeared
at 286.0 ± 0.2 eV (assigned to the C 1s→π * transition for carbonyl groups [5, 106]),
thus indicating an increase in the fraction of carbonyl groups in the near-surface
region upon sliding in oxygen. The detection of a positive peak at 287.8 ± 0.2 eV,
which can be assigned to carbon–hydrogen bonds [5, 80, 99], in the spectra extracted
from the tracks produced in oxygen can be due to some residual hydrogen in the vac-
uum chamber. The relative intensity of the peaks at 286.0 ± 0.2 eV and 287.8 ± 0.2 eV
did not significantly change with the oxygen partial pressure, but their absolute inten-
sity decreased.
To gain information about the silicon oxidation state and quantitatively evaluate
the composition of the near-surface region, in situ (post mortem) XPS analyses were
performed (not shown). When sliding in hydrogen, a slight increase in the fraction of
silicon atoms in low oxidation states was observed with gas pressure. On the contrary,
in the case of the experiments performed in oxygen, the silicon oxidation state first
decreased at the lowest gas pressure (10 mbar), and then increased upon increasing
the gas pressure. The composition of the near-surface region of the wear tracks also
changed with gas pressure: when sliding in hydrogen, no significant variations were
observed up to 2000 mbar. At this pressure, the oxygen concentration in the near-
surface region of the wear track slightly decreased. In the case of the experiments
performed in oxygen, a progressive increase in oxygen and silicon concentration
was observed in the wear track upon increasing the gas pressure during tribological
experiments.
On the basis of the NEXAFS and XPS results, the authors developed a model
for explaining the friction response of a-C:H:Si:O sliding against steel under differ-
ent environmental conditions. Independent of the gas environment, a stress-induced
conversion from sp3 - to sp2 -bonded (disordered) carbon–carbon bonds occurs in the
near-surface region. When sliding in hydrogen, the newly-generated, strained sp2
carbon layer reacts with hydrogen molecules to form a hydrogenated amorphous
carbon material. Upon increasing the hydrogen pressure, the fraction of carbon–hy-
drogen bonds increases. This is proposed to progressively lower the shear strength of
the material at the sliding interface, thus resulting in a decrease of friction with hydro-
gen pressure. When sliding in oxygen, the dissociative reaction of oxygen molecules
with strained sp2 carbon–carbon bonds leads to the formation of carbonyl groups, as
indicated by imaging NEXAFS measurements. Additionally, increasing the oxygen
pressure during tribological testing leads to an increase in oxygen concentration in
the near-surface region of a-C:H:Si:O together with an increase in the fraction of
silicon atoms in high oxidation states, which are proposed to increase friction with
oxygen gas pressure by progressively increasing the shear strength of the material
generated at the sliding interface.
96 F. Mangolini and J. B. McClimon

3.5 Conclusions

Fundamental research in tribology hinges on the development of new characterization

techniques to better analyze contacting, sliding interfaces. Synchrotron-based near
edge X-ray absorption fine structure (NEXAFS) is a powerful surface-analytical
tool thanks to its elemental specificity, surface sensitivity, and ability to provide
important information about local bonding configurations, such as hybridization,
chemical states, and bond orientations. Additionally, spatially-resolved NEXAFS
spectroscopic techniques, such as photoemission electron microscopy (PEEM) and
magnetically-guided imaging, enables chemical imaging of material surfaces with
high spatial resolution. In light of this, NEXAFS spectroscopy has effectively been
used for investigating tribological surfaces as well as for accessing spatial inhomo-
geneities in their surface chemistry and structure.
In the present contribution, the principles of NEXAFS spectroscopy were
described together with the methods for acquiring and processing NEXAFS. Exam-
ples of the application of NEXAFS spectroscopy for fundamental and applied
research in tribology were provided. In particular, case studies concerning the
thermally- and tribologically-induced structural evolution of amorphous carbon-
based materials were presented.

Acknowledgements This material is based upon work supported by the National Science Foun-
dation under Grant No. DMR-1107642 and by the Agence Nationale de la Recherche under grant
No. ANR-11-NS09-01 through the Materials World Network program. F.M. acknowledges support
from The University of Texas at Austin Startup Funding, the Marie Curie International Outgo-
ing Fellowship for Career Development within the 7th European Community Framework Program
under contract no. PIOF-GA-2012-328776 and the Marie Skłodowska-Curie Individual Fellow-
ship within the European Union’s Horizon 2020 Program under contract no. 706289. The authors
acknowledge support from the Advanced Storage Technology Consortium ASTC (grant 2011-012).
The authors would like to thank Dr. C. Jaye and Dr. D. A. Fischer for the kind assistance with the
NEXAFS measurements at the National Synchrotron Light Source. Use of the National Synchrotron
Light Source, Brookhaven National Laboratory, was supported by the US Department of Energy,
Office of Science, and Office of Basic Energy Sciences, under Contract No. DE-AC02-98CH10886.
The authors would like to acknowledge Prof. R. W. Carpick (University of Pennsylvania, Philadel-
phia, USA) for fruitful discussions, valuable suggestions, and guidance. Finally, the authors would
also like to thank Dr. K. D. Koshigan (Ecole Centrale de Lyon, Ecully-Cedex, France) and Dr. J.
Fontaine (Ecole Centrale de Lyon, Ecully-Cedex, France) for performing tribological experiments
on a-C:H:Si:O.

References

1. W.G. Sawyer, K.J. Wahl, Accessing inaccessible interfaces. In Situ Approaches Mater. Tribol.
MRS Bull. 33, 1145–1150 (2008)
2. W.G. Sawyer, N. Argibay, D.L. Burris, B.A. Krick, Mechanistic studies in friction and wear
of bulk materials. Annu. Rev. Mater. Res. 44(1), 395–427 (2014)
3. C. Donnet, in Problem-Solving Methods in Tribology with Surface-Specific Techniques, ed. by
J.C. Rivière, S. Myhra. Handbook of Surface and Interface Analysis: Methods and Problem-
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 97

Solving, 2nd edn. (CRC Press, Taylor & Francis Group: Boca Raton, FL, 2009), pp. 351–388
4. S. Mobilio, F. Boscherini, C. Meneghini (eds.), Synchrotron Radiation: Basics, Methods and
Applications. Springer (2015)
5. J. Stöhr, NEXAFS Spectroscopy. Springer (1992)
6. A. Balerna, S. Mobilio, in Introduction to Synchrotron Radiation, ed. by S. Mobilio,
F. Boscherini, C. Meneghini. Synchrotron Radiation: Basics, Methods and Applications
(Springer, Berlin Heidelberg, 2015), pp. 3–28
7. B.K. Agarwal, X-ray Spectroscopy. Springer (1991)
8. D.C. Koningsberger, R. Prins, X-ray Absorption: Principles, Applications, Techniques of
EXAFS, SEXAFS, and XANES (Wiley, New York, 1988)
9. P. Fornasini, in Introduction to X-Ray Absorption Spectroscopy, ed. by S. Mobilio, F.
Boscherini, C. Meneghini. Synchrotron Radiation: Basics, Methods and Applications
(Springer, Berlin, Heidelberg, 2015), pp. 181–211
10. G. Bunker, Introduction to XAFS: A Practical Guide to X-ray Absorption Fine Structure
Spectroscopy (Cambridge University Press, Cambridge, UK; New York, 2010)
11. P.A. Lee, G. Beni, New method for the calculation of atomic phase shifts: application to
extended X-ray absorption fine structure (EXAFS) in molecules and crystals. Phys. Rev. B
15(6), 2862–2883 (1977)
12. S. Gurman, Interpretation of EXAFS data. J. Synchrotron Radiat. 2(1), 56–63 (1995)
13. B.K. Teo, EXAFS: Basic Principles and Data Analysis. Springer (1986)
14. S. Anders, H.A. Padmore, R.M. Duarte, T. Renner, T. Stammler, A. Scholl et al., Photoemission
electron microscope for the study of magnetic materials. Rev. Sci. Instrum. 70(10), 3973–3981
(1999)
15. E. Bauer, M. Mundschau, W. Swiech, W. Telieps, Surface studies by low-energy electron
microscopy (LEEM) and conventional UV photoemission electron microscopy (PEEM).
Ultramicroscopy 31(1), 49–57 (1989)
16. W. Engel, M.E. Kordesch, H.H. Rotermund, S. Kubala, A. von Oertzen, A UHV-compatible
photoelectron emission microscope for applications in surface science. Ultramicroscopy
36(1–3), 148–153 (1991)
17. O. Renault, N. Barrett, A. Bailly, L.F. Zagonel, D. Mariolle, J.C. Cezar et al., Energy-filtered
XPEEM with NanoESCA using synchrotron and laboratory X-ray sources: principles and
first demonstrated results. Surf. Sci. 601(20), 4727–4732 (2007)
18. A. Konicek, C. Jaye, M. Hamilton, W. Sawyer, D. Fischer, R. Carpick, Near-edge X-ray
absorption fine structure imaging of spherical and flat counterfaces of ultrananocrystalline
diamond tribological contacts: a correlation of surface chemistry and friction. Tribol. Lett.
44(1), 99–106 (2011)
19. M. Nicholls, M.N. Najman, Z. Zhang, M. Kasrai, P.R. Norton, P.U.P.A. Gilbert, The contri-
bution of XANES spectroscopy to tribology. Can. J. Chem. 85(10), 816–830 (2007)
20. A.R. Konicek, D.S. Grierson, P.U.P.A. Gilbert, W.G. Sawyer, A.V. Sumant, R.W. Carpick,
Origin of ultralow friction and wear in ultrananocrystalline diamond. Phys. Rev. Lett. 100(23),
235502 (2008)
21. R. Lindsay, G. Thornton, Structure of atomic and molecular adsorbates on Low-Miller-Index
ZnO surfaces using X-ray absorption spectroscopy. Top. Catal. 18(1–2), 15–19 (2002)
22. M. Bauer, C. Gastl, X-Ray absorption in homogeneous catalysis research: the iron-catalyzed
Michael addition reaction by XAS, RIXS and multi-dimensional spectroscopy. Phys. Chem.
Chem. Phys. 12(21), 5575–5584 (2010)
23. D.E. Ramaker, D.C. Koningsberger, The atomic AXAFS and μ XANES techniques as
applied to heterogeneous catalysis and electrocatalysis. Phys. Chem. Chem. Phys. 12(21),
5514–5534 (2010)
24. J.B. MacNaughton, L.-A. Naslund, T. Anniyev, H. Ogasawara, A. Nilsson, Peroxide-like
intermediate observed at hydrogen rich condition on Pt(111) after interaction with oxygen.
Phys. Chem. Chem. Phys. 12(21), 5712–5716 (2010)
25. T. Anniyev, H. Ogasawara, M.P. Ljungberg, K.T. Wikfeldt, J.B. MacNaughton, L.-A. Naslund
et al., Complementarity between high-energy photoelectron and L-edge spectroscopy for
98 F. Mangolini and J. B. McClimon

probing the electronic structure of 5d transition metal catalysts. Phys. Chem. Chem. Phys.
12(21), 5694–5700 (2010)
26. J. Genzer, E.J. Kramer, D.A. Fischer, Accounting for Auger yield energy loss for improved
determination of molecular orientation using soft x-ray absorption spectroscopy. J. Appl.
Phys. 92(12), 7070–7079 (2002)
27. M. Gliboff, L. Sang, K.M. Knesting, M.C. Schalnat, A. Mudalige, E.L. Ratcliff et al., Orienta-
tion of phenylphosphonic acid self-assembled monolayers on a transparent conductive oxide:
a combined NEXAFS, PM-IRRAS, and DFT study. Langmuir 29(7), 2166–2174 (2013)
28. F. Cheng, L.J. Gamble, D.G. Castner, XPS, TOF-SIMS, NEXAFS, and SPR characterization
of nitrilotriacetic acid-terminated self-assembled monolayers for controllable immobilization
of proteins. Anal. Chem. 80(7), 2564–2573 (2008)
29. S. Turgman-Cohen, D.A. Fischer, P.K. Kilpatrick, J. Genzer, Asphaltene adsorption onto self-
assembled monolayers of alkyltrichlorosilanes of varying chain length. ACS Appl. Mater.
Interfaces 1(6), 1347–1357 (2009)
30. T. Hemraj-Benny, S. Banerjee, S. Sambasivan, M. Balasubramanian, D.A. Fischer, G. Eres
et al., Near-edge X-ray absorption fine structure spectroscopy as a tool for investigating
nanomaterials. Small 2(1), 26–35 (2006)
31. A.D. Winter, E. Larios, F.M. Alamgir, C. Jaye, D. Fischer, E.M. Campo, Near-edge X-
ray absorption fine structure studies of electrospun poly(dimethylsiloxane)/poly(methyl
methacrylate)/multiwall carbon nanotube composites. Langmuir 29(51), 15822–15830 (2013)
32. T. Breuer, G. Witte, Diffusion-controlled growth of molecular heterostructures: fabrication
of two-, one-, and zero-dimensional C60 nanostructures on pentacene substrates. ACS Appl.
Mater. Interfaces 5(19), 9740–9745 (2013)
33. H.-J. Lee, K.-S. Lee, J.-M. Cho, T.-S. Lee, I. Kim, D.S. Jeong et al., Novel aspect in grain
size control of nanocrystalline diamond film for thin film waveguide mode resonance sensor
application. ACS Appl. Mater. Interfaces 5(22), 11631–11640 (2013)
34. Y.S. Li, Y. Tang, Q. Yang, J. Maley, R. Sammynaiken, T. Regier et al., Ultrathin W–Al dual
interlayer approach to depositing smooth and adherent nanocrystalline diamond films on
stainless steel. ACS Appl. Mater. Interfaces 2(2), 335–338 (2010)
35. K.J. Sankaran, Y.-F. Lin, W.-B. Jian, H.-C. Chen, K. Panda, B. Sundaravel et al., Structural
and electrical properties of conducting diamond nanowires. ACS Appl. Mater. Interfaces 5(4),
1294–1301 (2013)
36. A. Saravanan, B.-R. Huang, K.J. Sankaran, S. Kunuku, C.-L. Dong, K.-C. Leou et al., Bias-
enhanced nucleation and growth processes for ultrananocrystalline diamond films in Ar/CH4
plasma and their enhanced plasma illumination properties. ACS Appl. Mater. Interfaces 6(13),
10566–10575 (2014)
37. W.S. Yeap, X. Liu, D. Bevk, A. Pasquarelli, L. Lutsen, M. Fahlman et al., Functionalization of
boron-doped nanocrystalline diamond with N3 dye molecules. ACS Appl. Mater. Interfaces
6(13), 10322–10329 (2014)
38. S. Zhong, J.Q. Zhong, H.Y. Mao, R. Wang, Y. Wang, D.C. Qi et al., CVD graphene as
interfacial layer to engineer the organic donor-acceptor heterojunction interface properties.
ACS Appl. Mater. Interfaces 4(6), 3134–3140 (2012)
39. J. Kikuma, B.P. Tonner, XANES spectra of a variety of widely used organic polymers at the
C K-edge. J. Electron Spectrosc. Relat. Phenom. 82(1–2), 53–60 (1996)
40. B. Watts, S. Swaraj, D. Nordlund, J. Luning, H. Ade, Calibrated NEXAFS spectra of common
conjugated polymers. J. Chem. Phys. 134(2), 024702 (2011)
41. H. Ade, A.P. Hitchcock, NEXAFS microscopy and resonant scattering: composition and
orientation probed in real and reciprocal space. Polymer 49(3), 643–675 (2008)
42. D. Park, J.A. Finlay, R.J. Ward, C.J. Weinman, S. Krishnan, M. Paik et al., Antimicrobial
behavior of semifluorinated-quaternized triblock copolymers against airborne and marine
microorganisms. ACS Appl. Mater. Interfaces 2(3), 703–711 (2010)
43. H.S. Sundaram, Y. Cho, M.D. Dimitriou, J.A. Finlay, G. Cone, S. Williams et al., Fluori-
nated amphiphilic polymers and their blends for fouling-release applications: the benefits of
a triblock copolymer surface. ACS Appl. Mater. Interfaces 3(9), 3366–3374 (2011)
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 99

44. A.F. Tillack, K.M. Noone, B.A. MacLeod, D. Nordlund, K.P. Nagle, J.A. Bradley et al.,
Surface characterization of polythiophene: fullerene blends on different electrodes using near
edge X-ray absorption fine structure. ACS Appl. Mater. Interfaces 3(3), 726–732 (2011)
45. S.P. Cramer, T.K. Eccles, F.W. Kutzler, K.O. Hodgson, L.E. Mortenson, Molybdenum x-ray
absorption edge spectra. The chemical state of molybdenum in nitrogenase. J. Am. Chem.
Soc. 98(5), 1287–1288 (1976)
46. G. Meitzner, G.H. Via, F.W. Lytle, J.H. Sinfelt, Analysis of x-ray absorption edge data on
metal catalysts. J. Phys. Chem. 96(12), 4960–4964 (1992)
47. D.H. Pearson, C.C. Ahn, B. Fultz, White lines and d-electron occupancies for the 3d and 4d
transition metals. Phys. Rev. B 47(14), 8471–8478 (1993)
48. D. Hübner, F. Holch, M.L.M. Rocco, K.C. Prince, S. Stranges, A. Schöll et al., Isotope effects
in high-resolution NEXAFS spectra of naphthalene. Chem. Phys. Lett. 415(1–3), 188–192
(2005)
49. A.P. Hitchcock, C.E. Brion, K-shell excitation of HF and F2 studied by electron energy-loss
spectroscopy. J. Phys. B: At. Mol. Phys. 14(22), 4399–4413 (1981)
50. F. Sette, J. Stöhr, A.P. Hitchcock, Determination of intramolecular bond lengths in gas phase
molecules from K shell shape resonances. J. Chem. Phys. 81(11), 4906–4914 (1984)
51. J.S. Stevens, A. Gainar, E. Suljoti, J. Xiao, R. Golnak, E.F. Aziz et al., Chemical speciation
and bond lengths of organic solutes by core-level spectroscopy: ph and solvent influence on
p-aminobenzoic acid. Chemistry 21(19), 7256–7263 (2015)
52. J. Stöhr, F. Sette, A.L. Johnson, Near-edge X-ray-absorption fine-structure studies of
chemisorbed hydrocarbons: bond lengths with a ruler. Phys. Rev. Lett. 53(17), 1684–1687
(1984)
53. A. Gainar, J.S. Stevens, C. Jaye, D.A. Fischer, S.L. Schroeder, NEXAFS sensitivity to bond
lengths in complex molecular materials: a study of crystalline saccharides. J. Phys. Chem. B
119(45), 14373–14381 (2015)
54. V.L. Shneerson, D.K. Saldin, W.T. Tysoe, On the dependence with bond lengths of the observed
energies of NEXAFS resonances of diatomic molecules. Surf. Sci. 375(2–3), 340–352 (1997)
55. N. Haack, G. Ceballos, H. Wende, K. Baberschke, D. Arvanitis, A.L. Ankudinov et al., Shape
resonances of oriented molecules: ab initio theory and experiment on hydrocarbon molecules.
Phys. Rev. Lett. 84(4), 614–617 (2000)
56. D. Arvanitis, N. Haack, G. Ceballos, H. Wende, K. Baberschke, A.L. Ankudinov et al., Shape
resonances of oriented molecules. J. Electron Spectrosc. Relat. Phenom. 113(1), 57–65 (2000)
57. B. Kempgens, H.M. Köppe, A. Kivimäki, M. Neeb, K. Maier, U. Hergenhahn et al., On the
correct identification of shape resonances in NEXAFS. Surf. Sci. 425(1), L376–L380 (1999)
58. M.N. Piancastelli, D.W. Lindle, T.A. Ferrett, D.A. Shirley, Reply to the “Comment on ‘The
relationship between shape resonances and bond lengths”’. J. Chem. Phys. 87(5), 3255 (1987)
59. M.N. Piancastelli, D.W. Lindle, T.A. Ferrett, D.A. Shirley, The relationship between shape
resonances and bond lengths. J. Chem. Phys. 86(5), 2765–2771 (1987)
60. K. Weiss, P.S. Bagus, C. Wöll, Rydberg transitions in X-ray absorption spectroscopy of alka-
nes: the importance of matrix effects. J. Chem. Phys. 111(15), 6834–6845 (1999)
61. S.G. Urquhart, R. Gillies, Rydberg-valence mixing in the carbon 1 s near-edge X-ray absorp-
tion fine structure spectra of gaseous alkanes. J. Phys. Chem. A 109(10), 2151–2159 (2005)
62. D. Briggs, J.T. Grant (eds.), Surface Analysis by Auger and X-Ray Photoelectron Spectroscopy
(IM Publications, Chichester (UK), 2003)
63. D. Briggs, M.P. Seah (eds.), Practical Surface Analysis (Wiley, New York, 1990)
64. T. Maruyama, Y. Ishiguro, S. Nartitsuka, W. Norimatsu, M. Kusunoki, K. Amemiya et al.,
Near-edge X-ray absorption fine structure study of vertically aligned carbon nanotubes grown
by the surface decomposition of SiC. Jpn. J. Appl. Phys. 51(Copyright (c) 2012 The Japan
Society of Applied Physics), 055102
65. T. Maruyama, S. Sakakibara, S. Naritsuka, K. Amemiya, Initial stage of carbon nanotube
formation process by surface decomposition of SiC: STM and NEXAFS study. Diam. Relat.
Mater. 20(10), 1325–1328 (2011)
100 F. Mangolini and J. B. McClimon

66. G. Margaritondo, in Characteristics and Properties of Synchrotron Radiation, ed. by S.

Mobilio, F. Boscherini, C. Meneghini. Synchrotron Radiation: Basics, Methods and Applica-
tions (Springer, Berlin, Heidelberg, 2015), pp. 29–63
67. G. Aquilanti, L. Vaccari, J.R. Plaisier, A. Goldoni, in Instrumentation at Synchrotron Radiation
Beamlines, S. Mobilio, F. Boscherini, C. Meneghini. Synchrotron Radiation: Basics, Methods
and Applications (Springer, Berlin, Heidelberg, 2015), pp. 65–104
68. D. Attwood, Soft X-Rays and Extreme Ultraviolet Radiation: Principles and Applications
(2007)
69. A. Scholl, H. Ohldag, F. Nolting, J. Stöhr, H.A. Padmore, X-ray photoemission electron
microscopy, a tool for the investigation of complex magnetic structures (invited). Rev. Sci.
Instrum. 73(3), 1362 (2002)
70. T. Schmidt, A. Sala, H. Marchetto, E. Umbach, H.J. Freund, First experimental proof for
aberration correction in XPEEM: resolution, transmission enhancement, and limitation by
space charge effects. Ultramicroscopy 126, 23–32 (2013)
71. E. Bauer, Surface Microscopy with Low Energy Electrons (Springer, New York, 2014)
72. B.H. Frazer, M. Girasole, L.M. Wiese, T. Franz, G. De Stasio, Spectromicroscope for the
Photoelectron imaging of nanostructures with X-rays (SPHINX): performance in biology,
medicine and geology. Ultramicroscopy 99(2–3), 87–94 (2004)
73. J. Feng, A. Scholl, in Photoemission Electron Microscopy (PEEM), ed. by P.W. Hawkes,
J.C.H. Spence. Science of Microscopy (Springer, New York, NY, 2007), pp. 657–695
74. G. De Stasio, M. Capozi, G.F. Lorusso, P.A. Baudat, T.C. Droubay, P. Perfetti et al.,
MEPHISTO: performance tests of a novel synchrotron imaging photoelectron spectromi-
croscope. Rev. Sci. Instrum. 69(5), 2062–2066 (1998)
75. G. De Stasio, L. Perfetti, B. Gilbert, O. Fauchoux, M. Capozi, P. Perfetti et al., MEPHISTO
spectromicroscope reaches 20 nm lateral resolution. Rev. Sci. Instrum. 70(3), 1740–1742
(1999)
76. B.P. Tonner, G.R. Harp, Photoelectron microscopy with synchrotron radiation. Rev. Sci.
Instrum. 59(6), 853–858 (1988)
77. A.V. Sumant, D.S. Grierson, J.E. Gerbi, J. Birrell, U.D. Lanke, O. Auciello et al., Toward the
ultimate tribological interface: surface chemistry and nanotribology of ultrananocrystalline
diamond. Adv. Mater. 17(8), 1039–1045 (2005)
78. S. Anders, T. Stammler, W. Fong, D.B. Bogy, C.S. Bhatia, J. Stöhr, Investigation of slider
surfaces after wear using photoemission electron microscopy. J. Vac. Sci. Technol. A: Vac.
Surf. Films 17(5), 2731–2736 (1999)
79. S. Anders, T. Stammler, W. Fong, C.-Y. Chen, D.B. Bogy, C.S. Bhatia et al., Study of tribo-
chemical processes on hard disks using photoemission electron microscopy. J. Tribol. 121(4),
961–967 (1999)
80. A.V. Sumant, P.U.P.A. Gilbert, D.S. Grierson, A.R. Konicek, M. Abrecht, J.E. Butler et al.,
Surface composition, bonding, and morphology in the nucleation and growth of ultra-thin,
high quality nanocrystalline diamond films. Diam. Relat. Mater. 16(4–7), 718–724 (2007)
81. D.S. Grierson, A.V. Sumant, A.R. Konicek, M. Abrecht, J. Birrell, O. Auciello et al., Tribo-
chemistry and material transfer for the ultrananocrystalline diamond-silicon nitride interface
revealed by x-ray photoelectron emission spectromicroscopy. J. Vac. Sci. Technol. B: Micro-
electron. Nanometer Struct. 25(5), 1700–1705 (2007)
82. A.R. Konicek, D.S. Grierson, A.V. Sumant, T.A. Friedmann, J.P. Sullivan, P.U.P.A. Gilbert
et al., Influence of surface passivation on the friction and wear behavior of ultrananocrystalline
diamond and tetrahedral amorphous carbon thin films. Phys. Rev. B 85(15), 155448 (2012)
83. C.M. Schneider, G. Schönhense, Investigating surface magnetism by means of photoexcitation
electron emission microscopy. Rep. Prog. Phys. 65(12), 1785 (2002)
84. M.R. Freeman, B.C. Choi, Advances in magnetic microscopy. Science 294(5546), 1484 (2001)
85. F. Nolting, A. Scholl, J. Stöhr, J.W. Seo, J. Fompeyrine, H. Siegwart et al., Direct observation of
the alignment of ferromagnetic spins by antiferromagnetic spins. Nature 405(6788), 767–769
(2000)
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 101

86. M. Kim, M. Bertram, M. Pollmann, Oertzen Av, A.S. Mikhailov, H.H. Rotermund et al.,
Controlling chemical turbulence by global delayed feedback: pattern formation in catalytic
CO oxidation on Pt(110). Science 292(5520), 1357 (2001)
87. S. Aggarwal, A.P. Monga, S.R. Perusse, R. Ramesh, V. Ballarotto, E.D. Williams et al.,
Spontaneous ordering of oxide nanostructures. Science 287(5461), 2235 (2000)
88. F.-J. Meyer zu Heringdorf, M.C. Reuter, R.M. Tromp, Growth dynamics of pentacene thin
films. Nature 412(6846), 517–520 (2001)
89. C. Morin, H. Ikeura-Sekiguchi, T. Tyliszczak, R. Cornelius, J.L. Brash, A.P. Hitchcock et al.,
X-ray spectromicroscopy of immiscible polymer blends: polystyrene–poly(methyl methacry-
late). J. Electron Spectrosc. Relat. Phenom. 121(1–3), 203–224 (2001)
90. H. Ade, D.A. Winesett, A.P. Smith, S. Anders, T. Stammler, C. Heske et al., Bulk and surface
characterization of a dewetting thin film polymer bilayer. Appl. Phys. Lett. 73(25), 3775–3777
(1998)
91. G. De Stasio, P. Casalbore, R. Pallini, B. Gilbert, F. Sanità, M.T. Ciotti et al., Gadolinium
in human glioblastoma cells for gadolinium neutron capture therapy. Can. Res. 61(10), 4272
(2001)
92. G. De Stasio, B.H. Frazer, B. Gilbert, K.L. Richter, J.W. Valley, Compensation of charging
in X-PEEM: a successful test on mineral inclusions in 4.4 Ga old zircon. Ultramicroscopy
98(1), 57–62 (2003)
93. M. Labrenz, G.K. Druschel, T. Thomsen-Ebert, B. Gilbert, S.A. Welch, K.M. Kemner et al.,
Formation of sphalerite (ZnS) deposits in natural biofilms of sulfate-reducing bacteria. Science
290(5497), 1744 (2000)
94. B. Gilbert, R. Andres, P. Perfetti, G. Margaritondo, G. Rempfer, G. De Stasio, Charging
phenomena in PEEM imaging and spectroscopy. Ultramicroscopy 83(1–2), 129–139 (2000)
95. A. Locatelli, E. Bauer, Recent advances in chemical and magnetic imaging of surfaces and
interfaces by XPEEM. J. Phys.: Condens. Matter 20(9), 093002 (2008)
96. C. Wiemann, M. Patt, I.P. Krug, N.B. Weber, M. Escher, M. Merkel et al., A new nanospec-
troscopy tool with synchrotron radiation: NanoESCA@Elettra. e-J. Surf. Sci. Nanotechnol.
9, 395–399 (2011)
97. J.E. Baio, C. Jaye, D.A. Fischer, T. Weidner, Multiplexed orientation and structure analysis
by imaging near-edge X-ray absorption fine structure (MOSAIX) for combinatorial surface
science. Anal. Chem. 85(9), 4307–4310 (2013)
98. K.D. Koshigan, F. Mangolini, J.B. McClimon, B. Vacher, S. Bec, R.W. Carpick et al., Under-
standing the hydrogen and oxygen gas pressure dependence of the tribological properties of
silicon oxide–doped hydrogenated amorphous carbon coatings. Carbon 93, 851–860 (2015)
99. F. Mangolini, J.B. McClimon, F. Rose, R.W. Carpick, Accounting for nanometer-thick adven-
titious carbon contamination in X-ray absorption spectra of carbon-based materials. Anal.
Chem. 86(24), 12258–12265 (2014)
100. A. Schöll, Y. Zou, T. Schmidt, R. Fink, E. Umbach, Energy calibration and intensity nor-
malization in high-resolution NEXAFS spectroscopy. J. Electron Spectrosc. Relat. Phenom.
129(1), 1–8 (2003)
101. B. Watts, H. Ade, A simple method for determining linear polarization and energy calibration
of focused soft X-ray beams. J. Electron Spectrosc. Relat. Phenom. 162(2), 49–55 (2008)
102. B. Watts, L. Thomsen, P.C. Dastoor, Methods in carbon K-edge NEXAFS: experiment and
analysis. J. Electron Spectrosc. Relat. Phenom. 151(2), 105–120 (2006)
103. M. Olla, G. Navarra, B. Elsener, A. Rossi, Nondestructive in-depth composition profile of
oxy-hydroxide nanolayers on iron surfaces from ARXPS measurement. Surf. Interface Anal.
38(5), 964–974 (2006)
104. M.A. Scorciapino, G. Navarra, B. Elsener, A. Rossi, Nondestructive surface depth profiles
from angle-resolved X-ray photoelectron spectroscopy data using the maximum entropy
method. I. A New Protocol. J. Phys. Chem. C 113(51), 21328–21337 (2009)
105. M. Seah, Ultrathin SiO2 on Si I. quantifying and removing carbonaceous contamination. J.
Vac. Sci. Technol. A 21(2), 34 (2003)
102 F. Mangolini and J. B. McClimon

106. I. Ishii, A.P. Hitchcook, The oscillator strengths for C 1s and O 1s excitation of some saturated
and unsaturated organic alcohols, acids and esters. J. Electron Spectrosc. Relat. Phenom. 46(1),
55–84 (1988)
107. J.F. Morar, F.J. Himpsel, G. Hollinger, G. Hughes, J.L. Jordan, Observation of a C-1s core
exciton in diamond. Phys. Rev. Lett. 54(17), 1960–1963 (1985)
108. S.C. Ray, R.M. Erasmus, H. Tsai, M. nbsp, C. Pao et al., Hydrogenation effects of ultra-
nanocrystalline diamond detected by X-ray absorption near edge structure and raman spec-
troscopy. Jpn. J. Appl. Phys. 51(Copyright (c) 2012 The Japan Society of Applied Physics),
095201
109. A.V. Sumant, D.S. Grierson, J.E. Gerbi, J.A. Carlisle, O. Auciello, R.W. Carpick, Surface
chemistry and bonding configuration of ultrananocrystalline diamond surfaces and their
effects on nanotribological properties. Phys. Rev. B 76(23), 235429 (2007)
110. J. Diaz, S. Anders, X. Zhou, E.J. Moler, S.A. Kellar, Z. Hussain, Combined near edge X-ray
absorption fine structure and X-ray photoemission spectroscopies for the study of amorphous
carbon thin films. J. Electron Spectrosc. Relat. Phenom. 101–103, 545–550 (1999)
111. R. Gago, I. Jiménez, J.M. Albella, A. Climent-Font, D. Cáceres, I. Vergara et al., Bonding and
hardness in nonhydrogenated carbon films with moderate sp3 content. J. Appl. Phys. 87(11),
8174–8180 (2000)
112. D.S. Grierson, A.V. Sumant, A.R. Konicek, T.A. Friedmann, J.P. Sullivan, R.W. Carpick,
Thermal stability and rehybridization of carbon bonding in tetrahedral amorphous carbon. J.
Appl. Phys. 107(3), 033523–033525 (2010)
113. C. Lenardi, P. Piseri, V. Briois, C.E. Bottani, A.L. Bassi, P. Milani, Near-edge x-ray absorption
fine structure and Raman characterization of amorphous and nanostructured carbon films. J.
Appl. Phys. 85(10), 7159–7167 (1999)
114. S.C. Ray, H.M. Tsai, J.W. Chiou, B. Bose, J.C. Jan, K. Krishna et al., X-ray absorption
spectroscopy (XAS) study of dip deposited a-C:H(OH) thin films. J. Phys.: Condens. Matter
16(32), 5713 (2004)
115. A. Saikubo, N. Yamada, K. Kanda, S. Matsui, T. Suzuki, K. Niihara et al., Comprehensive
classification of DLC films formed by various methods using NEXAFS measurement. Diam.
Relat. Mater. 17(7–10), 1743–1745 (2008)
116. D. Wesner, S. Krummacher, R. Carr, T.K. Sham, M. Strongin, W. Eberhardt et al., Synchrotron-
radiation studies of the transition of hydrogenated amorphous carbon to graphitic carbon. Phys.
Rev. B 28(4), 2152–2156 (1983)
117. J.G. Buijnsters, R. Gago, A. Redondo-Cubero, I. Jimenez, Hydrogen stability in hydrogenated
amorphous carbon films with polymer-like and diamond-like structure. J. Appl. Phys. 112(9),
093502–093507 (2012)
118. G. Comelli, J. Stöhr, C.J. Robinson, W. Jark, Structural studies of argon-sputtered amorphous
carbon films by means of extended x-ray-absorption fine structure. Phys. Rev. B 38(11),
7511–7519 (1988)
119. J. Díaz, S. Anders, X. Zhou, E.J. Moler, S.A. Kellar, Z. Hussain, Analysis of the π* and σ*
bands of the x-ray absorption spectrum of amorphous carbon. Phys. Rev. B 64(12), 125204
(2001)
120. J. Diaz, O.R. Monteiro, Z. Hussain, Structure of amorphous carbon from near-edge and
extended x-ray absorption spectroscopy. Phys. Rev. B 76(9), 094201 (2007)
121. J. Diaz, G. Paolicelli, S. Ferrer, F. Comin, Separation of the sp3 and sp2 components in the C
1s photoemission spectra of amorphous carbon films. Phys. Rev. B 54(11), 8064–8069 (1996)
122. Gago, I. Jiménez, J.M. Albella, Detecting with X-ray absorption spectroscopy the modifi-
cations of the bonding structure of graphitic carbon by amorphisation, hydrogenation and
nitrogenation. Surf. Sci. 482–485, Part 1: 530–536 (2001)
123. R. Gago, M. Vinnichenko, H.U. Jäger, A.Y. Belov, I. Jiménez, N. Huang et al., Evolution of
sp2 networks with substrate temperature in amorphous carbon films: experiment and theory.
Phys. Rev. B 72(1), 014120 (2005)
124. H.-S. Jung, H.-H. Park, I.R. Mendieta, D.A. Smith, Determination of bonding structure of Si,
Ge, and N incorporated amorphous carbon films by near-edge x-ray absorption fine structure
and ultraviolet Raman spectroscopy. J. Appl. Phys. 96(2), 1013–1018 (2004)
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 103

125. C. Lenardi, E. Barborini, V. Briois, L. Lucarelli, P. Piseri, P. Milani, NEXAFS characteriza-

tion of nanostructured carbon thin-films exposed to hydrogen. Diam. Relat. Mater. 10(3–7),
1195–1200 (2001)
126. C. Lenardi, M. Marino, E. Barborini, P. Piseri, P. Milani, Evaluation of hydrogen chemisorp-
tion in nanostructured carbon films by near edge X-ray absorption spectroscopy. Eur. Phys.
J. B – Condens. Matter Complex Syst. 46(3), 441–447 (2005)
127. O.R. Monteiro, M.-P. Delplancke-Ogletree, Investigation of non-hydrogenated DLC: Si pre-
pared by cathodic arc. Surf. Coat. Technol. 163–164, 144–148 (2003)
128. V. Palshin, R. Tittsworth, C. Fountzoulas, E. Meletis, X-ray absorption spectroscopy, simu-
lation and modeling of Si-DLC films. J. Mater. Sci. 37(8), 1535–1539 (2002)
129. B.J. Schultz, C.J. Patridge, V. Lee, C. Jaye, P.S. Lysaght, C. Smith et al., Imaging local
electronic corrugations and doped regions in graphene. Nat. Commun. 2, 372 (2011)
130. D. Pacilé, M. Papagno, A.F. Rodríguez, M. Grioni, L. Papagno, Ç.Ö. Girit et al., Near-edge
X-ray absorption fine-structure investigation of graphene. Phys. Rev. Lett. 101(6), 066806
(2008)
131. S. Banerjee, T. Hemraj-Benny, M. Balasubramanian, D.A. Fischer, J.A. Misewich, S.S. Wong,
Ozonized single-walled carbon nanotubes investigated using NEXAFS spectroscopy. Chem.
Commun. 7, 772–773 (2004)
132. S. Banerjee, T. Hemraj-Benny, S. Sambasivan, D.A. Fischer, J.A. Misewich, S.S. Wong,
Near-edge X-ray absorption fine structure investigations of order in carbon nanotube-based
systems†. J. Phys. Chem. B 109(17), 8489–8495 (2005)
133. D.A. Fischer, K. Efimenko, R.R. Bhat, S. Sambasivan, J. Genzer, Mapping surface chem-
istry and molecular orientation with combinatorial near-edge X-ray absorption fine structure
spectroscopy. Macromol. Rapid Commun. 25(1), 141–149 (2004)
134. F. Mangolini, F. Rose, J. Hilbert, R.W. Carpick, Thermally induced evolution of hydrogenated
amorphous carbon. Appl. Phys. Lett. 103(16), 161605 (2013)
135. P.E. Batson, Carbon 1s near-edge-absorption fine structure in graphite. Phys. Rev. B 48(4),
2608–2610 (1993)
136. A. Wada, T. Ogaki, M. Niibe, M. Tagawa, H. Saitoh, K. Kanda et al., Local structural analysis
of a-SiCx: H films formed by decomposition of tetramethylsilane in microwave discharge
flow of Ar. Diam. Relat. Mater. 20(3), 364–367 (2011)
137. F. Mangolini, J.B. McClimon, R.W. Carpick, Quantitative evaluation of the carbon hybridiza-
tion state by near edge X-ray absorption fine structure spectroscopy. Anal. Chem. 88(5),
2817–2824 (2016)
138. J. Robertson, Diamond-like amorphous carbon. Mater. Sci. Eng.: R: Rep. 37(4–6), 129–281
(2002)
139. S. Osswald, G. Yushin, V. Mochalin, S.O. Kucheyev, Y. Gogotsi, Control of sp2 /sp3 carbon
ratio and surface chemistry of nanodiamond powders by selective oxidation in air. J. Am.
Chem. Soc. 128(35), 11635–11642 (2006)
140. F.L. Coffman, R. Cao, P.A. Pianetta, S. Kapoor, M. Kelly, L.J. Terminello, Near-edge X-ray
absorption of carbon materials for determining bond hybridization in mixed sp2 /sp3 bonded
materials. Appl. Phys. Lett. 69(4), 568–570 (1996)
141. F. Mangolini, J. Hilbert, J.B. McClimon, J.R. Lukes, R.W. Carpick, Thermally induced
structural evolution of silicon- and oxygen-containing hydrogenated amorphous carbon: a
combined spectroscopic and molecular dynamics simulation investigation. Langmuir 34(9),
2989–2995 (2018)
142. F. Mangolini, B.A. Krick, T.D.B. Jacobs, S.R. Khanal, F. Streller, J.B. McClimon et al., Effect
of silicon and oxygen dopants on the stability of hydrogenated amorphous carbon under harsh
environmental conditions. Carbon 130, 127–136 (2018)
143. A.C. Ferrari, J. Robertson, Raman spectroscopy of amorphous, nanostructured, diamond-like
carbon, and nanodiamond. Philos. Trans. Royal Soc. Lond. Ser. A: Math. Phys. Eng. Sci.
2004(362), 2477–2512 (1824)
144. F. Rose, N. Wang, R. Smith, Q.-F. Xiao, H. Inaba, T. Matsumura et al., Complete characteri-
zation by Raman spectroscopy of the structural properties of thin hydrogenated diamond-like
carbon films exposed to rapid thermal annealing. J. Appl. Phys. 116(12), 123516 (2014)
104 F. Mangolini and J. B. McClimon

145. A.C. Ferrari, B. Kleinsorge, G. Adamopoulos, J. Robertson, W.I. Milne, V. Stolojan et al.,
Determination of bonding in amorphous carbons by electron energy loss spectroscopy, Raman
scattering and X-ray reflectivity. J. Non-Cryst. Solids 266–269 Part 2, 765–768 (2000)
146. J. Filik, P.W. May, S.R.J. Pearce, R.K. Wild, K.R. Hallam, XPS and laser Raman analysis of
hydrogenated amorphous carbon films. Diam. Relat. Mater. 12(3–7), 974–978 (2003)
147. S. Kaciulis, Spectroscopy of carbon: from diamond to nitride films. Surf. Interface Anal.
44(8), 1155–1161 (2012)
148. A. Mezzi, S. Kaciulis, Surface investigation of carbon films: from diamond to graphite. Surf.
Interface Anal. 42(6–7), 1082–1084 (2010)
149. J.C. Lascovich, V. Rosato, Analysis of the electronic structure of hydrogenated amorphous
carbon via Auger spectroscopy. Appl. Surf. Sci. 152(1–2), 10–18 (1999)
150. S. Kaciulis, A. Mezzi, P. Calvani, D.M. Trucchi, Electron spectroscopy of the main allotropes
of carbon. Surf. Interface Anal. 46(10–11), 966–969 (2014)
151. B. Lesiak, J. Zemek, P. Jiricek, L. Stobinski, A. Jóźwik, The line shape analysis of electron
spectroscopy spectra by the artificial intelligence methods for identification of C sp2 /sp3
bonds. Phys. Status Solidi (b) 247(11–12), 2838–2842 (2010)
152. G. Speranza, N. Laidani, Measurement of the relative abundance of sp2 and sp3 hybridised
atoms in carbon based materials by XPS: a critical approach. Part I. Diam. Relat. Mater. 13(3),
445–450 (2004)
153. J. Zemek, J. Zalman, A. Luches, XAES and XPS study of amorphous carbon nitride layers.
Appl. Surf. Sci. 133(1–2), 27–32 (1998)
154. Y. Mizokawa, T. Miyasato, S. Nakamura, K.M. Geib, C.W. Wilmsen, Comparison of the
CKLL first-derivative auger spectra from XPS and AES using diamond, graphite SiC and
diamond-like-carbon films. Surf. Sci. 182(3), 431–438 (1987)
155. Y. Mizokawa, T. Miyasato, S. Nakamura, K.M. Geib, C.W. Wilmsen, The C KLL first-
derivative x-ray photoelectron spectroscopy spectra as a fingerprint of the carbon state and
the characterization of diamond like carbon films. J. Vac. Sci. Technol. A: Vac. Surf. Films
5(5), 2809–2813 (1987)
156. S.D. Berger, D.R. McKenzie, P.J. Martin, EELS analysis of vacuum arc-deposited diamond-
like films. Philos. Mag. Lett. 57(6), 285–290 (1988)
157. J. Kulik, G.D. Lempert, E. Grossman, D. Marton, J.W. Rabalais, Y. Lifshitz, sp3 content of
mass-selected ion-beam-deposited carbon films determined by inelastic and elastic electron
scattering. Phys. Rev. B 52(22), 15812–15822 (1995)
158. Y. Wang, H. Chen, R.W. Hoffman, J.C. Angus, Structural analysis of hydrogenated diamond-
like carbon films from electron energy loss spectroscopy. J. Mater. Res. 5(11), 2378–2386
(1990)
159. M.J. Paterson, An investigation of the role of hydrogen in ion beam deposited a-C:H. Diam.
Relat. Mater. 7(6), 908–915 (1998)
160. C. Donnet, J. Fontaine, F. Lefebvre, A. Grill, V. Patel, C. Jahnes, Solid state 13 C and 1 H
nuclear magnetic resonance investigations of hydrogenated amorphous carbon. J. Appl. Phys.
85(6), 3264–3270 (1999)
161. J. Peng, A. Sergiienko, F. Mangolini, P.E. Stallworth, S. Greenbaum, R.W. Carpick, Solid
state magnetic resonance investigation of the thermally-induced structural evolution of silicon
oxide-doped hydrogenated amorphous carbon. Carbon 105, 163–175 (2016)
162. G. Kovach, A. Karacs, G. Radnoczi, H. Csorbai, L. Guczi, M. Veres et al., Modified π-states
in ion-irradiated carbon. Appl. Surf. Sci. 254(9), 2790–2796 (2008)
163. J.A. Leiro, M.H. Heinonen, T. Laiho, I.G. Batirev, Core-level XPS spectra of fullerene, highly
oriented pyrolitic graphite, and glassy carbon. J. Electron Spectrosc. Relat. Phenom. 128(2–3),
205–213 (2003)
164. R.F. Egerton, An Introduction to EELS. Electron Energy-Loss Spectroscopy in the Electron
Microscope. Springer, US, pp. 1–28 (2011)
165. P.J. Fallon, V.S. Veerasamy, C.A. Davis, J. Robertson, G.A.J. Amaratunga, W.I. Milne et al.,
Properties of filtered-ion-beam-deposited diamondlike carbon as a function of ion energy.
Phys. Rev. B 48(7), 4777–4782 (1993)
3 Near Edge X-Ray Absorption Fine Structure Spectroscopy … 105

166. D.G. McCulloch, D.R. McKenzie, C.M. Goringe, Ab initio simulations of the structure of
amorphous carbon. Phys. Rev. B 61(3), 2349–2355 (2000)
167. K.E. Sohn, M.D. Dimitriou, J. Genzer, D.A. Fischer, C.J. Hawker, E.J. Kramer, Determination
of the electron escape depth for NEXAFS spectroscopy. Langmuir 25(11), 6341–6348 (2009)
168. S. Anders, J. Diaz, J.W. Ager Iii, R.Y. Lo, D.B. Bogy, Thermal stability of amorphous hard
carbon films produced by cathodic arc deposition. Appl. Phys. Lett. 71(23), 3367–3369 (1997)
169. S. Takabayashi, K. Okamoto, H. Sakaue, T. Takahagi, K. Shimada, T. Nakatani, Annealing
effect on the chemical structure of diamondlike carbon. J. Appl. Phys. 104(4), 043512–043516
(2008)
170. N. Wang, K. Komvopoulos, F. Rose, B. Marchon, Structural stability of hydrogenated amor-
phous carbon overcoats used in heat-assisted magnetic recording investigated by rapid thermal
annealing. J. Appl. Phys. 113(8), 083517–083517 (2013)
171. J.P. Sullivan, T. Friedmann, A. Baca, Stress relaxation and thermal evolution of film properties
in amorphous carbon. J. Electron. Mater. 26(9), 1021–1029 (1997)
172. A.C. Ferrari, S.E. Rodil, J. Robertson, W.I. Milne, Is stress necessary to stabilise sp3 bonding
in diamond-like carbon? Diam. Relat. Mater. 11(3–6), 994–999 (2002)
173. C.M. Mate, Tribology on the Small Scale—A Bottom Up Approach to Friction, Lubrication,
and Wear (Oxford University Press, Oxford, 2007)
174. M.H. Kryder, E.C. Gage, T.W. McDaniel, W.A. Challener, R.E. Rottmayer, J. Ganping et al.,
Heat assisted magnetic recording. Proc. IEEE 96(11), 1810–1835 (2008)
175. J. Hilbert, F. Mangolini, J.B. McClimon, J.R. Lukes, R.W. Carpick, Si doping enhances the
thermal stability of diamond-like carbon through reductions in carbon-carbon bond length
disorder. Carbon 131, 72–78 (2018)
176. C. Venkatraman, D. Kester, A. Goel, D. Bray, in Diamond-Like Nanocomposite Coatings—A
New Class of Materials, ed. by T.S. Sudarshan, W. Reitz, J.J. Stiglich. Surface Modification
Technologies IX (The Minerals, Metals & Materials Society, 1996)
177. T.W. Scharf, J.A. Ohlhausen, D.R. Tallant, S.V. Prasad, Mechanisms of friction in diamondlike
nanocomposite coatings. J. Appl. Phys. 101(6), 063521–063511 (2007)
178. W.J. Yang, Y.-H. Choa, T. Sekino, K.B. Shim, K. Niihara, K.H. Auh, Thermal stability eval-
uation of diamond-like nanocomposite coatings. Thin Solid Films 434(1–2), 49–54 (2003)
179. C. Jongwannasiri, X. Li, S. Watanabe, Improvement of thermal stability and tribological
performance of diamond-like carbon composite thin films. Mater. Sci. Appl. 4, 630–636
(2013)
180. V.F. Dorfman, Diamond-like nanocomposites (DLN). Thin Solid Films 212(1–2), 267–273
(1992)
181. D. Neerinck, P. Persoone, M. Sercu, A. Goel, D. Kester, D. Bray, Diamond-like nanocomposite
coatings (a-C:H/a-Si:O) for tribological applications. Diam. Relat. Mater. 7(2–5), 468–471
(1998)
182. F. Demichelis, C.F. Pirri, A. Tagliaferro, Influence of silicon on the physical properties of
diamond-like films. Mater. Sci. Eng. B 11(1–4), 313–316 (1992)
183. R. Hatada, S. Flege, K. Baba, W. Ensinger, H.J. Kleebe, I. Sethmann et al., Temperature
dependent properties of silicon containing diamondlike carbon films prepared by plasma
source ion implantation. J. Appl. Phys. 107(8), 083307–083306 (2010)
184. G.J. Wan, P. Yang, R.K.Y. Fu, Y.F. Mei, T. Qiu, S.C.H. Kwok et al., Characteristics and
surface energy of silicon-doped diamond-like carbon films fabricated by plasma immersion
ion implantation and deposition. Diam. Relat. Mater. 15(9), 1276–1281 (2006)
185. W.-J. Wu, M.-H. Hon, Thermal stability of diamond-like carbon films with added silicon.
Surf. Coat. Technol. 111(2–3), 134–140 (1999)
186. C. Donnet, A. Erdemir (eds.), Tribology of Diamond-Like Carbon Films (Springer, New York,
2008)
187. A. Erdemir, C. Donnet, Tribology of Diamond, Diamond-Like Carbon, and Related Films, in
Modern Tribology Handbook, ed. by B. Bhushan (CRC Press, Boca Raton, 2001)
188. J. Andersson, R.A. Erck, A. Erdemir, Frictional behavior of diamondlike carbon films in
vacuum and under varying water vapor pressure. Surf. Coat. Technol. 163–164, 535–540
(2003)
106 F. Mangolini and J. B. McClimon

189. Y. Tzeng, Very low friction for diamond sliding on diamond in water. Appl. Phys. Lett. 63(26),
3586–3588 (1993)
190. M.N. Gardos, Surface chemistry-controlled tribological behavior of silicon and diamond.
Tribol. Lett. 2(2), 173–187 (1996)
191. M.N. Gardos, Tribological fundamentals of polycrystalline diamond films. Surf. Coat. Tech-
nol. 113(3), 183–200 (1999)
192. J.A. Harrison, D.W. Brenner, Simulated tribochemistry: an atomic-scale view of the wear of
diamond. J. Am. Chem. Soc. 116(23), 10399–10402 (1994)
193. J.A. Harrison, C.T. White, R.J. Colton, D.W. Brenner, Investigation of the atomic-scale fric-
tion and energy dissipation in diamond using molecular dynamics. Thin Solid Films 260(2),
205–211 (1995)
194. M.D. Perry, J.A. Harrison, Universal aspects of the atomic-scale friction of diamond surfaces.
J. Phys. Chem. 99(24), 9960–9965 (1995)
195. G. Zilibotti, M.C. Righi, M. Ferrario, Ab initio study on the surface chemistry and nanotri-
bological properties of passivated diamond surfaces. Phys. Rev. B 79(7), 075420 (2009)
196. O. Manelli, S. Corni, M.C. Righi, Water adsorption on native and hydrogenated diamond
(001) surfaces. J. Phys. Chem. C 114(15), 7045–7053 (2010)
197. Y. Qi, E. Konca, A.T. Alpas, Atmospheric effects on the adhesion and friction between non-
hydrogenated diamond-like carbon (DLC) coating and aluminum—a first principles investi-
gation. Surf. Sci. 600(15), 2955–2965 (2006)
198. R.J.A. van den Oetelaar, C.F.J. Flipse, Atomic-scale friction on diamond (111) studied by
ultra-high vacuum atomic force microscopy. Surf. Sci. 384(1–3), L828–L835 (1997)
199. M.N. Gardos, B.L. Soriano, The effect of environment on the tribological properties of poly-
crystalline diamond films. J. Mater. Res. 5, 2599–2609 (1990)
200. A. Erdemir, G.R. Fenske, A.R. Krauss, D.M. Gruen, T. McCauley, R.T. Csencsits, Tribological
properties of nanocrystalline diamond films. Surf. Coat. Technol. 120–121, 565–572 (1999)
201. S.E. Grillo, J.E. Field, The friction of CVD diamond at high Hertzian stresses: the effect of
load, environment and sliding velocity. J. Phys. D Appl. Phys. 33(6), 595 (2000)
Chapter 4
Tribochemistry of n-Alkane Thiols
Examined by Gas-Phase Lubrication
(GPL)

Maria-Isabel De Barros Bouchet and Jean-Michel Martin

Abstract Gas Phase Lubrication (GPL) is an approach allowing to simulate exper-

imentally the reaction mechanism of lubrication additives on a surface thanks to a
specific device, an environmentally controlled analytical tribometer (ECAT). Indeed
it is shown that GPL is able to simulate the boundary lubrication regime by bringing
directly the molecules on the surface in contact. The presence of in situ chemical
analysis clearly gives some insights on the chemical dissociation pathway and reac-
tion of additive molecules on the surface under shearing leading to the tribofilm
formation. This is a huge advantage compared to standard liquid phase lubrication
coupled with ex situ post mortem surface analyses. GPL approach also permits to
investigate the critical role of the freshly nascent surfaces obtained by friction in UHV
in the decomposition mechanism of the additive molecules. This brings new knowl-
edge of the complex phenomena that can occur in lubricated friction experiments.
Moreover, such experimental approach appears particularly suited for combination
with numerical simulations, offering unique capabilities for investigating boundary
lubricated interfaces. In this chapter, GPL is applied to investigate the tribochemical
reactions of sulfur-based compounds, N-alkanethiols. Undoubtedly, the primordial
role of nascent tribo-stressed metallic surfaces obtained after shearing in the disso-
ciation of lubricant additives under boundary lubrication conditions is highlighted.
The investigation into the reactivity of thiol molecules showed that they suffer dis-
sociative adsorption on the asperities of the tribo-stressed nascent metallic surface,
rapidly forming a thin protective film composed of iron sulfide reducing the friction.

M.-I. De Barros Bouchet (B) · J.-M. Martin

Laboratoire de Tribologie et Dynamique des Systèmes, Ecole Centrale de Lyon, CNRS,
UMR5513, Université de Lyon I, Lyon 69134, Ecully, France
e-mail: maria-isabel.de-barros@ec-lyon.fr
J.-M. Martin
e-mail: jean-michel.martin@ec-lyon.fr

© Springer Nature Switzerland AG 2018 107

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_4
108 M.-I. De Barros Bouchet and J.-M. Martin

4.1 Introduction

Currently, the development of new efficient and sustainable lubrication technologies,

such as low SAPS (low sulfated ash, phosphorus and sulfur) lubricants or lubricants
containing biodegradable additives, is crucial to meet both environmental and eco-
nomic concerns of the 21st century [1]. As formulated lubricants are highly complex
fluids due to the numerous additives mixed with the base oils, the reaction mech-
anisms of each, as well as possible synergistic or antagonist interactions between
them, need to be well understood. Given the complexity of the reaction mechanisms
at the origin of the properties of the lubricating additives and the physico-chemical
phenomena that may impact these reactions, it is necessary for researchers to control
and simplify tribological systems. Most studies in the literature were conducted with
traditional tribometers to understand the action mode of the additives by controlling
as many parameters as possible and using surface analysis techniques. However,
these studies at atmospheric pressure still have some restrictions, such as the lack
of control of the surface chemistry and post-mortem chemical analyses. To avoid
these problems and since chemical analyses are crucial for understanding physico-
chemical phenomena, studies combining tribometry and in situ surface analyses
were carried out using ultra-high vacuum (UHV) chambers. Initially, such tribome-
ters were used to investigate the friction of clean materials without contaminants [2],
with the advantage that the surface of these materials can be modified in a controlled
way by introducing molecules in the gaseous phase [3]. The coupling of tribological
experiments with in situ analytical techniques (XPS, AES, SIMS, etc.) allowed the
formation of well-defined surfaces with controlled molecular top layers and the deter-
mination of their absorption and decomposition properties [4, 5]. These tribometers
have also been used to develop and test solid lubricants for aerospace applications [6,
7]. These new in situ approaches, aimed at further simplifying tribological systems
by controlling the atmosphere, surfaces and products formed, appear to be perfectly
suited for investigating the action mechanisms of lubricating additives. As in all
simplified systems, the connection between these systems and reality remains to be
established. In particular, it is necessary to study the link between liquid-phase lubri-
cation and gas-phase lubrication. Many experiments, including those of Lauer [8],
illustrate the possibility of lubricating a surface with gaseous molecules and suggest
that the two above means of lubrication are not so different [9]. This point is partic-
ularly justified under boundary lubrication conditions, where friction mainly results
from the additives adsorbed on the asperities of the sliding surfaces where there is no
longer a liquid lubricant film and thus no effect related to the rheology of the fluid.
Therefore, how the additives are brought to the sliding surfaces does not matter. The
crucial point for boundary lubrication to be effective is their presence on the surface!
Consequently, we examined a new approach based on gas-phase lubrication (GPL)
combined with in situ surface analyses to experimentally simulate reactions of the
additives under boundary lubrication conditions. In the following sections of this
chapter, the GPL approach is detailed, and its application in the study of the adsorption
and tribochemistry of thiol compounds on steel-based substrates is presented. A
4 Tribochemistry of n-Alkane Thiols Examined … 109

specific experimental methodology allowed by GPL to separate the different sources

of activation of the tribochemical reaction described in the literature (i.e., thermal
effects, pressure effects, nascent surfaces, molecular shearing, exo-electron emission,
etc.) is also presented.

4.2 Experimental

4.2.1 Gas-Phase Lubrication (GPL) Coupled with In Situ

XPS/AES Surface Analyses

Under boundary lubrication conditions, surfaces asperities bear all the load of the
contact. In these specific conditions, the contact can be considered ‘dry’ because there
is practically no base oil fluid film at the contact. The friction is concentrated at the
lubricant additives adsorbed on the surface of each body in contact. In GPL exper-
iments, low-molecular-weight molecules, which are easily vaporized, are needed.
First, polished surfaces are heated to approximately 100 °C under UHV to remove
physisorbed molecules. Second, gaseous molecules are introduced and chemically
or physically adsorbed on the surfaces. These model molecules must display the
same chemical functionalities, i.e., long hydrocarbon chains containing additives,
in order to simulate the tribochemical reactions that occur under boundary lubrica-
tion conditions. Third, friction experiments can be carried out under a controlled
partial pressure of gas. In this way, the conditions are very similar to those existing
under boundary lubrication conditions because the friction occurs on the molecules
adsorbed on the contacting surfaces.
A device dedicated to GPL, coupling an environmentally controlled tribometer
and XPS/AES surface analysis system, named ECAT (Environmentally Controlled
Analytical Tribometer) was built in the LTDS laboratory (Fig. 4.1) [10]. This device
is composed of three independent parts: a UHV chamber for tribometry, a surface
analysis system and a UHV chamber for the preparation and cleaning of samples.
ECAT allows friction experiments to be carried out under UHV or controlled gas
pressures. Pure or mixed gases can be introduced, and the inlet gas pressure can be
monitored from 10−9 to 2000 hPa with a membrane vacuum gauge. The purity and
proportion of inlet gases can be controlled by a residual gas analyser. This anal-
yser also permits the nature of molecules generated during the friction experiments
to be analysed, providing complementary, interesting information about the tribo-
chemical mechanism [11]. Friction experiments under a controlled environment were
conducted with a reciprocating pin-on-flat tribometer over a wide range of contact
pressures and sliding speeds. The flat temperature can be varied from liquid nitrogen
temperature (approximately −100 °C) to 800 °C. For the experiments discussed in
the following paragraphs, a normal load of 3.5 N was applied, corresponding to a
maximum Hertzian contact pressure of 0.52 GPa and a track width of 114 µm. The
track length was adjusted to 2 mm, and the sliding speed of the reciprocating motion
110 M.-I. De Barros Bouchet and J.-M. Martin

Fig. 4.1 ECAT device equipped with a tribometry chamber (green), XPS/AES surface analysis
system (blue) and UHV preparation chamber (orange) [10]

was 0.5 mm/s. All experiments were performed at ambient temperature (flat tem-
perature of approximately 25 °C). This tribometer can measure friction coefficients
with both very low (milli range) and very high values (above 1). The applied normal
load was adjusted automatically as the partial pressure changed.
In addition, this device makes it possible to carry out in situ XPS and Auger anal-
yses. For these analyses, the tribometry chamber is connected to a surface analysis
system by an intermediate UHV chamber dedicated to the cleaning (heat treatment
and surface abrasion) and preparation (thermal evaporation) of samples (Fig. 4.1).
This intermediate chamber prevents the worn surfaces from being contaminated by
the gases used in the GPL experiments and allows transfer of the sample with-
out exposure to air. XPS analyses were carried out with a focused (250 µm) and
monochromatic Al X-ray source, and AES analyses were performed with an elec-
tron gun FEG1000 (spot size of approximately 0.1 µm). The photoelectrons emitted
by the surface during XPS/AES analyses were detected by a VG 220i spectrometer.
First, an XPS survey spectrum covering a range of 1200 eV was obtained to identify
all elements and the presence of contaminants. Afterwards, a more detailed scan-
ning of the individual elements of interest over a smaller range of 15–20 eV was
undertaken to establish the different chemical states of the species and to perform
a quantitative analysis. XPS photopeaks were fitted with a Shirley background, and
atomic quantifications were made with the Scofield table of sensitivity factors. All
peaks were calibrated against the C 1s level of adventitious carbon at a binding energy
4 Tribochemistry of n-Alkane Thiols Examined … 111

Molecules in M
Motion
vapour phase hν e¯

e¯
L
Load

1) Modell molecule for 2) GPL fricttion experiments 3) In situ XPSS/AES surfaces

Lubricatting Additive analyses of adso
orbed films and
(same chemical function) triboffilms

Fig. 4.2 Schematic of the experimental simulation of tribochemical reactions using GPL/in situ
XPS/AES analyses

(BE) of 284.8 eV. Argon ions from a VG EXO5 ion gun were used for etching to
clean the surfaces before the friction experiments and to measure XPS/AES depth
profiles. In comparison with XPS, AES analysis presents a strong spatial resolu-
tion of approximately 0.1 µm (versus 100 µm for XPS), making possible accurate
measurements inside the tribofilm.
Overall, GPL coupled with in situ XPS/AES surface analyses appears to be an
efficient way to experimentally simulate real, complex tribochemical reactions. The
operation of this approach, schematically displayed in Fig. 4.2, shows two main
advantages:
– No cleaning of the rubbed surfaces is required, avoiding any pollution and/or
deterioration of the surfaces by the cleaning solvents in the case of liquid-phase
lubrication;
– Perfect control of the chemical nature of the surfaces before and after the friction
experiment (due to surface etching in the preparation chamber before the exper-
iment) can be obtained without any issues of contamination by the surrounding
air.
This can bring new knowledge of the complex phenomena that can occur in lubri-
cated friction experiments. Moreover, such approach appears particularly suited for
combination with numerical simulations, offering unique capabilities for investigat-
ing boundary lubricated interfaces.

4.2.2 Materials

The tribochemical reactions of different kinds of lubricating additives currently used

in industry were previously investigated by the authors using GPL experimental
simulations. For example, phosphite, phosphate and borate additives were simu-
lated [12–17]. The friction results obtained by GPL were corroborated by the results
obtained in liquid-phase lubrication. New insights into the chemical composition of
the tribofilms generated in the presence of these additives as well as more detailed
112 M.-I. De Barros Bouchet and J.-M. Martin

information about their reaction mechanisms were obtained using GPL [13]. This
chapter focuses on the investigation of tribochemical reactions of sulfur-based com-
pounds. These compounds have been used in many applications, such as biosensors
[18, 19], corrosion inhibitors [20] and lubricant additives [21–26]. Different studies
in the literature have focused on N-alkanethiols and, in particular, their adsorption
mechanism on different metallic surfaces, such as gold, nickel, silver and iron. How-
ever, until now, the detailed adsorption mechanism is still not fully understood [27].
The authors have shown that these compounds can adsorb on metallic surfaces to
form stable self-assembled monolayers (SAMs). The structure and properties of
these SAMs have been studied by different surface techniques, such as spectroscopy,
microscopy and electrochemistry, to understand their formation mechanism. Self-
chemisorption seems to be a particularly attractive approach in this field. The growth
mechanism of SAMs on metallic surfaces strongly depends on the molecule-molecule
and molecule-substrate interactions [28, 29]. It has been shown that the lateral carbon
chain and deposition temperature play important roles in both the formation and the
stabilization of the SAMs [30–32]. The –SH functional group promotes the adhesion
of organic layers to the metal surface. The authors, by studying the mechanism of
the decomposition of thiol compounds on different metallic surfaces, showed that
S–H bond scission occurs upon chemisorption at low temperature, forming adsorbed
alkylthiolate (R–S–Fe) [21, 33–37]. It appears that the nature of the bond formed
between the adsorbate and substrate has an influence on the structure of the mono-
layer films formed by these species [21]. At high temperature, the S–C bond in the
adsorbed thiolate breaks, and atomic sulfur remains on the surface. This thermal
decomposition at high temperature results in the formation of gas products and the
adsorption of alkyl groups on the metallic surfaces [35–37]. Authors have also stud-
ied the effect of an oxide layer using different oxidized metallic surfaces. They found
that a thin oxide layer is sufficient to oxidize thiol to sulfonate on iron oxide surfaces
[38, 39]. Different components, such as sulfinates and sulfonates, were observed to
have similar behaviour on aluminium oxide [40] and titanium oxide surfaces [41]. In
contrast, on clean iron substrates without an oxide layer, a chemisorptive S–Fe bond
is observed between the thiol group and the substrate [20]. The adsorption kinetics
of thiols on oxidized metals are quite different from those on noble metals [39].
However, the formation and properties of SAMs on gold substrates, in particular,
those derived from alkanethiols, are among the most widely studied because such
SAMs are easily prepared and chemically stable and present well-ordered monolayer
structures [28]. Much less is known regarding the reaction between alkanethiols and
steel substrates. In a previous work [42], the friction and wear behaviour of thiol and
polysulfide compounds were compared using GPL. This study showed that the good
friction modification and anti-wear performances of thiols under moderate pressure
conditions are related to the formation of an iron sulfide-rich tribofilm. However, the
key mechanism controlling its formation could not be elucidated accurately because
different activation sources can operate simultaneously. Moreover, the microstructure
of such a tribofilm, in particular, its stoichiometry, structure, thickness and interface
with the substrate, is not known. In this paper, we investigate the adsorption and tri-
bochemical reaction mechanisms of 1-hexanethiol on two different surfaces: (i) steel
4 Tribochemistry of n-Alkane Thiols Examined … 113

with native iron oxide layers and (ii) an ion-etched steel surface, revealing metallic
iron. Moreover, we design a model GPL experiment to separate the different poten-
tial activation sources of tribofilm formation described in the literature (i.e., thermal
effects, nascent surfaces, molecular shearing and exo-electron emission). For our
study, we choose 1-hexanethiol (C6 H13 -SH), supplied by ARKEMA, which is liquid
at room temperature. 1-Hexanethiol can be easily evaporated and introduced into the
UHV chamber dedicated to tribometry at a limited pressure at room temperature,
corresponding to the saturation vapor pressure (SVP) (Fig. 4.2). This compound was
further purified by freeze–pumping–thaw cycles before being introduced into the
tribometry chamber. Iron sulfide (FeS) was used as a reference for the AES surface
analysis.
The friction materials, pin and flat, were made of AISI 52100 steel (composition
96.9 Fe–1.04 C–1.45 Cr–0.35 Mn–0.27 Si, in wt%), and both surfaces were polished
with a 1 µm diamond paste solution to obtain a surface roughness average (Ra)
of approximately 20–25 nm and then cleaned with n-heptane and 2-propanol in
ultrasonic baths. The use of steel is preferred to crystalline pure iron material because
the hardness of pure iron is too low to easily reach the required contact pressure in
real applications. The pin has a hemispherical radius of 8 mm. The 3–4 nm-thick
native oxide/hydroxide layer present at the top surface of the steel samples can be
removed using a VG EXO5 Ar ion gun (ions of 3 keV) to reveal a metallic pure iron
surface. Before starting the GPL experiments, the etched sample was first analysed
by XPS to check its iron purity, i.e., the absence of any adventitious carbon or residual
oxide layers.

4.3 Adsorption of Hexanethiol on Steel

4.3.1 Procedure for Adsorption Study

1-Hexanethiol was deposited as droplets on the steel substrate and kept in contact for
2 h. The excess liquid was removed with paper so that a very thin liquid layer remained
on the surface. The sample was then introduced into the XPS analysis chamber. After
XPS analysis, the sample was removed from the analysis chamber and cleaned with
heptane for 30 min in an ultrasonic bath. After ultrasonic cleaning, this sample was
analysed again by XPS to determine whether 1-hexanethiol chemisorption occurred
on the substrate. To study adsorption on a pure iron surface, gaseous 1-hexanethiol
was directly introduced in the tribometry chamber at a vapor pressure of 0.1 hPa to an
ion-etched steel surface. After the 1 h adsorption experiment, the flat was transferred
without exposure to air to the analytical chamber to be analysed by XPS (see Fig. 4.2).
114 M.-I. De Barros Bouchet and J.-M. Martin

S2p3/2 1-hexaneth
hiol

S2p11/2
S2p3/2
S2p1/2

Fig. 4.3 S2p XPS spectra of liquid 1-hexanethiol deposited on the native oxide layer of steel a
before and b after ultrasonic cleaning with heptane

4.3.2 Adsorption of Hexanethiol on the Native Oxide Layer

of Steel

At ambient temperature, it is well known that bearing steel is covered by a 2–3 nm-
thick native oxide/hydroxide layer, as revealed by XPS [43]. A survey spectrum,
covering a range of 1200 eV, was first recorded to identify the elements present after
the adsorption experiment in the presence of 1-hexanethiol. The results revealed the
presence of sulfur, carbon, oxygen and iron on the surface. Figure 4.3 compares the
S 2p XPS spectra of 1-hexanethiol covering the native oxide layer before and after
ultrasonic cleaning with heptane.
Notably, the sulfur S2p photopeaks recorded before the cleaning step are the result
of two contributions, the first peak at approximately 163 eV BE (S2p3/2 ) corresponds
to the H–S–C bond present in the 1-hexanethiol molecule. This peak completely dis-
appears after heptane cleaning, confirming that 1-hexanethiol does not chemically
adsorb as a thiolate on the native oxide layer of steel. However, another weaker pho-
topeak at higher binding energy (S2p3/2 core level at 168.2 eV BE) is detected before
and after cleaning, although the peak is quite smaller in the latter case. According to
the literature data [19, 20], this peak can be assigned to the sulfonate bond (R–SO3 )
fixed on the oxide surface. The oxygen XPS photopeak is composed of two contribu-
tions: one located at 530 eV and attributed to iron oxide species, and another located
at 532 eV and attributed to –O–C bonding. The iron Fe2P3/2 photopeak at 710.8 eV
is assigned to iron oxides, Fe2 O3 and Fe3 O4 . To summarize, the XPS results show
4 Tribochemistry of n-Alkane Thiols Examined … 115

(b)

Fig. 4.4 a S2p XPS spectrum of liquid 1-hexanethiol covering steel, b S2p and C1s XPS spectra
of gaseous hexanethiol adsorbed on metallic iron (etched steel), and c S2p XPS spectrum of iron
sulfide (FeS) for reference

that n-alkanethiol molecules are weakly bound to oxygen atoms from the oxide layer
by sulfur atoms, forming a few sulfonate species.

4.3.3 Adsorption of Hexanethiol on Metallic Iron (Etched

Steel Surface)

1-Hexanethiol was introduced at a vapor pressure of 10−1 hPa in the tribometry

chamber and kept in contact for 2 h with an iron surface obtained by ion-etching.
At the end of the adsorption experiment, the gas was pumped down to 10−8 hPa,
and the sample was transferred into the analytical chamber. Figure 4.4 compares
the XPS results obtained after the adsorption of 1-hexanethiol on a pure iron surface
(ion-etched steel surface) (Fig. 4.4b) and on the native oxide layer of steel (Fig. 4.4a).
The S2p spectrum of the iron sulfide standard is also shown for reference (Fig. 4.4c).
The S2p3/2 XPS spectrum (Fig. 4.4b) reveals a sulfur contribution of weak intensity
at 161.5 eV BE, which is very close to the value reported for iron sulfide (Fe–S bond),
as shown in Fig. 4.4c. This peak is easily distinguished from the R–S–H bond in the
1-hexanethiol molecule at 163 eV BE (Fig. 4.4a). The carbon C1s signal (Fig. 4.4b)
is composed of a first peak at 284.8 eV BE, corresponding to the C–C/C–H bond, and
a second peak at 282.8 eV BE, typical of the carbide state (C–Fe or C–Cr bond). It
was observed that both the sulfur peak (S2p3/2 at 161.5 eV BE) and the carbon peak
(C1s at 284.8 eV BE) disappear from the surface after ion-etching, but the carbide
peak is still visible and may originate from chromium carbide present in the AISI
52100 steel composition. The iron Fe2p3/2 peak at 706.8 eV is attributed to metallic
116 M.-I. De Barros Bouchet and J.-M. Martin

iron and/or iron carbide, because these species have almost the same binding energy
(not shown here).
This first section clarifies the reactivity of 1-hexanethiol on steel surfaces at room
temperature. We showed that 1-hexanethiol weakly interacts with iron oxide present
on steel substrates by forming a sulfonate bond. In contrast, 1-hexanethiol chemisorbs
on pure metallic iron by cleavage of the S–H bond, forming an iron sulfide and
possibly iron carbide. However, the kinetics of this reaction appear to be very slow,
as shown by the intensity of the S2p photopeak. This could be attributed to the
modification of the metallic iron surface due to ion implantation during the etching
process and to the lower molecular concentration in the gas phase in comparison
with the liquid phase. This result will be discussed in the following section with the
role of mechanically activated surfaces.

4.4 Model GPL Experiment for Understanding

the Formation Mechanism of Hexanethiol-Based
Tribofilm

It is still extremely challenging to provide an accurate description of the surface

chemistry and interfacial properties of a tribofilm, which are relevant for under-
standing the functionality of lubricant additives, due to the need to clean rubbed
surfaces before analysis (ex situ analyses). Part of the tribofilm may be removed by
the cleaning procedure, and exposure to air may partially and/or totally oxidize its
elements. This is particularly the case for adsorbed films, which are usually extremely
thin. That is the reason why the ECAT system was developed, and the tribometer is
schematically shown in Fig. 4.5. We illustrate the different steps used to accurately
investigate both the adsorption and tribofilm formation mechanisms of additives on
tribo-stressed metallic iron. First, the heptane- and propanol-cleaned steel surface
is introduced in the UHV chamber, and the oxide layers are removed by Ar+ iron
etching, the metallic nature of the surface after etching can be then checked by XPS
or AES (step 1 in Fig. 4.5). Second, a friction experiment under UHV (below 10−8
hPa) is performed to generate a nascent tribo-stressed metallic iron surface inside
the wear scar (red track formed during step 2 in Fig. 4.5). Third, both the etched
and tribo-stressed surfaces are exposed to gaseous hexanethiol for adsorption studies
(step 3 in Fig. 4.5). Fourth, the friction experiment is performed in the presence of
gaseous hexanethiol to form a tribofilm (blue track formed during step 4 in Fig. 4.5).
Finally, the three different surfaces (adsorbed film on etched steel, adsorbed film on
nascent tribo-stressed iron and thiol-derived tribofilm) are analysed by in situ XPS
and AES analyses (step 5 in Fig. 4.5). This approach, which combines GPL and
in situ surface analyses, is very efficient to study the tribochemistry mechanisms of
lubricant additives and highlight the critical role of nascent metallic surfaces in the
dissociation of the additives under boundary lubrication conditions.
4 Tribochemistry of n-Alkane Thiols Examined … 117

1- Ion etching of pin/flat 2- Friction under UHV 3- Exposure to gas 4- Friction under GPL Pg 5- XPS/AES analyses
Removal of oxide layers (1st track in red) partial pressure Pg (2nd track in blue) Inside tracks and outside

Fig. 4.5 Model GPL experiment for elucidating the origin of tribochemical reactions of lubricating
additives

4.4.1 Tribochemical Reactions of Hexanethiol on Oxidized

Steel and Metallic Iron

First, the friction behaviours of oxidized steel and its metallic iron counterpart in the
presence of gaseous hexanethiol were compared. For this, two-pass friction experi-
ments were performed in the presence of 10−1 hPa 1-hexanethiol with metallic iron
(pin and flat) surfaces (blue friction curve) and with oxidized steel (pin and flat)
surfaces (green friction curve). Figure 4.6 shows the central part of the friction coef-
ficient recorded during the back and forth motion in order to avoid the perturbing
effect of changes in the sliding direction. In the case of the etched surface (Fe/Fe),
the friction coefficient is initially low for a few cycles; then increases rapidly to high
values near unity, indicating welding between the iron atoms from each side; and then
begins to decrease until the end of the first pass. During the second (reverse) pass,
the friction is much lower. This suggests that the chemisorbed layer on the nascent
tribo-stressed iron flat was quickly formed during the first pass at the rear of the
contact zone and is highly efficient at reducing friction during the reverse pathway.
This tribolayer formed on metallic iron can be seen in the optical image of the track
presented in Fig. 4.6b. The situation is very different if the contacting surfaces are
oxidized (non-etched). The friction coefficient is almost stable during the first pass
and increases during the second pass, reaching a value of approximately 0.6–0.7 at
the end of the two passes. This result seems to indicate that no reaction occurred
between oxidized steel and hexanethiol during the two friction passes. This result is
in good agreement with the optical image of the track presented in Fig. 4.6c and the
weak gas adsorption on oxidized steel observed previously (Fig. 4.3).
At the end of these GPL experiments, gaseous hexanethiol is pumped down, and
AES analysis is performed on the different rubbed surfaces. It is important to note
that the time necessary to reach a high vacuum below 10−8 hPa in the chamber is
at least 1 h (3600 s), so some hydrocarbon adsorption from the residual gas on the
steel surfaces occurs before analysis. Indeed, at a gas partial pressure of 10 nPa,
the time for a monolayer to adsorb is approximately 104 s. Figure 4.7 compares
the in situ AES spectra recorded inside the tracks formed on both the oxidized and
118 M.-I. De Barros Bouchet and J.-M. Martin

(a) (b)
1.2 on metallic iron
Etched steel (metallic iron)

1 Steel (oxidized surface)

Passage 1
Passage 2
Friction coefficient

0.8
(Fe/Fe in 10-8 hPa)

0.6

Passage 1
0.4 (c)
on oxidized surface
0.2

Passage 2
0
0 50 100 150 200 250 300 350 400 450 500
Number of friction coefficient measurements

Fig. 4.6 a Evolution of the friction coefficient (for two passes, back and forth) obtained under GPL
in the presence of 10−1 hPa pressure of 1-hexanethiol at room temperature. The friction curves in
blue represent metallic iron surfaces (ion-etched steel), and the friction curves in green represent
oxidized steel surfaces. The steady-state friction level for an etched steel surface under UHV is
shown for reference. b, c Optical images of the corresponding tracks formed on the metallic iron
flat and the steel flat

etched steel surfaces. In the spectrum recorded inside the tribofilm formed on etched
steel, the formation of iron sulfide is clearly evidenced by the strong SLMM peak
at approximately 151 eV KE, confirming the decomposition of hexanethiol under
shear and the release of S elements to form S–Fe bonds. It is noticed that no sulfate
bonding is detected in the spectrum near 158.5 eV KE, which is in good agreement
with the quite low amount of oxygen in the spectrum. For the steel surface covered
by a native oxide layer, sulfur is not clearly detected in the AES spectrum. The
friction reduction property (Fig. 4.6) is correlated with the quick formation of iron
sulfide in the tribofilm on etched steel. These experimental results are in agreement
with previous first-principles calculation results, which revealed that sulfur is very
efficient at reducing the adhesion and shear strength at the iron-iron interface [17].

4.4.2 Study of the Role of Nascent Metallic Surfaces

on the Tribochemistry of Hexanethiol

Additional AES spectra were recorded after exposure of the tribo-stressed nascent
metallic surface (red) and the etched steel surface (outside the tracks) to 10−1 hPa hex-
anethiol to compare the nature of both adsorbed species and the tribofilm. Figure 4.8
compares the three different in situ AES spectra recorded. First, practically no oxy-
gen is detected on the different surfaces (O KLL at 530 eV KE), confirming that no
contamination by oxygen or water from the gaseous atmosphere occurred in any
4 Tribochemistry of n-Alkane Thiols Examined … 119

FeLMM
OKLL
SLMM CKLL
Track formed on etched steel
(metallic iron )

Track formed on steel

(oxidized surface)

100 200 300 400 500 600 700 800

Kinetic energy (eV)

Fig. 4.7 In situ AES spectra recorded inside the tracks formed after the two-pass friction experi-
ments with 10−1 hPa hexanethiol at room temperature on oxidized and etched steel surfaces

sequence (total duration of the experiment 2 h). As already observed by the authors,
iron sulfide is quickly formed by tribochemical reaction (blue), whereas a very small
amount of sulfur can be detected outside the tribofilm due to molecular adsorption
on the etched steel. In the AES spectrum in red, corresponding to the friction track
formed under UHV followed by exposure to the gas phase without friction, a small
but visible signal is detectable at approximately 151 eV KE, in comparison with the
spectrum registered outside the track. This peak indicates the possible formation of
iron sulfide. From this result, we can conclude that the tribo-stressed iron surface,
but not the ion-etched surface, chemically reacted with 1-hexanethiol. It is noticed
that the mean free path of 150 eV electrons in the substrate is very small (approx-
imately 1 nm), which is very different from the iron 700 eV electrons (mean free
path of 3.5 nm). Therefore, the sulfur signal results from the first atomic layers, and
the quantity of iron sulfide formed from the interaction of the molecules with the
mechanically activated surface is very small, i.e., a partial monolayer.
From this model experiment, we can deduce some important features of the tri-
bochemical reaction of 1-hexanethiol with steel surfaces:
– 1-Hexanethiol does not form iron sulfide in the presence of a native oxide layer
on steel, but some sulfonate bonding was detected. This is in agreement with the
chemical hardness model developed by Pearson [44], which states that a soft base
(R-SH) prefers to react with a soft acid (metallic iron).
– Surprisingly, hexanethiol does not react significantly with pure metallic iron
obtained by ion-etching the steel surface (free of carbon and oxygen). This is
supported by the adsorption experiments shown in Fig. 4.4. This could be tenta-
tively explained by the effect of Ar ion implantation, which can disturb the surface
and remove some crystal defects. However, the presence of Ar was not detected
in the AES spectra (peak at approximately 210 eV KE).
120 M.-I. De Barros Bouchet and J.-M. Martin

Fe LMM

O KLL

C KLL
S LMM

Tribofilm surface

Tribo-stressed nascent
surface

Etched steel
surface

100 200 300 400 500 600 700 800

Kinetic Energy (eV)

Fig. 4.8 In situ AES spectra recorded after the model GPL experiment on the tribofilm formed
during the two-passes friction experiment with 10−1 hPa hexanethiol (blue), on the tribo-stressed
nascent surface (red) and on the etched steel surface exposed to 10−1 hPa hexanethiol (purple)

– 1-Hexanethiol molecules dissociate on a tribo-stressed metallic iron surface

(Fig. 4.8) by forming sulfide bonds (Fe–S). This dissociative reaction occurred
approximately 1 h after the creation of the nascent tribo-stressed metallic iron sur-
face. Therefore, neither the contact temperature nor electron emission is respon-
sible for this dissociative reaction. The creation of crystal defects and dangling
bonds at the metallic surface caused by high friction under UHV is certainly the
origin of the activation energy of this thiol decomposition mechanism.
– 1-Hexanethiol rapidly forms a tribofilm containing iron sulfide in two passes on
a metallic iron surface (Fig. 4.6). The detailed analysis of the thiol-metallic iron
surface interaction as a function of the surface chemistry shows that the thiol
molecules completely dissociate on the nascent surface at the contact zone exit.
This confirms that the creation of the nascent metallic iron surface by mechanical
action is one of the energy sources responsible for tribofilm formation. Once
the iron sulfide accumulated in the tribofilm after successive passes, the friction
greatly decreased. The formation of iron sulfide appeared to be widely enhanced by
tribological solicitation, in particular, by shearing stress occurring at the colliding
asperities.

4.5 Ex Situ Analyses of the Iron Sulfide Tribofilm Formed

by Thiol

The GPL experiments coupled with in situ surface analyses showed that 1-
hexanethiol efficiently reduced friction. This is correlated with the formation of iron
4 Tribochemistry of n-Alkane Thiols Examined … 121

sulfide inside the tribofilm. We highlighted the role of the active nascent metallic
surface on the dissociation of adsorbed molecules and the activation of the tribo-
chemical reaction to rapidly form a protective iron sulfide tribofilm. However, the
microstructure of such tribofilm is not clearly known. Previous in situ XPS/AES
chemical analyses have identified the chemistry of the top surface of the tribofilm (a
few nm depth) but did not provide any information on the structure of the tribofilm
and its interface with the steel substrate. Moreover, the exact stoichiometry of iron
sulfide cannot be easily determined due to technical difficulties in accurately quanti-
fying such compound and the existence of various forms of iron sulfide (FeS, FeS2 ).
The aim of this final section is to fully characterize the iron sulfide tribofilm formed
by 1-hexanethiol on steel substrates, in particular, its stoichiometry, structure, thick-
ness and the interfacial region with the steel substrate. A multi-technique approach is
proposed, combining focused ion beam (FIB) sample preparation, transmission elec-
tron microscopy (TEM) observation, chemical analysis by energy dispersive spec-
troscopy (EDS) and time-of-flight secondary-ion mass spectrometry (ToF-SIMS).
The studied tribofilm was obtained using GPL, taking advantage of this approach to
control the contact surfaces, without the need to clean the surfaces after the friction
experiment. ToF-SIMS analysis was performed on the top surface of the tribofilm,
and TEM observations coupled with EDS analyses were carried out on a transversal
cross section of the tribofilm prepared by FIB.

4.5.1 Formation of a 1-Hexanethiol Tribofilm by GPL

To clearly identify the nature of the tribofilm and its interfacial zone with the
steel substrate, thicker tribofilms were generated by performing long-time friction
experiments. The operating tribological conditions were similar to those previously
described, except that 1-hexanethiol gas was introduced into the tribometry chamber
of the ECAT system at a high pressure of 3 hPa and the duration of the experiment
was approximately 1 h.
Figure 4.9 shows the evolution of the friction curve for the experiment in the
presence of 3 hPa 1-hexanethiol gas. First, the friction coefficient slightly increased
to 0.18 and then quickly decreased and stabilized at 0.15 (15th cycle). This value
is in agreement with the formation of a tribofilm containing iron sulfide [42]. This
tribofilm is clearly visible by optical microscopy, and some debris are present around
the track, as shown in the inset of Fig. 4.9. However, 1-hexanethiol seems to have
a good anti-wear property because both the width of the wear track on the steel flat
(inset in Fig. 4.9) and the diameter of the wear track on the hemispherical pin almost
correspond to the calculated initial Hertzian diameter, i.e., 114 mm.
122 M.-I. De Barros Bouchet and J.-M. Martin

4.5.2 ToF-SIMS Chemical Analysis of the 1-Hexanethiol

Tribofilm Formed by GPL

ToF-SIMS analysis is performed on a Physical Electronics TRIFT III ToF-SIMS

instrument operated with a pulsed 22 keV Au+ ion gun (ion current of approximately
2 nA) rastered over a 100 µm × 100 µm area. No charge compensation was needed.
The ion dose was kept below the static conditions limit. The data were analysed
using the WinCadence software. Mass calibration was performed on the hydrocarbon
secondary ions. The data were normalized to the total intensity of the secondary
ions minus the ion contributions resulting from hydrogen and contaminants. The
corrective eliminations were implemented because of their critical dependence on
slight variations in the experimental settings (such as H ±) or because they were
related to the signatures of contaminants usually observed in ToF-SIMS (such as
alkalis and fatty acids).
The ToF-SIMS analysis focused specifically on the question of the exact stoi-
chiometry of iron sulfides formed during the tribochemical reaction of 1-hexanethiol
on the steel surface. For this, four different analyses were undertaken to compare the
inside and outside of the tribofilm formed by hexanethiol and two references, i.e.,
the hexanethiol molecule and virgin steel. The standard deviations were calculated
from several measurements in different areas for each case. Figure 4.10 presents
the ion intensity ratios FeS+ /Fe+ and FeS2 + /Fe+ recorded inside and outside the tri-
bofilm. The chemical signatures of the steel substrate and the sulfur-containing thiol
molecule could easily be identified via characteristic Fe+ and S− ions. In contrast,

0.25
Hexanethiol (3 hPa)
121

0.20
Friction coefficient

0.15

0.10

0.05

0
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Number of cycles

Fig. 4.9 Evolution of the friction coefficient obtained for the etched steel/etched steel friction pair
under GPL in the presence of 3 hPa 1-hexanethiol versus the number of cycles at room temperature.
The corresponding track formed on the etched steel flat at the end of the experiment is shown in the
inset
4 Tribochemistry of n-Alkane Thiols Examined … 123

Fig. 4.10 ToF-SIMS

intensity ratios FeS+ /Fe+ and FeS+/Fe+
FeS2 + /Fe+ recorded in

Intensities ratio X104

FeS2+/Fe+
positive-ion mode inside
(2A-2D) and outside the
tribofilm (2Y-2Z) formed on
steel in the presence of 3 hPa
1-hexanethiol

no specific signature for the hexanethiol molecule could be found (CxHySz− ions
were not detected). The normalized intensities of the Fe+ , S+ , FeS+ and FeS2 + ions
and the FeS+ /Fe+ and FeS2 + /Fe+ ion intensity ratios were used to differentiate the
spectra (so-called univariate ToF-SIMS data analysis). The results obtained from this
univariate analysis (Fig. 4.10) confirmed the presence of a notable amount of these
species, depending on whether the zones are contacted or not. The results seem to
indicate that the tribofilm formed by hexanethiol is characterized by a higher relative
FeS+ intensity. However, this univariate data analysis made on a small number of
molecular signatures is inherently incomplete and may lead to artefacts of interpre-
tation. To confirm and complete our understanding of the nature of the iron sulfide
created by the tribochemical reaction of hexanethiol on a steel surface, TEM obser-
vations combined with EDS analyses were carried out on a transversal cross section
of the tribofilm.

4.5.3 TEM Observations and EDS Chemical Analyses

of the 1-Hexanethiol Tribofilm Cross Section

A cross section was taken at the centre of the tribofilm formed under GPL with 3 hPa
1-hexanethiol. The cross section was thinned by FIB after deposition of a protective
resin on the surface of the tribofilm. We used Ga+ ions for nanomachining the sample.
A high-resolution JEOL 2010F microscope equipped with a field-emission gun oper-
ating at 200 kV accelerating voltage was utilized to perform the TEM observations
as well as the EDS analyses with a point-to-point resolution of 2 nm.
Figures 4.11 and 4.12 show a cross-sectional TEM image of the tribofilm. From
Fig. 4.12 recorded with higher-resolution, it is possible to distinguish 4 zones: the
steel substrate, the interfacial region, the tribofilm, and finally the Pt protective resin
used for FIB preparation. Cross-sectional observations allow the measurement of the
thickness and observation of the structure of the tribofilm. Even though the interface
between the tribofilm and the steel substrate is not clearly marked, it appears smooth,
and the thickness of the tribofilm is quite homogeneous and can be estimated to be
approximately 15–30 nm. This smooth interface may be related to the fact that the
124 M.-I. De Barros Bouchet and J.-M. Martin

Fig. 4.11 TEM

cross-sectional view of a

el
tribofilm formed on steel

Ste
under GPL in the presence of
3 hPa 1-hexanethiol at room
temperature for 1180 sliding
cycles

ofilm
Trib

in
15-3

es
0 nm

er
tiv
tec
pro
Pt
GPL friction experiment was performed directly on etched steel (metallic iron), so
that the quick dissociation and reaction of thiol molecules was possible. The measured
thickness is on the same order of magnitude as that of the tribofilms obtained from
liquid-phase lubrication, even though in GPL, the concentration of the active element
(here, sulfur) is generally lower than that in the liquid phase. Thus, the growth kinetics
seem to be comparable. In fact, when molecules are introduced in vacuum, access
to the surface is favoured. In the liquid phase, interaction with other additives or
viscous base oils may limit access to the surface and further tribochemical reactions.
This observation confirms the original hypothesis that GPL is a good way to simulate
tribochemical reactions.
Further details of the tribofilm are given by the higher-resolution image in
Fig. 4.13a. In addition to the quite homogeneous thickness and the defined inter-
face, it is possible to see that the tribofilm is composed of nanocrystalline areas
of approximately 3–5 nm in diameter embedded in an amorphous matrix. Chemi-
cal analysis by EDS conducted on the cross section of the tribofilm is presented in
Fig. 4.13b. We carried out EDS analysis at three specific zones: the interface with the
steel substrate, the centre of the tribofilm and a crystallite inside the tribofilm. Some
expected contaminants and artefacts were detected in all spectra, including copper
from the microscope grid and gallium from the FIB process. As expected, EDS anal-
ysis of the substrate reveals an iron-rich composition (81 at.%) with 9 and 10 at.%
oxygen and carbon, respectively. The nanocrystals inside the tribofilm are mainly
composed of iron and sulfur with a similar ratio of approximately 40 and 42 at.%,
respectively, which can be assigned to the stoichiometry of iron sulfide (FeS). It is not
unusual to find a crystalline structure for iron sulfide-based compounds because dif-
ferent crystalline structures can exist for these compounds. A large amount of carbon
4 Tribochemistry of n-Alkane Thiols Examined … 125

Fig. 4.12 HRTEM

cross-sectional view of a
tribofilm formed on steel

el
under GPL in the presence of

S te
3 hPa 1-hexanethiol at room
temperature for 1180 sliding
cycles

Trib
o film

e
rfac

n
r es i
Inte

tive
c
rote
Pt P
but little oxygen was detected inside the amorphous matrix of the tribofilm, as shown
by the intermediate EDS spectrum in Fig. 4.13b. Carbon may have been introduced
during tribofilm formation (via the 1-hexanethiol molecule, which contains 6 carbon
atoms) or after FIB preparation (adventitious contamination of the tribofilm). The
EDS line scans (Fig. 4.14) show the distribution of the main elements: platinum,
iron, sulfur, carbon and oxygen. The profiles of the S and Fe elements are similar,
and the elements are homogeneously distributed in the main part of the tribofilm.
Near the interface with steel substrate, the amount of sulfur decreases, while that of
iron increases. However, the absence of oxygen in the interfacial zone with steel may
be because the GPL experiment was performed directly on etched steel (Fig. 4.14).
The TEM cross-sectional observations, EDS analyses, and ToF-SIMS analyses
converge on the same conclusion about the composition of the tribofilm formed on the
etched steel surface in the presence of 1-hexanethiol. This surface is mainly composed
of iron sulfide with a stoichiometry close to FeS. Taking into account thermodynamics
considerations, the preferential formation of iron monosulfide is not corroborated.
By considering that iron sulfide formation is the result of the reaction of metallic
iron Fe with atomic sulfur S after decomposition of 1-hexanethiol, all reactions in
the formation of FeSx are characterized by a negative Gf0 , and thus, iron sulfide
forms spontaneously. However, the Gibbs free energy of formation Gf0 of iron
disulfide FeS2 (−167 kJ mol−1 ) is more favourable than that of iron monosulfide
FeS (−100 kJ mol−1 ) [45]. Consequently, the nucleation of this last phase is not
the most preferential. The energy dissipated at the tribological contact allows this
phase to be stabilized under friction. In addition, the thiol molecule contains only
one sulfur atom, which is not considered to be very “active” since it is bonded to a
carbon atom of the alkyl chain.
126 M.-I. De Barros Bouchet and J.-M. Martin

(a)
el
S te (b)

m
o fil
Tr ib
e
rfa c
In te

Fig. 4.13 a HRTEM cross-sectional view of a tribofilm formed on steel under GPL in the presence
of 3 hPa 1-hexanethiol at room temperature for 1180 sliding cycles and b EDS spectra recorded on
three different parts of the tribofilm

450
n
si
re
400 Pt

350
ee
l - Pt
ilm St
300 ib
of - Fe
Tr
-S
250 -O
-C
200

150

100

50

0
0 20 40 60 80 100 120

Fig. 4.14 EDS line scan analyses conducted on the cross section of the hexanethiol tribofilm
(yellow arrow)
4 Tribochemistry of n-Alkane Thiols Examined … 127

4.6 Conclusions

An approach combining gas-phase lubrication (GPL) and in situ XPS/AES analy-

ses was found to be very efficient to investigate the tribochemistry mechanisms of
lubricant additives under boundary lubrication conditions. Using a specific device
called ECAT, it is possible to perform friction experiments under different gaseous
atmospheres and to perfectly control the surface chemistry of the friction surfaces.
In situ XPS/AES analyses allowed us to clearly detail the chemistry of the tribofilm
formed by the lubricant additive under friction. This permitted us to determine the
principal activation energy of the tribochemical reaction due to the model GPL exper-
iments. Undoubtedly, we highlighted the primordial role of nascent tribo-stressed
metallic surfaces obtained after shearing in the dissociation of lubricant additives
under boundary lubrication conditions. The investigation into the reactivity of thiol
molecules showed that they suffer dissociative adsorption on the asperities of the
tribo-stressed nascent metallic surface, rapidly forming a thin protective film com-
posed of iron sulfide. Comparison with the tribofilm obtained by tribological action
under GPL showed that iron sulfide formation is largely enhanced by friction, in
particular, by shearing stress. For severe collisions at the asperities, the preferen-
tial reaction between thiol species and nascent iron metal leads to the formation
of an iron sulfide tribofilm, promoting the lubrication effect and efficient friction
reduction. The detailed analysis of the tribofilm formed with hexanethiol using a
multi-technique approach revealed that the tribofilm was mainly composed of iron
and sulfur. The tribofilm displays a nanocrystalline structure made of nanograins
of iron sulfide embedded in an amorphous carbon-rich matrix. The EDS analysis
showed a Fe/S ratio close to unity for thiol-derived tribofilm corresponding to iron
monosulfide.
Overall, this study indicates that the approach combining GPL and in situ sur-
face analyses can be a promising way to study the potential competitive/synergic
mechanisms occurring between additives in a formulated lubricant at the sliding
surface.

References

1. H. Spikes, Tribol. Lett. 60(5), 1–26 (2015)

2. F.P. Bowden, D. Tabor, The Friction and Lubrication of Solids, part II. Oxford University Press
(1964)
3. D.H. Buckley, J. Appl. Phys. 39, 4224 (1968)
4. A.G. Gellman, J. Vac. Sci. Tech. A 10, 180 (1992)
5. S. Mori, International Tribology Conference. Yokohama (1995)
6. K. Miyoshi, F.K. Honecy, P.B. Abel, S.V. Pepper, T. Spalvins, D.R. Wheeler, STLE Trans.
36(3), 351 (1993)
7. J. Fontaine, J.-L. Loubet, Th Le-Mogne, A. Grill, Tribol. Lett. 17(4), 709 (2009)
8. J.L. Lauer, S.R. Dwyer, Proceeding of the Japan International Tribology Conference, Nagoya,
vol. 989 (1990)
9. M. Boehm, J.M. Martin, C. Grossiord, Th Le Mogne, Tribol. Lett. 11(2), 83–90 (2001)
128 M.-I. De Barros Bouchet and J.-M. Martin

10. M.I. De Barros-Bouchet, M.C. Righi, D. Philippon, S. Mambingo-Doumbe, T. Le Mogne, J.M.

Martin, A. Bouffet, RSC Adv. 5, 49270 (2015)
11. J.M. Martin, M.I. De Barros Bouchet, C. Matta, Q. Zhang, W. Goddard III, S. Okuda, T.
Sagawa, J. Phys. Chem. C 114, 5003 (2010)
12. D. Philippon, M.I. De Barros Bouchet, Th Le Mogne, E. Gresser, J.M. Martin. Tribology 1(3),
113 (2007)
13. D. Philippon, M.-I. De Barros-Bouchet, Th Le Mogne, O. Lerasle, A. Bouffet, J.-M. Martin,
Tribol. Int. 44(6), 684 (2009)
14. D. Philippon, M.I. De Barros Bouchet, T. Le-Mogne, O. Lerasle, J.M. Martin, Tribol. Lett.
41(1), 73 (2011)
15. D. Philippon, M.I. De Barros-Bouchet, Th Le-Mogne, B. Vacher, O. Lerasle, J.M. Martin, Thin
Solid Films 524(1), 191 (2012)
16. M.C. Righi, S. Loehle, M.I. De Barros-Bouchet, D. Philippon, S. Mambingo-Doumbe, T. Le
Mogne, J.M. Martin, A. Bouffet, RSC Adv. 5, 101162 (2015)
17. M.C. Righi, S. Loehle, M.I. De Barros-Bouchet, S. Mambingo-Doumbe, J.M. Martin, RSC
Adv. 6, 47753 (2016)
18. H.H. Weetall, M.J. Lee, Appl. Biochem. Biotechnol. 22, 311 (1989)
19. S. Kulin, R. Khishore, J.B. Hubbard, Biophys. J. 83, 1965 (2002)
20. M. Volmer-Uebing, M. Stratmann, Appl. Surf. Sci. 55, 19 (1992)
21. B. Parker, A.J. Gellman, Surf. Sci. 292, 223 (1993)
22. W. Davey et al., Wear 1, 291 (1958)
23. E.S. Forbes, A.J.D. Reid, ASLE Trans. 16, 50 (1973)
24. S. Mori, K. Hori, Y. Tamai, Junkatsu 27, 505 (1982)
25. Z. Li, Y. Li, Y. Zhang, T. Ren, Y. Zhao, RSC Adv. 4, 25118 (2014)
26. J. Li, B. Fan, T. Ren, Y. Zhao, Tribol. Int. 88, 1 (2015)
27. M. Cohen-Atiya, D. Mandler, J. Electroanal. Chem. 550–551, 267 (2003)
28. F. Schreiber, Prog. Surf. Sci. 65, 151 (2000)
29. C. Vericat, M.E. Vela, G.A. Benitez, J.M. Gago, X. Torrelles, R.C. Salvarezza, J. Phys. Condens.
Matter 18, R867–R900 (2006)
30. N. Camillone, T. Leung, P. Schwartz, P. Eisenberger, G. Scoles, Langmuir 12, 2737 (1996)
31. S. Xu, S.J.N. Cruchon-Dupeyart, J.C. Garno, G.Y. Liu, G.K. Jennings, T.H. Yong, P.E. Laibinis,
J. Chem. Phys. 108, 5002 (1998)
32. W. Azzam, A. Bashir, A. Terfort, T. Strunskus, C. Woll, Langmuir 22, 3647 (2006)
33. M.R. Albert, J.P. Lu, L. Bernasek, S. Cameron, J.L. Gland, Surf. Sci. 206, 348 (1988)
34. J.A. Syed, S.A. Sardar, S. Yagi, K. Tanaka, Surf. Sci. 556–568, 597 (2004)
35. S.A. Sardar, J.A. Syed, E. Ikenaga, S. Yagi, T. Sekitani, S. Wada, M. Taniguchi, K. Tanaka,
Nucl. Instrum. Methods Phys. Res. 199, 240 (2003)
36. C. Kodama, T. Hayashi, H. Nozoye, Appl. Surf. Sci. 169–170, 264 (2001)
37. D.M. Jaffey, R. Madix, Surf. Sci. 311, 159 (1994)
38. M. Volmer, M. Stratmann, H. Viefhaus, Surf. Interface Anal. 16, 278 (1990)
39. C. Pirlot, J. Delhalle, J.J. Pireaux, Z. Mekhalif, Surf. Coat. Technol. 138, 166 (2001)
40. Z. Mekhalif, J. Delhalle, P. Lang, F. Garnier, J.J. Pireaux, J. Electrochem. Soc. 146, 2913 (1999)
41. Z. Mekhalif, J. Delhalle, P. Lang, F. Garnier, J.J. Pireaux, Synth. Metals 96, 16 (1998)
42. J. Tannous, B.M.I. de Bouchet, T. Le Mogne, P. Charles, J.M. Martin, Tribology 1, 98 (2007)
43. A. Rossi, F.M. Piras, D. Kim, A.J. Gellman, N.D. Spencer, Tribol. Lett. 23, 197 (2006)
44. R.G. Pearson, Chemical Hardness, 1st edn. (Wiley-VCH, Weinheim, 1995), p. 198
45. D.D. Wagman, W.H. Evans, V.B. Parker, R.H. Schumm, I. Halow, S.M. Bailey, K.L. Churney,
R.L. Nuttall, Am. Chem. Soc. Washington, DC (1982)
Chapter 5
In-Situ Measurement of Tribochemical
Processes in Ultrahigh Vacuum

Wilfred T. Tysoe

Abstract Monitoring chemical reactions occurring at a solid-solid interface is par-

ticularly challenging because of the problem of analyzing a buried interface with
surface-sensitive spectroscopic techniques. This can, to some extent, be addressed if
one of the contacting materials is transparent. In the case of optically opaque mate-
rials that are often of the greatest tribological interest, truly in-situ techniques are
limited to the detection of gas-phase products formed by rubbing in high vacuum
using a mass spectrometer, or by monitoring the contact resistance or friction coeffi-
cient variations during sliding. Optical techniques such as infrared spectroscopy can
be used when one of the materials is transparent. The results of such in-situ analy-
ses can be corroborated by using so-called pseudo in-situ techniques to analyze the
surfaces after rubbing without exposing the samples to the atmosphere. Examples of
such techniques are Auger spectroscopy and low-energy electron diffraction. Finally,
the use of these approaches is illustrated using a simple model tribochemical reaction
consisting of the gas-phase lubrication of copper by dimethyl disulfide.

5.1 Introduction

Being able to understand tribochemical processes and their reaction pathways is per-
haps the most challenging of surface science problems since the reactions occur at an
inaccessible, solid-solid interface. However, understanding such reaction pathways
is crucial since the majority of lubricants include additive molecules that react at the
sliding interface to form a film that reduces friction or wear, or both, and are used
in applications as diverse as machining, in the so-called extreme pressure regime,
where the interfacial temperatures are exceedingly high [1], to lubricants used in
engines and machines, where the conditions are less severe. A recent study suggests
that ~20% of energy used in the world is used to overcome friction and that it could

W. T. Tysoe (B)
Department of Chemistry and Biochemistry, University of Wisconsin
Milwaukee, Milwaukee, WI, USA
e-mail: wtt@uwm.edu

© Springer Nature Switzerland AG 2018 129

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_5
130 W. T. Tysoe

be possible to save as much as 17.5% of the energy use in road transportation, much
of it by improved lubricants [2].
The majority of the work carried out to understand the way in which lubricant
additives or gas-phase lubricants operate relies on post-mortem analyses, in which the
nature of the film is analysed after it has formed, to correlate the chemical composition
of the film with its mechanical properties. However, while this approach is useful to
understand the film properties that lead to improved tribological behavior, it generally
yields few insights into the surface chemical processes that result in the formation
of tribofilms. This is, in part, because chemical (including tribochemical) reactions
generally consist of a sequence of elementary-step reactions that constitute an overall
“reaction mechanism”. Only by understanding such mechanisms for tribochemical
reactions can the design of gas-phase lubricants and lubricant additives be placed on
a firm scientific basis, rather than relying on trial-and-error approaches. Achieving
such understanding requires that the nature of the reaction products be analysed as a
function of time.
Reaction rate constants depend on temperature and are often analysed in terms of
transition state theory [3]. For a simple, elementary-step reaction, A → B, the trans-
formation of A into B involves an initial increase in energy, known as the activation
energy, E act , before forming the product B. If this were not the case, A would spon-
taneously transform into B, and thus be an unstable compound. The configuration
of the system at the peak of this activation barrier is known as the transition state.
Transition state theory
 shows that the rate constant for the reaction, k, is given by:
k  A exp − kB T , where k B is the Boltzmann constant, T the absolute temperature
E act

in degrees Kelvin (K), and A is a pre-exponential factor, which is weakly temperature

dependent, so that the overall temperature dependence of the rate constant is dom-
inated by the exponential term, and this follows a so-called Arrhenius dependence.
The derivation of this equation from transition-state theory assumes that the reactant
(A) is in thermal equilibrium with the transition state. Thus, correctly modelling tri-
bochemical reactions requires both a knowledge of the interfacial temperature during
sliding and that the interface be in thermal equilibrium. This can impose restrictions
on the sliding conditions since molecule dynamics simulations have suggested that
the interface can deviate from thermal equilibrium during sliding [4], thereby inval-
idating the use of simple kinetic models. This also suggests that one mechanism by
which the rates of tribochemical reactions can be accelerated is simply by frictional
heating in the contact. Since the temperature appears as an exponent in the Arrhe-
nius form of the temperature-dependence of the rate constant shown above, relatively
small temperature changes can cause significant increases in chemical reaction rates.
The above considerations place quite severe strictures on carrying out-in-situ
analyses of the kinetics of tribochemical reactions. In addition, depending on the
contact geometry, which generally consists of a ball sliding on a flat surface, that
size of the contact region may be quite small, thus often requiring the use of surface
analytical techniques with good spatial resolution.
The most useful but most difficult systems to study are those which are of the
greatest interest to understanding tribochemistry, namely metal and oxide surfaces
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 131

since they are the most common in engineering systems. However, since they are not
transparent to electrons and photons that are most commonly used as the basis for
surface spectroscopic techniques, there are relatively few in-situ techniques that can
be applied to such systems.
A solution to such a problem is to use contacts in which one of the materials is
transparent to enable photon-in/photon-out techniques to be used for in-situ analyses
of the sliding interface [5]. While this does extend the range of techniques that can
be used, it does limit the tribochemical systems to ones that may not be relevant to
engineering interfaces. In addition, photon-based techniques tend to be less surface
sensitive that those based on electrons.
Since the range of truly in-situ techniques, even when using contacts in which
one of the materials is transparent, is rather limited, we include a class of pseudo
in-situ techniques. This consists of techniques that cannot be directly implemented
at the solid-solid contact, but can be sequentially applied to one of the surfaces after
it has been rubbed. These technique should analyze with the sample in the same
location at which rubbing occurs to allow the evolution of the nature of the interface
to be examined as a function of the number of times the sample has been rubbed.
This definition is distinct from truly ex-situ techniques where the sample has to be
removed from the tribometer in order to perform the analyses. Such ex-situ analyses
can potentially suffer from the problem that the nature of the surface may be different
from that at the sliding interface, either because it continues to evolve after having
been rubbed or because of contamination from the ambient.
The techniques that are available for the three hierarchies of in-situ analysis are
discussed in the next section.

5.2 In-Situ Analyses of Sliding Interfaces

5.2.1 In-Situ Analyses with Opaque Interfaces

Since such opaque interfaces, for example for metal-metal or ceramic-metal inter-
faces are opaque to both electrons and photons (light), there are no spectroscopic
techniques for studying such interfaces.

5.2.1.1 Detection of Gas-Phase Products During Sliding

While the nature of the film formed on the surface is of primary tribochemical interest,
the reaction of adsorbed gas-phase lubricants may evolve gas-phase products and the
formation of such products can provide insights into the reactions occurring on the
surface and their rates. However, in general, only a small amount of product is formed
from the small contact area usually found in sliding contacts, thus requiring sensitive
detection. In particular, it is not feasible to carry out such an experiment for lubricants
132 W. T. Tysoe

additives in a base oil, except after a very large number of rubbing cycles, and is most
informative for tribochemical reactions in high [6, 7] or ultrahigh vacuum [8], where
the background pressure is low. In this case, the products are detected using a mass
spectrometer that is placed in-line-of sight of and as close as possible to the contact.
The experiments can be carried out either with the sample pressurized by a reactant
(vapor phase lubrication) or by adsorbing a monolayer of adsorbate of the surface and
measuring the evolution of gas-phase species as the surface is rubbed. In the former
case, the mass spectrometer fragmentation masses of the products that are formed
by rubbing must be distinct from those of the vapor-phase lubricant to allow them to
be distinguished. This experiment allows the nature of the gas-phase product(s) to
be identified from the relative intensities of the features at various masses.
In the second case, where a monolayer of adsorbate is rubbed, the nature of the
desorbing species can similarly be identified from its mass spectrometer fragmenta-
tion pattern, and the surface reaction kinetics measured from the decay in the amount
of products evolved from the surface as a function of the number of times that it has
been rubbed, where the signal eventually decays to zero when all adsorbed species
have completely reacted.
This is illustrated for methyl thiolate (CH3 –S) species adsorbed onto copper by
dosing dimethyl disulfide, which reacts rapidly on copper by S–S bond scission at
~200 K. The resulting methyl thiolate thermally decompose to desorb hydrocarbons
(methane, ethylene and ethane) at ~425 K in temperature-programmed desorption
(TPD), corresponding to an activation energy of ~100 kJ/mol [9], resulting in the
deposition of sulfur on the surface [10, 11]. Here the additional hydrogen required
to form methane from methyl thiolate species was provided by formation of C2
hydrocarbons.
Sliding a tip over copper covered with methyl thiolate species at a substrate tem-
perature of ~300 K (well below the thermal decomposition temperature), causes the
formation of gas-phase methane where the amount produced by each scan of the tip
over the surface is displayed in Fig. 5.1. Monitoring signals at other masses confirmed
methane and C2 hydrocarbon formation, and control experiments were carried out to
establish that the methane was formed by rubbing [12]. Analyses of the surface after
rubbing (see Sect. 5.2) reveal that only sulfur remains on the surface after rubbing.
The integrated areas of the methane signal as a function of the number of passes of
the tip over the surface decrease (see inset to Fig. 5.1), eventually decreasing to zero
as the surface becomes completely depleted of adsorbed methyl thiolate species. This
reveals that methyl thiolate species decompose via a tribochemical (shear-induced
reaction) to evolve methane and deposit sulfur on the surface. The rate of this simple
reaction can be written as:
dΘ
−  kΘ n (5.1)
dt
where Θ, the coverage (in monolayers) defines the proportion of the surface covered
by the adsorbate, k is the rate constant of the tribochemical reaction, and n the reaction
order where n  1 for a first-order and n  2 for a second-order reaction. Here, a
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 133

1.8
1.6 Mass Spectrometer. Load = 0.44 N

Desorption Yield/a.u.
1.4 Scan speed=4x10-3 m/s
1.2
First Dose
1.0
Second Dose
16 amu Signal
0.8 Third Dose
0.6 Fourth Dose
0.4 Fifth Dose
0.2
0.0
1 2 3 4 5 6 7 8
Number of Scans

1 2 3 4 5 6 7 8 9
Scan Number

Fig. 5.1 The 16 amu (methane) signal measured by sliding a tungsten carbide tip with a copper
transfer film on a copper foil as a function of the number of scans at a sliding speed of 1 × 10−3
m/s with a normal load of 0.44 N. The inset shows the desorption yield measured from the area
under each methane pulse as a function of the number of scans. After collecting data for the first
DMDS dose while sliding at 4 × 10−3 m/s, until no more methane was detected (), the sample
was re-saturated with DMDS and the 16 amu signal again monitored (●). This experiment was
repeated for the third (▲), fourth (▼) and fifth (˛) DMDS doses

first-order reaction would imply that the reaction occurs by a single decomposition
step of an adsorbed thiolate species and a second-order reaction that it occurs via the
reaction between two adsorbates and so can provide mechanistic information.
If the time that an element on the surface spends in the contact per pass is given
by t C , then the proportion of the adsorbate that decomposes during each pass can be
calculated by solving (5.1) over a time t C . The total yield of gas-phase products for
each pass is simply the difference in coverages before and after the tip passes over
the surface. Writing k   ktc gives the yield of gas-phase products for the pth pass,
Y p , for a first-order reaction (n  1):
   ( p−1)
Y p  (1 − exp −k  ) exp −k  . (5.2)

Similar considerations for a second-order reaction (n  2) yields:

1 1
Yp  
− . (5.3)
1 + ( p − 1)k 1 + pk 

A linear plot of ln(Y p ) versus (p − 1) yields a straight line indicating that shear-
induced thiolate decomposition occurs via a first-order reaction, thereby allowing k 
to be measured [8, 13]. Measurement of the rate constant k requires a knowledge
134 W. T. Tysoe

of the contact time per pass tC and thus a knowledge of the contact area. However,
under a constant load, tC will vary as ν1 .

5.2.1.2 In-Situ Measurements of Coverages from Friction Forces

Perhaps the most common measurement in tribology is of the friction force or friction
coefficient (friction force divided by normal load). It has been found that the friction
coefficient of a surface covered by several species is proportional to the proportion
of the surface covered by each species (denoted as the coverage, Θ) [14]. Thus, if
during a tribochemical reaction, the surface contains a number of species, where
the coverage varies with time in the contact as the reaction proceeds, then if the
ith adsorbate has a time-dependent coverage Θi (t), with an associated characteristic
friction coefficient μi , then the time evolution of the friction coefficient μ(t) can be
written as:

μ(t)  μi Θi (t). (5.4)
i


with the constraint that i Θi  1. This method requires that the characteristic
friction coefficients of each species participating in the surface tribochemical reaction
be known, and is generally the most useful in confirming a proposed tribological
model.
However, this approach is particularly useful for analyzing simple, first-order
reactions, A(ads) → B(ads) since, μ A and μ B can be measured before and after
reaction to obtain the time-dependent coverages. In this case, (5.4) simplifies to:

μ(t)  (μ A − μ B )Θ A (t) − μ B , (5.5)

because of the constraint that the coverages sum to unity. The value of μ B is the
asymptotic friction coefficient of the sample after the reaction is complete and μ A
is the initial friction coefficient assuming that the extent of decomposition induced
by sliding during the first pass is negligible. For example, for a first-order reaction,
(5.5) becomes:
 
μ( p)  (μ A − μ B )exp −k  p − μ B . (5.6)

This approach has recently been exploited to follow shear-induced reactions in

an AFM [15].

5.2.1.3 Contact Resistance Measurements

In principle, contact resistances can provide information on the nature of the inter-
faces between opaque materials, and provide estimates of film thickness, t:
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 135

ρ AC
t , (5.7)
R
where ρ is the resistivity of the film material, R the contact resistance and AC the real
area of contact. Since the resistivity of the film material is not known, either because
of its uncertain composition and non-uniformity, and the real contact area is often not
well defined, even this simple interpretation makes it difficult to obtain qualitative
film thickness and this approach is primarily useful for qualitative estimates of film
thickness. The analysis is further complicated since the contact between real surface
occurs at asperity tips that have small dimensions that may influence their conductive
properties.

5.2.2 In-Situ Analyses of Transparent Interfaces

The notion that lubricating films could be interrogated with optical techniques with
transparent contacts was pioneered by Cameron, who used interferometric methods
to measure lubricant film thickness in-situ during sliding [16]. While such tech-
niques have not yet been applied to sliding interfaces in ultrahigh vacuum, pre-
sumably because of the difficulty in implementing optical techniques in ultrahigh
vacuum because of the potential utility of such techniques in surface analysis, the
applications of such techniques are briefly discussed. The majority of optical tech-
niques such as Raman spectroscopy have been used to study tribofilm formation in
air since the Raman scattering cross section is generally too low to measure molec-
ularly thin layers except in exceptional circumstance in which there is a surfaces
enhancement in scattering intensity on rough coinage metals in a technique known a
surface-enhanced Raman spectroscopy (SERS). In principle, such techniques could
be applied to examining tribo-film growth in ultrahigh vacuum [17].
Infrared spectroscopy is one of the most commonly used technique for examining
the surface chemistry of monolayer adsorbates primarily on metal surfaces. Since
metals substrates are conductive, electric fields that are oriented parallel to the surface
are screened, while electric field with components normal to the surface are enhanced
[13, 18]. As a consequence, reflection-absorption infrared spectroscopy (RAIRS)
is carried out on metal surfaces at grazing incident to enhance the component of
the electric field normal to the surface and is sufficiently sensitive to detect sub-
monolayer coverages of adsorbates on metals. This leads to the so-called surface
selection rules in which only vibrations that have atomic motions with a component
perpendicular to the surface are detected. This yields information on the orientation
of the adsorbate on the surface.
Perhaps most importantly, infrared spectroscopy can provide detailed information
on the chemical nature of adsorbates through characteristic “group frequencies”
[19, 20]. Here, a functional group of an adsorbate is found to generally exhibit a
characteristic range of frequencies that can be used to identify the nature of the
adsorbate and is particularly useful for understanding surface chemistry. However,
136 W. T. Tysoe

Fig. 5.2 Schematic depiction of the ATR-IR experiment

while potentially powerful for following surface tribochemical reactions, RAIRS

cannot provide a truly in-situ interrogation of the sliding interface.
An alternative approach which does overcome this limitation is attenuated total
internal reflection infrared spectroscopy (ATR-IR). A typical experimental geometry
of the implementation of an ATR-IR experiment is illustrated in Fig. 5.2. This tech-
nique uses an ATR crystal which is fabricated from an infrared-transparent material
with a high refractive index, typically germanium, or zinc selenide. The ends of the
crystal have bevelled edges cut at an angle such that the infrared light impinges on
the long faces of the crystal below the critical angle ΘC defined as ΘC  nn 21 , where
n 2 is the refractive index of the ATR crystal and n 1 is the refractive index of the
medium outside the crystal, typically vacuum or air so that n 1 ∼ 1. In this case,
the infrared radiation repeatedly reflects off the inner faces of the crystal and exits
from the opposite bevelled face, which is cut at the same angle as the opposite face.
However, the infrared light does not terminate abruptly when reflecting from the
inner faces of the crystal but can penetrate as an evanescent (decaying) wave some
distance into the low-refractive-index region of the sample. The decay length for this
evanescent wave is of the order of the wavelength of the light passing through the
crystal. Thus, the spectrum of any material placed on the face of the ATR crystal that
absorbs infrared radiation can be recorded.
However, since the infrared signal passes through the whole of the crystal, using a
ball-on-flay geometry (where the ATR crystal comprises the flat) will result in only a
relatively small signal so that the technique is generally used with a cylinder sliding
on the ATR crystal to provide a wider detection area and greater sensitivity [21, 22]. A
major advantage of this approach is that the ATR crystal can be coated with a thin layer
of metal which, if sufficiently thin, still allows the infrared radiation to penetrate it,
thus allowing tribochemical reactions to be studied in-situ for a metal-metal interface.
In this configuration, the method has a transparent interface while maintaining the
ability to study tribochemically realistic interfaces. While this approach has not yet
been used to study tribochemical reactions in vacuum, there are no technical reasons
that this should not be done.
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 137

5.2.3 Pseudo In-Situ Analyses

This section describes the use of what is termed “pseudo-in-situ” techniques which
include surface-sensitive techniques that, while they are not able to directly probe the
contacting interface while rubbing, they can be used to analyze the surface immedi-
ately after rubbing, as a function of the number of times that it is rubbed. This allows
kinetic information on the nature of the surface to be obtained. In this case, it is
helpful to be able to correlate the results of true in-situ analytical techniques with the
pseudo-in-situ measurements. Strategies for accomplishing this will be discussed in
Sect. 5.4.
Analytical techniques that are the most surface sensitive and those based on the
use of the electron since the mean-free path of electrons in most materials is of the
order of a few nanometers [23]. Such techniques include those that are excited by
high-energy photons and where electrons are detected to render the technique surface
sensitive, or those that are themselves excited by electrons. In general, to be useful
in understanding tribochemistry, they should have good spatial resolution. However,
photon-based techniques such as X-ray Photoelectron Spectroscopy (Chaps. 4 and
6) tend to require complex light sources that can be difficult to incorporate into the
same chamber as the tribometer itself and are generally carried out by transferring
the sample, in vacuo, from the chamber in which the surface is rubbed to an analysis
chamber. However, they do have the advantage of being able to provide quite detailed
analytical information, for example on the chemical state of the surface region.
Perhaps the simplest pseudo-in-situ analytical technique that is excited by elec-
trons is Auger spectroscopy. This is generally excited by electrons with a few keV in
energy, which can be generated using compact electron sources and can be focused
to spot sizes of less than a few microns.

5.2.3.1 Surface Analysis by Auger Spectroscopy

Materials are ionized by incident high-energy electrons. In particular, high-binding

energy electrons, so-called core states, can be removed if the incident electron kinetic
energy is greater than the ionization energy of the electron in that state. In general,
the ionization cross section, or the probability that an electron is removed, increases
as the electron energy increases above the threshold ionization energy, reaching its
maximum value when the incident electron kinetic energy reaches about 2–3 times
the ionization energy.
The ionized atom in the material is unstable and can relax by lower-binding-
energy electrons undergoing a transition onto the vacant energy level. The energy
gained by such a transition can be released either by emitting a photon or another
electron. Since the photons emitted during this process generally have energies in
the X-ray region of the spectrum, this process is known as X-ray fluorescence. The
emission of an electron during the decay process was first described by Pierre Auger
and this three-electron process forms the basis for Auger stectroscopy [24].
138 W. T. Tysoe

Fig. 5.3 Schematic depiction of the Auger process

The energy of the emitted Auger electron therefore depends on the energies of
three electron states in the atom; the energy of the state which was initially ionized,
the state from which the decaying electron originated and finally, the state from
which the Auger electron was emitted (Fig. 5.3). Since the energies of the electrons
in a particular atom are characteristic of that particular atom, the pattern of kinetic
energies of the electrons emitted by a particular atom are characteristic of that atom so
that the Auger spectrum provides detailed information on the chemical composition
of the sample.
In order to collect an Auger spectrum, the sample is illuminated with high-energy
electrons (with kinetic energies of a few keV), and the energies of the emitted elec-
trons are analysed using an electron energy analyser. They are invariably electrostatic
analysers that are essentially parallel plates across which a voltage is applied to create
an electric field, which deflects the electrons in an arc of radius r. The magnitude
electric field E across parallel plates separated by a distance d is given by E  Vd ,
so that the force F exerted on an electron with charge e is F  eV d
. This causes the
2
electron to move in an arc of radius r such that the centripetal force ( mvr ) balances
2
the electrostatic force. Since the kinetic energy is given by K  mv2 , this leads to a
relationship between the kinetic energy and voltage across the plates:
er
K  V. (5.8)
2d
In practice, the parallel plates are bent to have some fixed radius so that r and
d are fixed and depend on the geometry of the electron energy analyser. There are
several designs of electrostatic analysers, but they all operate on the same physical
principle, and an example of the schematic of a typical electron energy analyser is
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 139

Fig. 5.4 Schematic depiction of a hemispherical electron energy electron analyzer

shown in Fig. 5.4. Here the plates have been bent into the form of two concentric
hemispheres and is known as the hemispherical analyzer. Other configurations such
as the cylindrical mirror analyzer and the 127° sector analyzer are used [24]. Similar
types of analyzer are used for X-ray photoelectron spectroscopy. In order to provide
good energy resolution, narrow slits are placed at the entrance and exit of the analyzer
plates. The system shown in Fig. 5.4 includes an electron lens at the entrance of the
analyzer to optimize the collection of electrons and an electron detector located after
the exit slit. Based on (5.8), the spectrum is recorded by monitoring the number of
electrons passing through the analyzer as a function of the voltage across the analyser
plates, and this technique is used in Auger spectroscopy. It should be noted that the
resolution of electrostatic analyzers also depends on the electron kinetic energy.
While this is not an issue for measuring Auger spectra since the lines are generally
quite broad and the measured kinetic energies relatively low, it is important in XPS
where a higher resolution is required and electron analyzer is therefore configured
slightly differently for XPS (see Chap. 4).
An complication with Auger spectroscopy is that, since electrons interact strongly
with solids (and thus have a short mean-free path), high-energy electrons create an
intense background of inelastically scatted electrons so that the peaks in the Auger
spectrum are generally superimposed on a large background intensity. This is illus-
trated in Fig. 5.5, where the top curve shows the relatively small Auger peaks super-
imposed on an intense background due to the emission of inelastic electrons. Since
the inelastic background intensity varies slowly with kinetic energy, while the fea-
tures due to Auger emission are narrower, the first-derivative spectrum (lower curve),
shows well-defined peaks with both positive and negative excursions. Note also that
140 W. T. Tysoe

Fig. 5.5 Auger spectra of an

oxidized Fe–Cr–Nb alloy.
The top spectrum shows the
number of counts as a
function of electron kinetic
energy showing the large
inelastic electron
background. The lower curve
shows the differentiated
spectrum. Reproduced with
permission from [25]

Fig. 5.6 Carbon KLL Auger

spectra of various
carbon-containing materials
illustrating the effect of
chemical environment on
Auger lineshape.
Reproduced with permission
from [26]

the measured kinetic energies are characteristic of the elements in the sample; in this
case, distinct features due to Nb, Cr and Fe are clearly evident, while the large oxygen
peak indicates that the alloy is oxidized and also contains some carbon contamination.
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 141

The spectrum in Fig. 5.5 illustrates the utility of Auger spectroscopy for analyzing
the elements in a sample and, because it relies on electrons, it is sensitive only to
the outermost few nanometers of the sample. However, unlike XPS, it is somewhat
insensitive to the chemical nature of the species near the surface. The exception
is cases in which the Auger process involves the valence electrons of the element
since these are strongly modified by changes in the chemical environment [26].
An example is shown in Fig. 5.6 for various carbon-containing species including
carbides, graphite and diamond. Since the carbon Auger transition involves an initial
ionization of the carbon 1 s electron (with a binding energy of ~280 eV), followed by a
decay from the valence orbitals of the carbon, and the emission of an Auger electron,
also from the carbon valence orbitals, significant changes occur in the lineshapes.
The composition of the surface region can also be estimated by measuring the
peak-to-peak amplitudes of the features for each element in the first-derivative spec-
trum. The atomic composition of the sample can then be estimated using tabulated
values of sensitivity factors [27].

Application of Auger Spectroscopy to Tribological Processes; Shear-Induced

Loss of Sulfur

This section discusses the use of Auger spectroscopic surface analyses for study-
ing tribochemical kinetics. The experiments are carried out in an ultrahigh vacuum
(UHV) chamber, operating at a base pressure of ~1 × 10−10 Torr, where the sample
can be rubbed using a tungsten carbide ball that slides on the surface with various
sliding speeds and loads [8]. Sliding creates a wear track on the surface that is approx-
imately 100 µm wide, so that the analysis spot size must be less than this to ensure
that only the composition inside the wear track is analyzed. In order to carry out
the experiment, a clean copper sample is rubbed ~70 times to create an initial wear
track. The surface is then exposed to a model gas-phase lubricant, dimethyl disulfide
(DMDS, CH3 –S–S–CH3 ), which reacts by S–S bond scission to deposit methyl thi-
olate species (CH3 –S) on the surface. An initial Auger spectrum of the wear track is
collected prior to rubbing and the peak-to-peak amplitude of the sulfur KLL Auger
signal at ~160 eV kinetic energy is normalized to the peak-to-peak amplitude of the
Cu LMM feature at ~920 eV. Subsequent measurements are collected in this way to
eliminate the effect of variations in experimental conditions such as electron beam
current.
It is a challenge in any tribological experiment to ensure that only the worn region
of the sample is analyzed. In the case of the Auger spectra collected here, the spatial
resolution was measured using a 100 µm diameter silver wire attached to the surface
to mimic a wear track and Auger spectra collected to ensure that only silver and no
copper is detected. Note that smaller electron beam spot sizes generally yield lower
electron fluxes and thus less intense spectral signals so the experiment involves a
compromise between surface sensitivity and spatial resolution.
In order to carry out Auger analyses sequentially on the same rubbed region of the
sample, the electron gun was oriented such that the electron beam was incident on the
142 W. T. Tysoe

Fig. 5.7 Loss of sulfur from

the surface as a function of 1.0
the number of rubbing cycles
Dimethyl disulfide/Cu foil
following the adsorption of

Relative S/Cu Auger Ratio

dimethyl disulfide on copper.
0.8 Load; 0.44 N
Reproduce with permission Sliding Speed; 4 mm/s
from [12] Temperature; 298 K
0.6

0.4

0.2

0.0

0 20 40 60 80
Number of Scans

wear track to allow the sample to move repeatably between the rubbing experiments
and the surface analyses. The electron gun used for these experiments also included
deflection plates that allowed the electron beam to be rastered across the wear track.
Scattered electrons were detected using a channeltron located inside the vacuum
chamber which allowed scanning electron microscope images to be collected and
the position of the wear track to be identified.
After obtaining the initial sulfur signal, the surface was rubbed at a normal load
of 0.44 N using a sliding speed of 4 mm/s with the sample at room temperature.
The Auger spectrum was then collected by rotating the sample towards a cylindrical
mirror electrostatic electron energy analyzer (CMA) to measure the peak-to-peak
ratio of the sulfur Auger signal ratioed to that of the copper substrate. The sample
was then rotated back to its original position and rubbed once again to yield a plot of
S/Cu Auger signal as a function of the number of passes of the ball over the surface
(Fig. 5.7). This shows a rapid decrease in the amount of surface sulfur, so that, after
~80 passes over the surface, no sulfur is detected.
Because the experiment is carried out in UHV, the amount of adsorption from
the gas phase when the experiment is being carried out is negligible so that the data
shown in Fig. 5.7 are due to the shear-induced effects that result in the removal of
sulfur from the surface. Since the data yield a time sequence of the surface coverage,
the results can be analyzed to yield kinetic parameters. Accomplishing this requires
a kinetic model to be developed, which, in turn, relies on being able to identify the
elementary steps occurring during the tribochemical reaction. This will be discussed
in greater detail in Sect. 5.3.4.
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 143

5.2.3.2 Surface Analysis by Low Energy Electron Diffraction

Low-energy electron diffraction (LEED) is a surface-sensitive electron diffraction

technique that provides information on the crystal structure and order of surfaces.
It is based on de Broglie’s proposal that particles can have wave-like properties
[28], where the wavelike properties (wavelength, λ) are related to particle properties
(momentum, p) through: λ  hp , where h is Planck’s constant. This indicated that
electrons should diffract from crystalline surfaces in a manner analogous to X-rays
and this was confirmed by the first LEED experiments, although they did not call it
that, by Davisson and Germer [29]. Because of the short mean-free path of electrons
in solids, the diffraction pattern is sensitive to the outermost layers of the sample and
this was one of the first techniques applied to the study of surfaces. The principle of
the LEED experiment is relatively straightforward and consists of a monoenergetic
beam of electrons with energies ~100 eV, to give de Broglie wavelengths of the order
of lattice spacing in metals, impinging on a metal single crystal and observing the
resulting diffractions pattern. However, as noted above, high-energy electrons pro-
duce large number of inelastically scattered electrons when interacting with surfaces
and must be filtered out to allow the elastic LEED electrons to be observed. By
definition, the energies of the inelastic electrons are lower than the incident beam
energy and so can be filtered by a retarding grid. A typical LEED apparatus is shown
in Fig. 5.8 and consists of concentric, electron-transparent grids and a fluorescent
screen to detect the diffracted electrons. The sample is located at the center of cur-
vature of the grids so that electrons diffracted from the sample impinge normal to
the grids and a fluorescent screen to image the diffracted electrons. An electron gun
protrudes from the center of the LEED screen and is incident onto the single crystal
sample. From the de Broglie relation above, the electron wavelength can be varied
merely by changing the electron kinetic energy E p as λ  √ h where m is the
2mE p
electron mass. In order to remove any inelastic electron, a retarding voltage of +Er
is applied to one of the grids so that Er < E p to allow only elastically scattered
electrons to pass through the grid. Grounded grids are placed between the retarding
grids and the sample, and the sample and the screen to ensure that there are field-free
regions between them. The LEED pattern is imaged using a fluorescent screen using
a high applied potential E s to accelerate the electrons to the screen.
The beam is diffracted through an angle Θ given by the Bragg equation:

nλ  2d sin Θ, (5.9)

where d is the lattice spacing and n is the diffraction order and can be used to
precisely measure surface lattice spacings. Since the beam energy, and hence the
electron wavelength, can be easily varied, the diffraction pattern can be collected at a
variety of electron wavelengths and the resulting intensity versus beam voltage data
can be used to obtain detailed structural information [30].
144 W. T. Tysoe

Fig. 5.8 Schematic diagram

of a LEED apparatus. An
example of a LEED pattern
is also shown

Application of Low Energy Electron Diffraction to Material Transfer

in a Contact

Buckley used LEED experiments to explore the transfer of material from the (111)
surface of gold to the (111) face of copper-aluminum alloys as a function of alloy
composition [31]. The experiments were carried out in a UHV chamber using various
aluminium-copper alloys by bringing a clean gold sample into contact with it. The
applied load was controlled by external magnets through an arm that contained the
sample at the end. This arrangement allowed the adhesion force between gold and the
copper-aluminum alloys to be measured after contacting them. Figure 5.9 display the
adhesive force after contacting the surfaces for 10 s with a 20 mg load. The surfaces
of the alloys were examined by LEED after contacting with gold where, for example,
epitaxial gold films were found to be transferred to the copper-aluminum alloys with
less than 0.5% aluminum.

5.3 Elementary Steps in Tribochemical Reactions

As indicated in the Introduction, fully understanding tribochemical reactions, their

rates and the processes that control them require an understanding of the elementary
steps in the reaction that lead to the formation of an antiwear and/or friction-reducing
film, the so-called reaction pathway. A general outline of the processes that lead to the
formation of a boundary film are illustrated in Fig. 5.10. The first step is necessarily
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 145

Fig. 5.9 Adhesive force of (111) gold to a (111) surface of aluminum and various aluminum-copper
alloys. Initial applied load, 20 mg, Contact time 10 s. Reproduced with permission from [31]

the adsorption of the lubricant additive or gas-phase lubricant onto the surface. This
chemistry will depend both on the nature of the surface and additive as well as the
experimental conditions such as the temperature and concentration.
These initial thermal reactions may then be followed by sliding effects that may
induce further mechano- or tribochemical reactions. These may be deleterious, as
in the case of the decomposition of organic friction monolayers (comprising self-
assembled monolayers of adsorbates) or beneficial when it eventually leads to the
formation of a tribofilm. The final step is the formation of the tribofilm. This may,
in principle, either occur by the sliding-induced products growing a film on the
surface [32, 33] or by penetrating the subsurface region of the sample. The latter
process will create a new surface on which the lubricant can adsorb, thus start-

Fig. 5.10 Schematic Lubricant (AddiƟve)

diagram indicating possible adsorpƟon on the IniƟal thermal chemical
elementary steps in a surface. reacƟon on the surface.
tribochemical reaction to
form an antiwear or friction
reducing film

Sliding-induced adsorbate
Tribofilm formaƟon
decomposiƟon.
146 W. T. Tysoe

ing the second tribochemical reaction cycle. As this cycle proceeds, the nature of
the surface will evolve as the tribofilm forms so that both the thermal and sliding-
induced reactions and their rates may be different for each cycle, thus emphasizing
the potential complexity of fully understanding tribochemical reaction pathways. In
addition, many engineering surfaces themselves may be structurally and chemically
rather complex, being covered by contaminants and oxides from reaction with atmo-
spheric gases, thus emphasizing the importance of carrying out experiments in UHV
on well-characterized samples. Furthermore, the complexity of the reaction cycle
shown in Fig. 5.10 suggests that it is important to use simple model systems. For this
purpose, we use the gas-phase lubrication of copper by dimethyl disulfide (DMDS)
to illustrate the use of in-situ and “pseudo” in-situ approaches to understanding the
elementary steps in a tribochemical reaction cycle and their kinetics. While this is
not a commercial lubricant, it does contain the key aspects of commercial lubri-
cant additive since it contains a S–S linkage, typical of sulfur-containing lubricant
additives, as well as a hydrocarbon functionality. The discussion will address the
following key steps. It is first necessary to demonstrate that DMDS does act as a
gas-phase lubricant and then to identify the nature of the elementary steps in the
reaction cycle shown in Fig. 5.10. The first two steps in the cycle in Fig. 5.10 involve
investigations of the thermal surface chemistry and the nature and stability of the
initially adsorbed species arising from exposure to DMDS and are explored using
standard UHV surface science techniques. The final steps in the reaction occur at
the sliding interface and illustrate the use of the in-situ techniques described above.
Finally, once the nature of the elementary steps has been identified, this will allow
a simple kinetic model to be constructed to establish how the surface composition
evolved during rubbing.

5.3.1 Gas-Phase Lubrication of Copper by DMDS

The gas-phase lubrication of copper by DMDS is explored in UHV. As noted above,

DMDS is selected because it mimics commonly used lubricant additives and sulfur-
containing molecules are used as to lubricate iron [34–37]. The tribological exper-
iments were carried out using the following protocol. First, an initial wear track
was made on the clean copper substrate without DMDS dosing to obtain an initially
constant friction coefficient to ensure that any subsequent changes in friction were
due to DMDS dosing. As shown in Fig. 5.11, this resulted in a friction coefficient
decrease from an initial value of ~0.82 to ~0.53 after rubbing 40 times, and remained
essentially constant thereafter. The sample was then pressurised with 5 × 10−8 Torr
of DMDS to cause a rapid initial decrease in friction, followed by a slower change
for longer rubbing times, thereby showing that DMDS does act as a gas-phase lubri-
cant. Experiments were also carried out using a model lubricant that consisted of
5% DMDS dissolved in a poly α-olefin and found a friction coefficient of ~0.05
for sliding on copper. This work established that DMDS does act as a lubricant for
copper [38].
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 147

0.9
T = 300 K
0.8 Load 0.44 N
Speed 4 mm/s
DMDS
0.7

Friction coefficient
5 10-8 Torr

0.6

0.5

0.4

0.3

0.2

0 50 100 150 200 250

Number of Scans

Fig. 5.11 Plot of friction coefficient versus number of cycles measured in the ultrahigh vacuum
tribometer at a sliding speed of 4 × 10−3 m/s and a normal load of 0.44 N. The clean surface was
initially rubbed 70 times to reach a steady-state value of friction coefficient, and then dimethyl
disulfide was introduced via a dosing tube at a background pressure of ~5 × 10−8 Torr and the
friction coefficient then recorded in the presence of gas-phase DMDS. Reproduced with permission
from [39]

5.3.2 Thermal Chemistry of DMDS on Copper

The initial steps in the tribochemical reaction cycle involve the adsorption and ther-
mal reaction of DMDS on the clean copper surface (Fig. 5.10) in UHV. Such studies
are also required to explore the stability of the resulting surface species. Many of
the techniques that are used to study tribological interfaces such as XPS (see previ-
ous chapter), Auger spectroscopy (Sects. 5.2.3.1), LEED (Sect. 5.2.3.3) and infrared
spectroscopy (Sect. 5.2.3) are equally applicable to studying surface in UHV. An
additional technique, that is not applicable to in-situ studies of the tribological inter-
face, but is extremely useful for following surface chemical reaction pathways and
their kinetics in UHV is temperature-programmed desorption (TPD). This technique
involves adsorbing the molecule of interest onto a clean sample held at a sufficiently
low temperature (usually ~80 K, by cooling the sample with liquid nitrogen) to pre-
vent decomposition of the adsorbate. The sample is then placed close to and in-line-of
sight of a mass spectrometer and is heated linearly as a function of time and the des-
orbing products monitored as a function of the sample temperature. The detection
of gas-phase products is analogous to the detection of gas-phase products during
sliding described in Sect. 5.2.1.1. The resulting plot of mass spectrometer signal as
a function of time generally consists of peaks, where the onset temperature of the
peak depends on the activation energy of the surface reaction giving rise to the peak.
However, as the surface becomes depleted of reactants, the mass spectrometer signal
decreases once again to produce a peaked structure.
148 W. T. Tysoe

Fig. 5.12 TPD profile of

DMDS on Copper
DMDS adsorbed on copper.
Reproduced with permission 0.5 L
from [10] 1.25 L
159 K
426 K 2.5 L
3.75 L

16 amu (methane)

0 100 200 300 400 500 600 700 800

Temperature, K

The nature of the molecular species giving rise to a particular peak can be obtained
by measuring the peak intensity at various masses and comparing with mass spec-
trometer ionizer fragmentation patterns of various compounds. A typical series of
TPD profiles for DMDS adsorbed on copper at ~100 K are shown in Fig. 5.12 when
monitoring the 16 amu mass spectrometer signal [10, 11]. The data are collected for
various initial exposures of DMDS, which are indicates in Langmuirs (1 Langmuir
(L)  1 × 10−6 Torr s). This shows two distinct peaks, a sharp feature centered at
~159 K, and a broader feature at 426 K. By measuring the signals at other masses,
it can be shown that the low-temperature (~179 K) peak is due to the desorption of
condensed, molecular DMDS, while the higher-temperature (~426 K) is due to the
reactive formation of methane due to the decomposition of adsorbed methyl thiolate
(CH3 –S(ads)) species [10].
The maximum temperature of a desorption peak, T p can be used to estimate the
reaction activation energy assuming that the sample temperature varies linearly with
time t:

T  T0 + βt, (5.10)

where T0 is the initial temperature and β is the heating rate. For a first order reaction
(5.1), and taking the amplitude of signal at a particular temperature in the TPD profile
(Fig. 5.12) to be proportional to the reaction rate, r, gives:
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 149

dΘ E act
−  r  A exp Θ, (5.11)
dt kB T

where E act is the activation energy for the process that gives rise to the desorption
product, T is the absolute temperature and k B the Boltzmann constant. The peak
occurs when the desorption rate is the maximum. Putting dr dt
 0 gives, for a first-
order process, an equation for the activation energy in terms of T p [9]:
 
E act A E act
 exp . (5.12)
kB T p2 β kB T p

This reveals that, for a first-order process, the value of T p is independent of the
adsorbate coverage, indicating that the methane formed in the TPD profile is due to a
first-order process (methyl thiolate decomposition). In order to simplify this equation,
it is found that a plot of E act versus T p is approximately linear for 1013 > βA > 108
and can be approximated by:

E act ATp
 ln − 3.64. (5.13)
kB T p β

Applying this “Redhead” equation [9] to the methane desorption peak in Fig. 5.12,
and using a typical pre-exponential factor A for a first-order process of 1013 s−1 , gives
an activation energy of ~100 kJ/mol. The reaction is confirmed using XPS, which
shows the disappearance of carbon over a temperature range which methane desorbs
and the persistence of sulfur on the surface. These results confirm that methyl thiolate
species are stable on the surface at ~300 K, the temperature at which the tribological
measurements are carried out.

5.3.3 Identifying Elementary Steps During Sliding

The subsequent sliding-induced elementary reaction steps are explored using a com-
bination of the in-situ and pseudo in-situ techniques described above. It should be
emphasized that the temperature rise for the relatively mild conditions used in the
UHV tribometer is negligible [8]. The products formed during sliding were discussed
in Sect. 5.2.1.1 and Fig. 5.1 shows the desorption of methane pulses as the pin slides
over a methyl thiolate-covered surface. The rate of thiolate decomposition to form
methane (and C2 hydrocarbons) is accelerated by interfacial shear, since it does not
occur thermally at this temperature and there are now a number of reactions for which
this has been found including the decomposition of oxygen-containing adsorbates
on graphene [15] and the growth of anti-wear films from ZDDP [32, 33]. However,
150 W. T. Tysoe

unlike the thermal chemistry on copper, where adsorbed sulfur remains on the sur-
face, Auger analyses of the wear track reveal that the sulfur is removed during sliding
(Fig. 5.7, Sect. 5.2.2.2).
Because of the strong binding of sulfur to copper, sulfur is unlikely to desorb
and no sulfur was detected in the gas phase, implying that shear causes sulfur to
penetrate the subsurface region of the copper sample. This idea is tested by taking
advantage of the observation made above that sulfur is thermodynamically stable
on the surface of copper rather than in the bulk. That is, any subsurface sulfur is
metastable so that heating the sample should cause sulfur to diffuse to the surface
once again. However, the DMDS-dosed copped sample is covered by methyl thiolate
species which must be removed for this experiment to be carried out. The results of
this experiment are displayed in Fig. 5.13, which display the profiles of the sulfur
coverage across a wear track measured using small-spot-size Auger spectroscopy.
The center of the wear track is located at 300 µm. The baselines of the scans shown
in Fig. 5.13a have as S/Cu Auger ratio of ~0.42, corresponding to the sulfur in a
saturated monolayer of methyl thiolate species, while in Fig. 5.13b, the baseline has
a S/Cu ratio of zero. Profile I shows a scan across a previously formed wear track for
a copper surface that has been saturated with methyl thiolate species. It seems that
slightly more methyl thiolate species can adsorb in the wear track, possibly because
it is somewhat rougher than the surface outside the wear track. Profile II shows the
effect of rubbing the surface 50 times, confirming that sulfur has been removed in
the rubbed region as found in the data plotted in Fig. 5.7. This sulfur is proposed to
have penetrated the bulk of the sample. In order to test this, the surface sulfur was
carefully removed by Argon ion bombardment and the resulting profile of the sulfur
Auger signal is shown in scan III; all of the surface sulfur has been removed. Profile
IV shows the result of heating to ~780 K where now sulfur is detected on the surface.
This is proposed to have diffused from the subsurface region and is wider than the
original wear track presumably due to lateral diffusion of sulfur on the copper surface.
This effect provides a shear-induced pathway for the low-temperature formation of
a tribofilm and formally results in an oxidation of the copper by sulfur.
A possible explanation for this shear-induced surface-to-bulk transport come from
molecular dynamics simulations [40–43] which suggest that shear can induced mix-
ing of the surface regions of soft metals during sliding. Analytical models of this
effect [44] show that the surface-to-bulk transport depends on the strain-rate sensi-
tivity of the material. For copper, which has a low value of the strain-rate sensitivity
[45–52], this theory predicts that the distance that the sulfur on the surface penetrates
the bulk is proportional to the number of times that the surface has been rubbed and
this dependence has recently been verified for subsurface sulfur in copper [53]. The
elementary steps for the tribochemical reaction of DMDS with copper are summa-
rized in Fig. 5.14.
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 151

Fig. 5.13 Plots of sulfur peak-to-peak LMM Auger intensity (152 eV kinetic energy) ratioed to
the peak-to-peak intensity of the Cu LMM Auger transition (920 eV kinetic energy) as a function
of position through the wear track, which is centered at ~300 µm. a Shows the effect of rubbing,
where profile I plots the sulfur coverage after the sample was dosed with DMDS after rubbing the
clean surface, and profile II shows the result of subsequently rubbing the surface 50 times under a
normal load of 0.44 N at a sliding speed of 4 × 10−3 m/s. b Shows that all of the sulfur is removed
from the surface by argon ion bombardment (profile III), but that subsequently heating the sample
to ~780 K causes sulfur to segregate to the surface only in the region of the wear track (profile IV).
Reproduced with permission from [12]

Fig. 5.14 A schematic of the elementary steps in the tribochemistry of DMDS on copper
152 W. T. Tysoe

5.3.4 Microkinetic Modelling of Tribofilm Formation

for DMDS on Copper

Having defined the elementary steps in the tribochemical reaction pathway sets the
stage for developing microkinetic models to describe the film formation reaction
rates based on the reaction pathway (Fig. 5.14), which consists of a sequence of two
shear-induced reactions: RS (ads) → S (ads) → S (subsurface) , where the rate constant for
the first reaction is k 1 and for the second is denoted as k 2 .
The data in Fig. 5.1 for the formation of gas-phase hydrocarbons indicated that
methyl thiolate decomposes via a first-order reaction (Sect. 5.2.1):
dΘth
−  k1 Θth , (5.14)
dt

where Θth is the thiolate coverage. Writing k1  k1 tC and integrating (5.14) gives:
 
Θth  exp −k1 p , (5.15)

where p is the number of passes, where the method for obtaining k1 is discussed in
Sect. 5.2.1.
Denoting the coverage of adsorbed sulfur, S (ads) formed by this reaction as Θ S
then, from the sequential reaction shown in Fig. 5.14, adsorbed sulfur is formed by
alkyl thiolate decomposition and removed by surface-to-bulk transport:
dΘ S
 k1 Θth − k2 Θ S , (5.16)
dt
and assuming that surface-to-bulk transport occurs via a first-order process as will
be discussed in greater detail below, substituting for the alkyl thiolate coverage Θth
from (5.15) and solving gives:

k1   −k1 p −k2 p

Θ S ( p)  e − e , (5.17)
k2  − k1 

where k2  k2 tC . The rate constant k2 is that for the transport of adsorbed sulfur
into the subsurface of copper where the film penetrates a distance d( p) into the
subsurface region as a function of the number of scans as: d( p)  d p p x , where d p is
the distance that the sulfur penetrates the subsurface region per pass [44] and it has
been shown the x ~ 1 [53]. If the initial film thickness at t  0 is d 0 (corresponding
to the thickness of the original saturated sulfur overlayer), then it can be shown that
k2  tCp [54].
d

Since the Auger analyses of the surface during rubbing (Fig. 5.7) measure the
sulfur from both adsorbed sulfur and alkyl thiolate species, the coverage of all sulfur-
containing species is calculated as Θ S (tot)  Θ S + Θth , and this gives:
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 153

k2 −k1 p k 
Θ S (tot)   e −  1  e−k2 p . (5.18)
k2 − k1 k2 − k1

Finally, the coverage of the clean surface, Θclean is calculated from:

Θclean  1 − Θ S − Θth . (5.19)

However, surface spectroscopies, including Auger spectroscopy, are sensitive not

only to adsorbates but also include signals from the sub-surface region [24, 55]. It
has been previously shown that the Auger signal due to subsurface sulfur varies as
1/ p [53] so that the Auger signal contribution arising from subsurface sulfur is taken
to be ∝ Θsubsurface
p+1
where Θsubsurface is the subsurface sulfur coverage and p + 1 is used
to avoid a singularity at p  0. Therefore the sulfur Auger signal IAuger ( p) is given
by:
αΘsubsurface
IAuger ( p) ∝ IAuger (S) + IAuger (th) + , (5.20)
p+1

where α is a constant. Since Θsubsurface  1 − Θth − Θ S , this yields:

  
k2 −k1 p k1 −k2 p α α
IAuger ( p)  e − e 1 − + , (5.21)
k2 − k1 k2 − k1 p+1 p+1

which is normalized to the initial Auger signal of the clean surface.

The above kinetic equations are used to analyze the shear-induced elementary
reaction steps of alkyl thiolates on copper (Fig. 5.14). The rate of decomposition of
methyl thiolate species, analyzed as described in Sect. 5.2.1.1 gives k1  0.63 ± 0.03.
In order to obtain k2 , the sulfur Auger data (Fig. 5.7) are fit to (5.21) to yield values
of α  0.9 ± 0.1 and k2  2.5 ± 0.5. The line through the experimental data in
Fig. 5.14 is a fit to this equation.
Since these kinetic parameters derive from a pseudo in-situ technique, Auger
spectroscopy, the results are verified using an in-situ method measuring the friction
coefficient as a function of number of rubbing cycles (Fig. 5.15). Adapting (5.4)
(Sect. 5.2.1.2) to the reaction sequence shown in Fig. 5.14 gives an equation for the
friction coefficient μ( p) as a function of the number of passes p:

μ( p)  μth Θth ( p) + μS Θ S ( p) + μclean (1 − Θth ( p) − Θ S ( p)), (5.22)

where μth , μS and μclean are the characteristic friction coefficients of the thiolate-
covered, sulfur-covered and clean surfaces, respectively. Here, the rate constants
were constrained to be in the range determined above and μclean was allowed to vary
between 0.45 and 0.55 found for sliding on clean copper. The fit to the data is shown
as a solid line in Fig. 5.15 and yields μth = 0.07 ± 0.02, μS  0.39 ± 0.06, and
μclean  0.47 ± 0.05. The agreement between experiment and theory suggests that
the value of k2 measured from the Auger data is reasonable. The resulting plot of
154 W. T. Tysoe

Fig. 5.15 Plot of the friction

coefficient as a function of
the number of scans while
rubbing a
methyl-thiolate-covered
copper surface at a normal
load of 0.44 N and a sliding
speed of 4 mm/s. The line
shows a fit to the data.
Reproduced with permission
from [54]

Fig. 5.16 Plot of the

calculated coverages of
adsorbed methyl thiolate
species (), adsorbed sulfur
(●) and the clean copper
surface (▲) as a function of
the number of scans over a
methyl-thiolate covered
surface with a normal load of
0.44 N and a sliding speed of
4 mm/s. Reproduced with
permission from [54]

the adsorbate coverages as a function of the number of scans for an initially methyl
thiolate-saturated surface, at a load of 0.44 N and a sliding speed of 4 mm/s, is shown
in Fig. 5.16. This shows that a clean surface is produced after ~8 scans, since the
clean surface coverage (▲) becomes almost unity after this number of scans, thereby
allowing the tribochemical cycle to be repeated to eventually form a sulfur-containing
boundary film. In addition, as a consequence of k2 being much larger than k1 , the
sulfur is transported into the subsurface region almost as quickly as it is formed on
the surface resulting in the adsorbed sulfur coverage always remaining less than 0.2
monolayers (●).
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 155

5.4 Conclusions

This Chapter describes a number of in-situ and pseudo in-situ techniques that can
be used to explore the elementary steps in tribochemical reactions to form friction-
reducing or antiwear films under UHV conditions. The application of some of these
techniques is illustrated using a simple model system of DMDS reacting with copper
to form a sulfur-containing tribofilm. The results of these experiments allow the ele-
mentary steps in the reaction pathway to be identified. In particular, two shear-induced
processes were found. The first consisted of an increase in the rate of decomposition
of methyl thiolate species into adsorbed sulfur and gas-phase hydrocarbons. Here,
the shear-induced reaction pathway was identical to that found for the thermal reac-
tion. This behaviour can be rationalized using model that was first applied to cell
adhesion [56] but has its roots in earlier models proposed by Prandl and Eyring [57]
and has also been shown to apply to adsorbate decomposition measured in the AFM
[8]. This approach of in-situ measurements of tribochemical reaction pathways on
model systems will lead to a more detailed fundamental understanding of the way in
which interfacial shear accelerates the rates of surface chemical reactions and lead
to the formation of tribofilms.
The second shear-induced step in the reaction pathway consisted of the surface-to-
bulk transport of sulfur to initiate the formation of a tribofilm. As shown in Fig. 5.16,
this results in the regeneration of a clean surface to allow methyl thiolate species to
once again adsorb onto copper to provide a tribochemical reaction cycle (Fig. 5.10).
However, alternative film formation processes have also been observed with ZDDP,
where the growing film forms on top of the substrate [32, 33].
Finally, while this Chapter has focussed on the film formation chemistry because
of its relevance to tribology, the reaction shown in Fig. 5.14 could equally as well be
viewed as a tribocatalytic reaction in which a combination of surface reactivity and
interfacial shear combine to accelerate the rate of DMDS decomposition into small
hydrocarbons and form a copper sulfide film.

References

1. T.J. Blunt, P.V. Kotvis, W.T. Tysoe, Determination of interfacial temperatures under extreme
pressure conditions. Tribol. Lett. 2(3), 221–230 (1996)
2. K. Holmberg, P. Andersson, A. Erdemir, Global energy consumption due to friction in passenger
cars. Tribol. Int. 47, 221–234
3. K.J. Laidler, Chemical kinetics (McGraw-Hill, New York, 1965)
4. O.A. Mazyar, H. Xie, W.L. Hase, Nonequilibrium energy dissipation at the interface of sliding
model hydroxylated α-alumina surfaces. J Chem. Phys. 122(9), 094713 (2005)
5. W.G. Sawyer, K.J. Wahl, Accessing inaccessible interfaces: in situ approaches to materials
tribology. MRS Bulletin 33(12), 1145–1150 (2008)
6. S. Mori, W. Morales, Tribological reactions of perfluoroalkyl polyether oils with stainless steel
under ultrahigh vacuum conditions at room temperature. Wear 132(1), 111–121 (1989). https://
doi.org/10.1016/0043-1648(89)90206-8
156 W. T. Tysoe

7. I. Minami, T. Kubo, S. Fujiwara, Y. Ogasawara, H. Nanao, S. Mori, Investigation of tribo-

chemistry by means of stable isotopic tracers: TOF-SIMS analysis of langmuir-Blodgett films
and examination of their tribological properties. Tribol. Lett. 20(3–4), 287–297 (2005). https://
doi.org/10.1007/s11249-005-9068-2
8. H.L. Adams, M.T. Garvey, U.S. Ramasamy, Z. Ye, A. Martini, W.T. Tysoe, Shear-induced
mechanochemistry: Pushing molecules around. J. Phys. Chem. C 119(13), 7115–7123 (2015).
https://doi.org/10.1021/jp5121146
9. P.A. Redhead, Thermal desorption of gases. Vacuum 12, 9 (1962)
10. O.J. Furlong, B.P. Miller, Z. Li, J. Walker, L. Burkholder, W.T. Tysoe, The surface chemistry
of dimethyl disulfide on copper. Langmuir 26(21), 16375–16380 (2010). https://doi.org/10.10
21/la101769y
11. O. Furlong, B. Miller, Z. Li, W.T. Tysoe, The surface chemistry of diethyl disulfide on copper.
Surf. Sci. 605(5–6), 606–611 (2011). https://doi.org/10.1016/j.susc.2010.12.026
12. O. Furlong, B. Miller, W. Tysoe, Shear-induced surface-to-bulk transport at room temperature
in a sliding metal-metal interface. Tribol. Lett. 41(1), 257–261 (2011). https://doi.org/10.100
7/s11249-010-9711-4
13. R.G. Greenler, Reflection method for obtaining the infrared spectrum of a thin layer on a metal
surface. J. Chem. Phys. 50(5), 1963–1968 (1969). https://doi.org/10.1063/1.1671315
14. G. Wu, F. Gao, M. Kaltchev, J. Gutow, J.K. Mowlem, W.C. Schramm, P.V. Kotvis, W.T. Tysoe,
An investigation of the tribological properties of thin KCl films on iron in ultrahigh vacuum:
modeling the extreme-pressure lubricating interface. Wear 252(7–8), 595–606 (2002)
15. J.R. Felts, A.J. Oyer, S.C. Hernández, K.E. Whitener Jr., J.T. Robinson, S.G. Walton, P.E. Shee-
han, Direct mechanochemical cleavage of functional groups from graphene. Nat. Commun. 6,
6467 (2015). https://doi.org/10.1038/ncomms7467
16. A. Cameron, R. Gohar, Theoretical and experimental studies of the oil film in lubricated point
contact. Proc. Royal Soc. A Mathematical, Phys. Eng. Sci. 291(1427), 520–536 (1966)
17. P.L. Stiles, J.A. Dieringer, N.C. Shah, R.P. Van Duyne, Surface-enhanced Raman spectroscopy.
Annu. Rev. Anal. Chem. 1(1), 601–626 (2008)
18. R.G. Greenler, Infrared study of adsorbed molecules on metal surfaces by reflection techniques.
J. Chem. Phys. 44(1), 310–315 (1966). https://doi.org/10.1063/1.1726462
19. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds (Wiley,
New York, 1978)
20. N.B. Colthup, L.H. Daly, S.E. Wiberley, Introduction to Infrared and Raman Spectroscopy,
2nd edn. (Academic Press, New York, 1975)
21. F.M. Piras, A. Rossi, N.D. Spencer, Growth of tribological films: a in situ characterization based
on attenuated total reflection infrared spectroscopy. Langmuir 18(17), 6606–6613 (2002)
22. F. Mangolini, A. Rossi, N.D. Spencer, Chemical reactivity of triphenyl phosphorothionate
(TPPT) with Iron: an ATR/FT-IR and XPS investigation. J. Phys. Chem. C 115(4), 1339–1354
(2010)
23. E.H. Sondheimer, The mean free path of electrons in metals. Adv. Phys. 50(6), 499–537 (2001).
https://doi.org/10.1080/00018730110102187
24. B. Briggs, M.P. Seah, Practical Surface Analysis: Auger and X-ray Photoelectron Spectroscopy
(Wiley, New York, 1996)
25. J.C. Vickerman, Surface Analysis: The Principal Techniques (Wiley, Chichester, 1997)
26. T.W. Haas, J.T. Grant, G.J. Dooley III, Chemical effects in auger electron spectroscopy. J. Appl.
Phys. 43(4), 1853–1860 (1972)
27. S. Mroczkowski, D. Lichtman, Calculated Auger yields and sensitivity factors for KLL–NOO
transitions with 1–10 kV primary beams. J. Vac. Sci. Technol. A: Vac. Surf. Films 3(4),
1860–1865 (1985)
28. L.D. Broglie, A tentative theory of light quanta. Philos. Mag. Ser. 6 47(278), 446–458 (1924).
https://doi.org/10.1080/14786442408634378
29. C. Davisson, L.H. Germer, Diffraction of electrons by a crystal of nickel. Phys. Rev. 30(6),
705–740 (1927)
5 In-Situ Measurement of Tribochemical Processes in Ultrahigh Vacuum 157

30. J.B. Pendry, Low Energy Electron Diffraction : The Theory and Its Application to Determination
of Surface Structure (Academic Press, London, 1974)
31. D.H. Buckley, A LEED Study of the Adhesion of Gold to Copper and Copper-Aluminum
Alloys. NASA Technical Report NASA-TN-D-5351 (1969)
32. N.N. Gosvami, J.A. Bares, F. Mangolini, A.R. Konicek, D.G. Yablon, R.W. Carpick, Mecha-
nisms of antiwear tribofilm growth revealed in situ by single-asperity sliding contacts. Science
(2015). https://doi.org/10.1126/science.1258788
33. J. Zhang, H. Spikes, On the Mechanism of ZDDP antiwear film formation. Tribol. Lett. 63(2),
1–15 (2016). https://doi.org/10.1007/s11249-016-0706-7
34. W. Davey, E.D. Edwards, The extreme-pressure lubricating properties of some sulphides and
disulphides, in mineral oil, as assessed by the four-ball machine. Wear 1(4), 291–304 (1958).
https://doi.org/10.1016/0043-1648(58)90002-4
35. E.S. Forbes, The load-carrying action of organo-sulphur compounds—A review. Wear 15(2),
87–96 (1970). https://doi.org/10.1016/0043-1648(70)90002-5
36. M. Kaltchev, P.V. Kotvis, T.J. Blunt, J. Lara, W.T. Tysoe, A molecular beam study of the
tribological chemistry of dialkyl disulfides. Tribol. Lett. 10(1), 45–50 (2001). https://doi.org/
10.1023/a:1009020725936
37. J. Lara, T. Blunt, P. Kotvis, A. Riga, W.T. Tysoe, Surface chemistry and extreme-pressure
lubricant properties of dimethyl disulfide. J. Phys. Chem. B 102(10), 1703–1709 (1998).
https://doi.org/10.1021/jp980238y
38. O.J. Furlong, B.P. Miller, P. Kotvis, W.T. Tysoe, Low-temperature, Shear-induced tribofilm
formation from dimethyl disulfide on copper. ACS Appl. Mater. Interfaces 3(3), 795–800
(2011)
39. O.J. Furlong, B.P. Miller, P. Kotvis, W.T. Tysoe, Low-temperature, shear-induced tribofilm
formation from dimethyl disulfide on copper. ACS Appl. Mater. Interfaces 3(3), 795–800
(2011). https://doi.org/10.1021/am101149p
40. D.A. Rigney, Transfer, mixing and associated chemical and mechanical processes during the
sliding of ductile materials. Wear 245(1–2), 1–9 (2000). https://doi.org/10.1016/s0043-1648(
00)00460-9
41. X.-Y. Fu, D.A. Rigney, M.L. Falk, Sliding and deformation of metallic glass: Experiments and
MD simulations. J. Non-Cryst. Solids 317(1–2), 206–214 (2003). https://doi.org/10.1016/s00
22-3093(02)01999-3
42. H.J. Kim, W.K. Kim, M.L. Falk, D.A. Rigney, MD simulations of microstructure evolution
during high-velocity sliding between crystalline materials. Tribol. Lett. 28(3), 299–306 (2007).
https://doi.org/10.1007/s11249-007-9273-2
43. A. Emge, S. Karthikeyan, H.J. Kim, D.A. Rigney, The effect of sliding velocity on the tribo-
logical behavior of copper. Wear 263, 614–618 (2007). https://doi.org/10.1016/j.wear.2007.0
1.095
44. S. Karthikeyan, H.J. Kim, D.A. Rigney, Velocity and strain-rate profiles in materials subjected
to unlubricated sliding. Phys. Rev. Lett. 95(10), 106001 (2005). https://doi.org/10.1103/physr
evlett.95.106001
45. A. Mishra, M. Martin, N.N. Thadhani, B.K. Kad, E.A. Kenik, M.A. Meyers, High-strain-rate
response of ultra-fine-grained copper. Acta Mater. 56(12), 2770–2783 (2008). https://doi.org/
10.1016/j.actamat.2008.02.023
46. M.A. Meyers, A. Mishra, D.J. Benson, Mechanical properties of nanocrystalline materials.
Prog. Mater. Sci. 51(4), 427–556 (2006). https://doi.org/10.1016/j.pmatsci.2005.08.003
47. A. Mishra, B.K. Kad, F. Gregori, M.A. Meyers, Microstructural evolution in copper subjected
to severe plastic deformation: Experiments and analysis. Acta Mater. 55(1), 13–28 (2007).
https://doi.org/10.1016/j.actamat.2006.07.008
48. T. Zhu, J. Li, A. Samanta, H.G. Kim, S. Suresh, Interfacial plasticity governs strain rate sensi-
tivity and ductility in nanostructured metals. Proc. Natl. Acad. Sci. 104(9), 3031–3036 (2007).
https://doi.org/10.1073/pnas.0611097104
49. R. Schwaiger, B. Moser, M. Dao, N. Chollacoop, S. Suresh, Some critical experiments on the
strain-rate sensitivity of nanocrystalline nickel. Acta Mater. 51(17), 5159–5172 (2003). https://
doi.org/10.1016/s1359-6454(03)00365-3
158 W. T. Tysoe

50. Y.F. Shen, L. Lu, M. Dao, S. Suresh, Strain rate sensitivity of Cu with nanoscale twins. Scripta
Mater. 55(4), 319–322 (2006). https://doi.org/10.1016/j.scriptamat.2006.04.046
51. H.W. Höppel, J. May, M. Göken, Enhanced strength and ductility in ultrafine-grained alu-
minium produced by accumulative roll bonding. Adv. Eng. Mater. 6(9), 781–784 (2004).
https://doi.org/10.1002/adem.200306582
52. G.T. Gray III, T.C. Lowe, C.M. Cady, R.Z. Valiev, I.V. Aleksandrov, Influence of strain rate
and temperature on the mechanical response of ultrafine-grained Cu, Ni, and Al–4Cu–0.5Zr.
Nanostruct. Mater. 9(1–8), 477–480 (1997). https://doi.org/10.1016/s0965-9773(97)00104-9
53. B. Miller, O. Furlong, W. Tysoe, The kinetics of shear-induced boundary film formation from
dimethyl disulfide on copper. Tribol. Lett. 49(1), 39–46 (2013). https://doi.org/10.1007/s1124
9-012-0040-7
54. H. Adams, B.P. Miller, P.V. Kotvis, O.J. Furlong, A. Martini, W.T. Tysoe, In situ measurements
of boundary film formation pathways and kinetics: Dimethyl and diethyl disulfide on copper.
Tribol. Lett. 62(1), 1–9 (2016). https://doi.org/10.1007/s11249-016-0664-0
55. P.J. Cumpson, Angle-resolved XPS and AES: Depth-resolution limits and a general comparison
of properties of depth-profile reconstruction methods. J. Electron Spectrosc. Relat. Phenom.
73(1), 25–52 (1995). https://doi.org/10.1016/0368-2048(94)02270-4
56. G. Bell, Models for the specific adhesion of cells to cells. Science 200(4342), 618–627 (1978).
https://doi.org/10.1126/science.347575
57. H. Spikes, W. Tysoe, On the commonality between theoretical models for fluid and solid
friction, wear and tribochemistry. Tribol. Lett. 59(1), 1–14 (2015). https://doi.org/10.1007/s1
1249-015-0544-z
Chapter 6
Tribochemistry and Morphology
of P-Based Antiwear Films

A. Dorgham, A. Neville and A. Morina

6.1 Introduction

Zincdialkyldithiophosphate (ZDDP) is one of the most widely used additives,

whether in oils or greases. ZDDP is multifunctional as it can work as antiwear,
antioxidant and anticorrosion additive [1]. It exhibits these beneficial functionalities
by its ability to decompose under rubbing, heating or possibly high pressure to form
a protective film. In the case of this film is formed during heating without rubbing it
is called a thermal antiwear film whereas in the case of rubbing, the film is called a
tribofilm [2]. The antiwear mechanism of this tribofilm originates from its capability
to separate the contacting surface [3–7], to digest the abrasive wear debris [8–10]
and to decompose peroxides and peroxy-radicals [11–13].
Despite the numerous benefits of using the antiwear ZDDP additive, it has disad-
vantages as well. Firstly, ZDDP increases micropitting of the steel surface and thus
decreases its bearing life [14–17]. Secondly, it can poison the catalyst in the cars
equipped with a catalytic converter and therefore degrades the converter’s capability
to reduce harmful emissions in the exhaust gas [18, 19]. This led to an ever expanding
strict rules on the allowed concentrations of ZDDP, phosphorus and zinc in the oil
[20, 21]. The trend of these rules suggests that the ZDDP should be replaced. One
of the available options is to use neutral or acidic ashless dialkyldithiophosphate
(DDP). The ashless additives can provide better [22, 23], comparable [24] or worse

A. Dorgham · A. Neville · A. Morina (B)

School of Mechanical Engineering, Institute of Functional Surfaces, University of Leeds, Leeds
LS2 9JT, UK
e-mail: a.morina@leeds.ac.uk
A. Dorgham
e-mail: A.Dorgham@leeds.ac.uk
A. Neville
e-mail: a.neville@leeds.ac.uk

© Springer Nature Switzerland AG 2018 159

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_6
160 A. Dorgham et al.

[25, 26] antiwear properties than the ZDDP, which seem to depend greatly on the
operating conditions, properties of contacting surfaces and chemistries of base oil
and additives.
Replacing the ZDDP completely is not an easy task due to the lack of complete
understanding of the complex pathways involved in its decomposition reactions and
the possible interactions with the other additives in the oil. This statement is also true
for the case of the DDP, which shares in general the same complexity. Nonetheless,
despite these difficulties, a myriad of experimental works were carried out over the
last 70 years in order to understand the ZDDP decomposition reactions. These works
have been discussed in different reviews [21, 27, 28]. The review of Barnes et al.
[27] discussed the role of the ZDDP in the oil and its functions as an antioxidant and
antiwear additive. In addition, the review covered the interactions between the ZDDP
and various other components that can be present in the oil. The ZDDP review of
Spikes [21] presented a historical overview of the additive starting from its inception
until its current use. The review focused on the reaction mechanisms of the ZDDP
in the oil and the composition of the formed protective tribofilm and its formation
and removal kinetics. It also examined the contribution of different experimental
techniques to our current understanding of the ZDDP tribochemistry. Finally, the
review of Nicholls et al. [28] focused on the decomposition reactions of the ZDDP
in the absence and presence of other components in the oil, the composition of the
formed tribofilms and their mechanical properties. This review is the last one focusing
on the ZDDP and was presented more than a decade ago.
It should be noted that the early works in the field of tribology focused on using a
single surface characterisation technique such as XPS, FTIR and EDX to study the
complex tribochemistry of the ZDDP and DDP tribofilms. However, every technique,
despite its numerous advantages, has certain limitations as summarised in Table 6.1,
which are mainly related to the experimental atmosphere; whether ambient (A),
high vacuum (HV) or ultra-high vacuum (UHV); detection sensitivity, chemical and
structural information capability, lateral resolution and sampling depth. Therefore,
as more and more of these techniques become available added to the insufficient
information gathered from a single technique only, the recent tribological studies
have started employing what is commonly known as the multi-technique approach.
In this approach, observations and evidence are collected from various experimental
techniques to help understand not only the composition but also the tribological and
mechanical properties of ZDDP and DDP tribofilms.
Despite the extensive research in the tribological and mechanical properties of the
ZDDP and DDP tribofilms and their compositions, the link between these properties
is still not clear. This review aims at discussing these properties and highlighting the
different links between them in order to provide a clearer picture on their tribochem-
istry in the base oil and the properties of the formed antiwear films. As the main
body of the literature is related to the ZDDP and due to the large similarities between
the ZDDP and DDP additives, the focus of this review will be mainly on the ZDDP
while making a direct comparison to the DDP whenever appropriate.
6 Tribochemistry and Morphology of P-Based Antiwear Films 161

Table 6.1 Comparison between the surface characterisation techniques

Method Atmospherea Sensitivity Chemical Structure Lateral Detected
information resolution depth
XPS UHV 1% Yes No 10 µm 3–8 nm
AES UHV 1% Yes No 20 nm 3–8 nm
XAS UHV-A 50 ppm Yes Yes 50 nm-mm 1 nm—bulk
EDX HV 0.1% No No 1 µm 1 µm
Raman A 0.1 at.% Yes Yes 1 µm 0.2–10 µm
FTIR A 1 ppm Yes Yes 10–100 µm 20 Å–1 µm
SIMS HV <1 ppm No No 60 µm 0–10 µm
XRD A 1% No Yes 30 µm 0.1–10 µm
XRF A 1 ppb No No 1 cm Bulk
AFM A NA No Yes <0.1 Å 0.1 Å
a Experimental atmosphere whether ambient (A), high vacuum (HV) or ultra-high vacuum (UHV)

6.2 Chemical Nature of ZDDP and DDP

Most of the added ZDDP to the base oil is either neutral, basic or hybrid of both [21,
27, 29, 30], which are depicted in Fig. 6.1. Similarly, the DDP can be either neutral,
acidic or a mixture of both, as depicted in Fig. 6.2. The neutral ZDDP usually exists
in equilibrium state between the monomeric structure (Zn[PS2 (RO)2 ]2 ) shown in
Fig. 6.1(I) and the dimeric one (Zn2 [PS2 (RO)2 ]4 ) shown in Fig. 6.1(II) [31]. On
the other hand, the basic ZDDP is stable and has a structure with the chemical
formula (Zn4 [PS2 (RO)2 ]6 O) [29, 30, 32, 33], which is depicted in Fig. 6.1(III). Within
these types of ZDDP and DDP additives, new categories can be defined as primary,
secondary, tertiary, etc., depending on the number of carbon atoms, i.e. one, two
or three respectively, that are attached to the carbon atom of the alkyl (R) group
[32]. These functional groups predetermine the thermal stability of the additive,
which can be arranged as follows [30]: aryl > primary alkyl > secondary alkyl. It
is interesting to note that the least thermally stable additive provides the best wear
protection. Therefore, according to the alkyl functional group (R), the wear protection
performance can be arranged as follows [21]:

secondary alkyl > primary alkyl > aryl

This can be attributed to the fact that primarily the decomposition process of
the additive is thermally activated, which will be discussed in more detail in the
subsequent sections.
162 A. Dorgham et al.

Fig. 6.1 The different structures of ZDDP: (I) neutral dimeric in equilibrium with (II) neutral
monomeric; (III) basic ZDDP. Reprinted from Harrison and Kikabhai [31]
Fig. 6.2 The different
structures of DDP: (I) neutral
and (II) acidic DDP.
Reprinted from Kim et al.
[34]
6 Tribochemistry and Morphology of P-Based Antiwear Films 163

6.3 Reactions of ZDDP and DDP

Different theories were proposed to explain the reactions of P-based additives, such as
ZDDP and DDP, in lubricating oils, which can be grouped into three main categories
[21]: (i) ligand exchange, (ii) peroxides and peroxy-radicals decomposition and (iii)
thermal, oxidative or hydrolytic degradation. These reactions can have very complex
pathways due to the fact that the commercial oils containing ZDDP or DDP may
as well contain some impurities, other additives, detergents or dispersant that might
alter these paths and ultimately alter the precursors and the final reaction products
forming the tribo- or thermal films [35].
To avoid such complexities, the three reactions mentioned earlier will be discussed
in detail in the following sections assuming that only the P-based additive is present
in the oil.

6.3.1 Ligand Exchange

During the ligand exchange reaction, the zinc cations (Zn2+ ) in the ZDDP or any of
its decomposition products can be exchanged by another cation. For example, the
Zn2+ in the monomeric form of the neutral ZDDP can undergo a ligand exchange
reaction with another metal ion (M2+ ), as follows [21]:
   
Zn PS2 (RO)2 2 + M2+ → M PS2 (RO)2 2 + Zn2+ (6.1)

The relative order of the ability of one cation to replace another one is as follows
[36]:

Pd2+ > Au3+ > Cu+ > Cu2+ > Fe3+ > Pb2+ > Zn2+ (6.2)

This means that Zn2+ can be replaced easily by Fe3+ , which can be released from
the metal surface as a result of wear of parts of the contacting surfaces. This reaction
is very important as the nature of the metal cations in the substrate or the metal dialkyl
dithiophosphate molecules can alter the thermal decomposition process by changing
the decomposition temperature and kinetics and possibly the final composition of
the formed tribofilm [37].

6.3.2 Decomposition of Peroxides and Peroxy-Radicals

Peroxides and peroxy-radicals can oxidise the steel surface as follows [11]:

ROO· + Fe → RO· + FeO (6.3)

164 A. Dorgham et al.

2ROOH + Fe → 2RO· + 2OH− + Fe2+ (6.4)

· −
ROOH + Fe 2+
→ RO + OH + Fe 3+
(6.5)

Several studies [11, 12] found that in the presence of hydrogen peroxide, the oil
that has antiwear additives, such as ZDDP, results in less wear than the oil without
the additive. This suggested that one of the different antiwear mechanisms to protect
the contacting surfaces is by decomposing the peroxides and hence terminating the
oxidation chain reaction. However, other authors noted that when the antiwear addi-
tive acts as an antioxidant additive it cannot completely protect the surface from wear
[38]. This indicates that the additive is used in decomposing the peroxides instead of
forming a protective film. Therefore, there is a need to tailor the proactive tribolog-
ical considerations based on not only the general application but also the operating
conditions while inspecting the compatibility between the additive and contacting
surfaces [38].

6.3.3 Formation of Tribo- and Thermal Films

The currently accepted view on the formation of tribo- and thermal films regards
the decomposition of P-based antiwear additives as a thermally-activated and stress-
assisted reaction [40–42], which can be catalysed by either heat or mechanical action
in the form of rubbing or shear at the interface of the contacting bodies. Nevertheless,
the classical view of this decomposition reaction considers it as a chemical reaction
that can be either thermal, oxidative, hydrolytic or hybrid. These mechanisms will be
discussed in detail in the following sections while highlighting the effect of rubbing
on the reaction kinetics and the final decomposition products.

6.3.3.1 Thermal Decomposition

Numerous studies suggested that the decomposition of the ZDDP or DDP addi-
tive occurs thermally [35, 36, 43–45], which means that the extent and rate of the
decomposition process are temperature dependent. The high temperature is espe-
cially needed at the early stage of the decomposition process as reported by Jones
and Coy [46]. This was based on the observation that after the initial decomposition
of the ZDDP at high temperature, the decomposition continued at a high rate even
at low temperature as indicated by the continuous formation of white deposition in
the oil. This suggested that the reaction can be multistage where intermediates must
be formed before the final products. The initial stage of this multistage reaction can
occur at elevated temperature in solution [36, 40, 47−49] or on the steel surface,
which was argued to be essential for this process [4, 50–52].
6 Tribochemistry and Morphology of P-Based Antiwear Films 165

In the case that the steel surface is not required, it was suggested that the formation
of the protective film at high temperature can occur due to an in situ deposition process
on the steel surface by one or more of the following mechanisms [49]:
1. Polymerisation of small molecules to form a complex large molecule on the steel
surface.
2. Isomerisation of one or more of the additive molecules to other molecules
deposited on the steel surface as a protective layer.
3. Decomposition of the additive molecules and the deposition of the resulting
products on the steel surface.
4. Chemical reaction between two or more intermediate molecules to form the
surface layer.
On the other hand, Coy and Jones [50] suggested that the availability of the steel
surface can play a vital role in the thermal decomposition process. A reaction with
the steel surface can transform the organic phosphates into inorganic ones, which
can subsequently polymerise to a range of polyphosphates [46].
Jones and Coy [46] explained these reactions based on Pearson’s hard and soft
acids and bases (HSAB) principle [53–57]. The principle suggests stable pathways
for any chemical reaction according to the general observation that hard acids prefer
to form bonds with hard bases whereas soft acids favour forming stable compounds
with soft bases. Following the hardness classification of the acids and bases of the
most relevant compounds to ZDDP and DDP additives, which are summarised in
Table 6.2, Jones and Coy [46] proposed that the following chain of reactions can take
place during the thermal decomposition of the ZDDP:
1. Migration of the soft acid alkyl from the hard base oxygen atoms to the soft base
sulphur atoms of thiophosphoryl (P=S)
2. Formation of phosphorus acid as a result of the elimination of the thioalkyl (_SR)
functions.
3. Formation of phosphates P_O_P as a result of the nucleophilic reaction of one
hard acid tetravalent phosphorus O_P in one short phosphate segment with a hard
base oxygen bonded to another phosphate segment O_P.

Table 6.2 Classification of the hard and soft acids and bases relevant to ZDDP [55]
Type Hard Soft Borderline
Acid Tetravalent Tetravalent carbon Ferrous Fe2+ , Zn2+
phosphorus –P– (CH4 )
Base H2 O, OH–, O2 – Thiophosphoryl (P=S) Sulphite (SO−23 )
Phosphoryl (P=O), Thiolate (RS− )
PO− 3,
4

ROH, RO–, SO−24 Phosphine (PR3 )

166 A. Dorgham et al.

4. Formation of zinc mercaptide Zn(SR)2 as intermediate, which upon reaction with

an alkylating agent forms dialkyl sulphide whereas upon reacting with mercaptide
forms dithiophosphate, trithiophosphate and finally tetrathiophosphate.
5. Formation of a mixture of zinc thiophosphate, zinc pyro-thiophosphate and zinc
polypyro-thiophosphate as a final deposition.
Although the predictions of Jones and Coy [46] do not entirely match the current
consensus that the final decomposition product is a mixture of zinc pyro, poly- and
meta-phosphate glass, the way they utilised the HSAB principle to explain the thermal
decomposition reaction was revolutionary at the time. They were one of the first to
explore the thermal decomposition mechanism not as a black box but as a tractable
chain of chemical reactions in which several intermediates can be identified. It took
nearly two decades until Martin et al. [8] explored the HSAB in more detail to explain
various important parts of the ZDDP reactions using the chemical hardness approach
while matching the predictions with the experimental findings. This will be discussed
in the following section.

6.3.3.2 Thermo-Oxidative Decomposition

A number of studies proposed that the decomposition reaction of the P-based addi-
tives is a thermo-oxidative in nature. Willermet et al. [58] suggested that the antiwear
additive decomposition takes an oxidative pathway in the case of an equimolar quan-
tity of free radicals is present in the oil, otherwise a thermally controlled decom-
position occurred. However, the authors pointed out that in some localised areas
of a thermally controlled decomposition, products of a thermooxidative controlled
decomposition coexisted as well. Based on this observation, they concluded that apart
from the availability of free radicals, stress plays a controlling role in determining the
predominant decomposition pathway. To test the hypothesis of the thermo-oxidative
controlled decomposition, which is controlled by the availability of O2 , Willermet
et al. [58] conducted their tribological experiments in air and argon atmospheres. In
the two cases, they did not observe any differences in the tribofilm composition. This
should have necessary ruled out the oxidative decomposition mechanism. However,
the authors argued that the experimental conditions somehow did not allow the oxida-
tive decomposition to compete well with the thermal one. This apparent discrepancy
can be related to the role of the ZDDP as an antiwear additive and oxidation inhibitor
[21, 27], i.e. helps decompose the peroxy-radicals [59, 60].
Yin et al. [1] proposed a thermo-oxidative mechanism for the ZDDP tribofilm
growth starting with the strong chemisorption of the ZDDP to the oxide layer on the
steel surface followed by the fast formation of long polyphosphate chains and the
slow formation of short polyphosphate chains, which can be summarised as follows:
• Step 1: Physisorption or chemisorption
 
Zn (RO)2 PS2 2 (solution) → Zn[(RO2 )PS2 ]2 (adsorbed) (6.6)
6 Tribochemistry and Morphology of P-Based Antiwear Films 167

• Step 2: Formation of long polyphosphate chains

 
Zn (RO)2 PS2 2 + O2 (or ROOH) → Zn(PO3 )2 + sulphur species (6.7)

• Step 3: Formation of short polyphosphate chains

Zn(PO3 )2 + FeO → FeZnP2 O7 (6.8a)

or

Zn(PO3 )2 + 2FeO → Fe2 Zn(PO4 )2 (6.8b)

Following Yin et al. [1], Najman et al. [25] proposed similar steps for the decom-
position of the DDP additive, as follows:
• Step 1: Physisorption or chemisorption

(RO)3 PS2 (solution) → (RO)3 PS2 (adsorbed) (6.9)

• Step 2: Formation of long polyphosphate chains

(RO)3 PS2 + O2 + FeO → Fe(PO3 )2 + sulphur species (6.10)

• Step 3: Formation of short polyphosphate chains

Fe(PO3 )2 + FeO → Fe2 P2 O7 (6.11)

• Step 4: Formation of sulphate species

3
S + O2 + FeO → FeSO4 (6.12)
2
Similar to the above reactions, Martin et al. [8, 9, 61] suggested that the depoly-
merisation of the initially formed long phosphate chains of the ZDDP tribofilm into
shorter ones is a result of their reaction with iron oxides on the steel surface, as
follows:
θ,τ,P
5Zn(PO2 )2 + Fe2 O3 → Fe2 Zn3 P10 O31 + 2ZnO (6.13)

which can be accompanied by other intermediate reactions, such as:

4Zn6 (P10 O29 S2 ) + 10Fe2 O3 → 10Fe2 Zn(P2 O7 )2 + 8ZnS + 6ZnO (6.14)

These proposed reactions were justified based on Pearson’s principle of hard and
soft acids and bases (HSAB) [53]. According to this principle, the harder acid Fe3+
than Zn2+ will react preferentially with the hard base tetravalent phosphorus. Thus,
168 A. Dorgham et al.

under high temperature, shear and pressure, iron can be easily digested by the zinc
phosphate glass to form mixed iron and zinc phosphates. However, under severe
conditions of shear, the authors suggested that the digestion of iron can convert the
long zinc phosphate chains completely into short iron phosphate chains according to
the following reaction:
θ,τ,P
Zn(PO3 )2 + Fe2 O3 −−→ 2FePO4 + ZnO (6.15)

As these reactions need iron oxide and thus wear to commence near the metal
surface, the tribofilm is expected to have a uniform structure of zinc phosphate free of
iron in the case of mild wear conditions whereas it should have a layered structure in
the case of severe wear. In this structure, the short chains of zinc and iron phosphate
or iron phosphate are formed on the metal surface whereas the long chains of zinc
phosphate are continuously formed away from the metal surface.
Fuller et al. [62] suggested a thermo-oxidative decomposition mechanism similar
to one proposed by Yin et al. [1] but with two main modifications. The first one
concerns the additive adsorption [step 1 in (6.6)]. It was suggested that when the
antiwear additive such as the ZDDP adsorbs to the steel surface, it undergoes a trans-
formation into a rearranged linkage isomer in which the alkyl groups have migrated
from O to S atoms. This alkylation reaction is based on the mechanism proposed by
Jones and Coy [46], which was discussed in the previous section. However, Fuller
et al. [62] further suggested that in this linkage isomer all the sulphurs are partially
or totally replaced by oxygen. The other modification concerns the formation of the
short phosphate chains [step 3 in (6.8a) and (6.8b)]. Instead of being a product of the
reaction of the long phosphate chains with iron oxide, the short phosphate chains can
also be formed as a product of the reaction between the long phosphate chains and
water, which increases with increasing temperature. These steps can be summarised
for the case of the ZDDP additive as follows:
• Step 1: Physisorption or chemisorption of ZDDP to the metal surface
   
Zn (RO)2 PS2 2 (solution) → Zn (RO)2 PS2 2 (ZDDP adsorbed) (6.16)

• Step 2: Decomposition of ZDDP to LI-ZDDP

 
Zn (RO)2 PS2 2 (solution) → Zn[O2 P(SR)2 ]2 (LI-ZDDP insolution) (6.17)

• Step 3: Physisorption or chemisorption of any LI-ZDDP in oil to the metal surface

   
Zn O2 P(SR)2 2 (solution) → Zn O2 P(SR)2 2 (LI-ZDDP adsorbed) (6.18)

• Step 4: Thermal-oxidation of ZDDP and LI-ZDDP to form long polyphosphate

chains
 
Zn O2 P(SR)2 2 + O2 (or ROOH) → Zn(PO3 )2 + sulphur species (6.19)
6 Tribochemistry and Morphology of P-Based Antiwear Films 169

• Step 5: Hydrolysis of the long polyphosphate chains to form short polyphosphate

chains

7Zn(PO3 )2 + 6H2 O → Zn7 (P5 O16 )2 + 4H3 PO4 (6.20)

2Zn(PO3 )2 + 3H2 O → Zn2 P2 O7 + 2H3 PO4 (6.21)

Bell et al. [6] suggested that the adsorption step described earlier occurs only
on localised areas of the wear scar where asperities are rubbing against each other.
This heterogeneity in the adsorption was suggested to be responsible for the noticed
heterogeneity of the tribofilm thickness and composition. In addition, the authors
suggested that after the adsorption of the rearranged linkage isomer (step 3) and
before the formation of the long polyphosphate chains (step 4), sulphide products
can react with the steel surface to form iron sulphide (FeS).

6.3.4 Hydrolytic Decomposition

Few studies suggested that the decomposition of P-based additives such as the ZDDP
is hydrolytic in nature, i.e. catalysed by water. In order to provide evidence for
this mechanism, Spedding and Watkins [35] showed that in the absence of water,
e.g. by heating the sample up to 100–170 °C in order to evaporate all water as
well as by flooding the sample with dry nitrogen, the decomposition reaction was
suppressed. In contrast, when the sample was flooded by watersaturated nitrogen, a
rapid decomposition rate was observed. Therefore, they proposed that the following
hydrolysis reaction chain can take place:
1. alcohol formation

RO−P− + H2 O → ROH + HO−P− (6.22)

2. alcohol dehydration to olefin

H+
ROH → R  CH2 + H2 O (6.23)

The sum of the above two reactions demonstrates the catalytic action of water in
the overall reaction, as follows:

H2 O
RO−P −−→ R  CH2 + HO−P− (6.24)

3. polyphosphate formation

−P−OH + HO−P → −P−O−P− + H2 O (6.25)

170 A. Dorgham et al.

4. hydrogen sulphide formation due to the consumption of water by the sulphur

species

−P−− S + H2 O → −P−− O + H2 S (6.26)

It should be noted that when Spedding and Watkins [35] conducted tests at 200 °C,
at which most of the water should evaporate, the decomposition proceeded with-
out any significant reduction in the reaction rate. Willermet et al. [13] argued that
even lower temperatures than 200 °C could not slow down the reaction rate. This
either disproves the hydrolytic decomposition mechanism or simply suggests that
more complex reaction pathways or multiple mechanisms, e.g. thermo-hydrolytic or
thermo-oxidative-hydrolytic, can take place at the same time.

6.4 Composition of Antiwear Films

The tribo- and thermal films of P-based additives such as the ZDDP and DDP have a
dynamic tribochemical nature as their compositions change continuously during the
decomposition process [1, 63–66]. Nevertheless, the typical decomposition products
of such additives are some volatile products, e.g. olefins and mercaptans, and some
non-volatile products [45]. The non-volatile products can be oil soluble, e.g. organic
sulphur-phosphorus compounds, or can be oil insoluble, e.g. phosphate polymers,
zinc or iron thio- and poly-phosphate [43, 45, 46]. This indicates that the formed
antiwear films, in general, consist mainly of Zn or Fe, P, S, O and C. This was
confirmed using electron probe micro analysis (EPMA) [6, 67], scanning electron
microscopy (SEM) [5, 6, 43, 67–70], Auger electron spectroscopy (AES) [5, 58,
67–69, 71], X-ray fluorescence (XRF) spectroscopy [3, 4, 44, 72, 73], X-ray pho-
toelectron spectroscopy (XPS) [3, 6, 51, 52, 58, 69, 71, 74–77], X-ray absorption
near edge spectroscopy (XANES) [1, 62, 63, 65, 75, 78–80] and secondary ion beam
spectroscopy (SIMS) [3, 6, 35]. Willermet et al. [69] found that the carbon can only
be detected in the thinnest parts of the tribofilm. This suggested that this carbon
was originated from the environment or steel substrate but not a real part of the tri-
bofilm. Ancillary experiments conducted by Lindsay et al. [71] supported the same
conclusion that carbon existed as a result of adventitious sources, i.e. contaminants.
Furthermore, the results of Martin et al. [81] showed that no carbon existed in the
samples under study, which confirms its adventitious nature.
It should be noted that although the composition of the tribo- and thermal films
are similar they are not completely identical. Kasrai et al. [75] found differences
in the ratio of the elements forming the thermal film and tribofilm. Hence, other
factors than the high temperature, e.g. reaction with the steel surface, rubbing and
high contact pressure, can also play a role in determining the final decomposition
products [50, 75].
In the subsequent sections, the decomposition species of the ZDDP and DDP
additives will be discussed in detail. Phosphorus, zinc and iron species will be exam-
6 Tribochemistry and Morphology of P-Based Antiwear Films 171

ined first followed by sulphur species. The role of the operating conditions on the
formation of these species will be discussed as well.

6.4.1 Phosphorus, Zinc and Iron Species

The solid precipitates of the decomposition reaction of the ZDDP or DDP additive
is a complex mixture of zinc or iron polyphosphates depending on the available
cations. DDP forms, in general, tribofilms of short chains of iron polyphosphate
[25, 34, 66, 82, 83] with a minor concentration of sulphur species consisting mainly
of iron sulphate under high contact pressure [84, 85] and iron sulphides, e.g. FeS
or FeS2 , under low pressure [66, 84]. On the other hand, the ZDDP forms tribo-
and thermal films of zinc phosphate of different chain lengths and a small content of
sulphur in the reduced form of sulphides [46]. The chain length is typically a complex
function of the operating conditions, e.g. load and temperature, as well as the type
of additives, dispersant and contaminants such as water that can be present in the
oil. For instance, in the case of thermal films, increasing the temperature can lead
to the formation of longer chains of polyphosphates [35]. However, in the case of
tribofilms formed under rubbing, increasing the temperature can lead to the formation
of short chain pyrophosphate [45]. This apparent controversy can be attributed to the
observation that regardless of the oil temperature short phosphate chains are likely to
be formed near the steel surface as a result of the depolymerisation reaction occurring
to the long phosphate chains [8, 9, 86–89]. The depolymerisation can occur due to
the high shear stress at the asperity-asperity contacts, which can possibly cleave the
long phosphate chains into shorter ones, and to wear that can remove the weakly
adhered long phosphate chains from the surface [89]. In addition, other studies [8,
9, 86] suggested that in the presence of iron oxide the long phosphate chains are
depolymerised into short ones of mixed Fe–Zn or Fe phosphates, as discussed in
Sect. 10.3.3. The depolymerisation effect is significant especially in the case of
ZDDP tribofilms consisting initially of long to medium phosphate chains, whereas
it might not affect the short chains of iron phosphate predominantly formed in the
case of the DDP additive [84].
The ratios between metal cations (M+ ) and phosphorus (P) [90] on the one hand,
and between oxygen (O) and phosphorus (P) [58] on the other hand, are useful
indicators for the length of the phosphate chains. For instance, in the case of the
ZDDP, assuming that the zinc phosphates have a general formula of x(ZnO).(1 −
x)P2 O5 , then the chain length n can be related to the ratio of the mole fraction of
ZnO, x, to the one of P2 O5 , 1 − x, as follows [91]:
x n+2
 (6.27)
1−x n

Alternatively, if the zinc phosphate is assumed to have the general formula

Znn+2 [Pn O3n+1 ]2 , for odd n and Zn(n+2)/ 2 Pn O3n+1 for even n, then the chain length
172 A. Dorgham et al.

can be related to ratio of the atomic concentration of phosphorus to the one of oxy-
gen, as follows [9]:
P n
 (6.28)
O n+1
These two ratios above appeared to give comparable results for the chain length of
phosphates [92]. Alternatively, another option to quantify the length of the phosphate
chains would be the intensity ratio of the bridging oxygen (BO) to non-bridging
oxygen (NBO) [86, 92–97], as follows:
B 1n−1
 (6.29)
N BO 2 n+1
Based on (6.27), the phosphate chains seem to change from metaphosphate (n →
∞) at x  0.5, to orthophosphate (n → 1) at x  0.75. As the chain length changes
drastically over a small concentration range of ZnO, or FeO and Fe2 O3 in the case of
DDP, one should examine the presence of zinc and iron more closely. Increasing the
metal oxides content of these cations in the phosphate can increase the fragmentation
of the long chains and thus depolymerises them into shorter chains [97–99].

6.4.2 Sulphur Species

There is no clear consensus on the evolution of the sulphur species present in the
tribo- and thermal films. For instance, Bird and Galvin [51] suggested that the thermal
film contains sulphur in the form of sulphate and free sulphur whereas the tribofilm
contains large patchy areas of sulphide and a small amount of sulphate. Zhang et al.
[26] showed that for the ZDDP tribofilms, the evolution of P/S ratio was nearly con-
stant over time, whereas it showed a gradual increase in the case of DDP tribofilms.
Kim et al. [34] observed that for the ZDDP the sulphate concentration increases with
heating while sulphide decreases possibly due to its oxidation to sulphate. For one of
the tested DDPs, both sulphides and sulphates were detected whereas for the other
no sulphides were observed but only sulphates with a similar behaviour to the ZDDP.
Similarly, Najman et al. [85] showed that the DDP reacts rapidly with the substrate
covered with oxides, which leads to the oxidation of the sulphur species into iron
sulphate. This was confirmed for both neutral and acidic DDPs [100]. Other studies
[25, 84] showed that under high contact pressure, DDP tribo- or thermal films con-
tain Fe sulphate near the steel surface whereas under less harsh conditions initially
mixed iron sulphide, as FeS and FeS2 , and sulphate, as FeSO4 , are formed where the
sulphides can oxidise over rubbing time yielding primarily sulphates at the end.
The role of the operating conditions can explain some of the conflicting results.
For instance, Zhang et al. [66] reported that the DDP additive forms mainly FeS
whereas the ZDDP forms FeS in the early stage near the metal surface and ZnS in
the later stages. The observed sulphides in the case of DDP, which is in contrast to
6 Tribochemistry and Morphology of P-Based Antiwear Films 173

the previously discussed studies suggesting sulphates, can be related to the different
operating conditions used while generating the various tribofilms.
The complex decomposition reaction of the antiwear additives might have several
intermediates. One of these intermediates, which might be formed initially in the
solution even at low temperature and be deposited on the metal surface, can be
a sulphur-rich thiophosphate [43]. Subsequently, when the temperature of the oil
is raised or the local temperature at the asperity-asperity contacts of the rubbing
surfaces increases due to frictional heating, the reaction between the thiophosphate
deposit and rubbing surfaces becomes possible to occur.
The amount of the deposits of the sulphur-rich thiophosphate and its rate depend
closely on the thermal stability of the additive. In the case of the ZDDP, Spedding
and Watkins [35] found that by increasing the temperature, more of the sulphur is
consumed. In contrast, Kim et al. [34] found that for the ZDDP the progression of
heating makes the thermal films richer in sulphur whereas for the DDP it makes the
films richer in phosphorus. Despite the trend, this indicates that the local temperature
during the tribofilm formation can be inferred from the local composition. This is
possible by examining the local ratio of zinc to sulphur or phosphorus to sulphur,
which indicates the minimum temperature attained during the formation of that part
of the tribofilm [43].
Several other studies [3, 5, 6, 25, 26, 66, 82, 85, 101] suggested that the decompo-
sition products can be a result of a direct reaction between the additive and contacting
surfaces instead of being formed in the oil and deposited on the surface. This starts
with the adsorption of the additive to the steel surface with the maximum cover-
age occurs when the additive molecules are flat on the surface, which means that
the sulphur atoms lay near the surface [102]. Several authors [51, 103–105] have
already found that the amount of sulphur chemisorption products are higher on the
steel surface. Bell et al. [6] suggested that immediately after the adsorption of the
ZDDP to the steel surface, sulphide products from the decomposition of the ZDDP
react with the steel surface to form iron sulphides. In agreement with these results,
Loeser et al. [76] also found that the amount of sulphur is localised in the areas of
high pressure, i.e. asperity-asperity contacts. These various reports indeed corrob-
orate the observation of Dacre and Bovington [102] that initially the four sulphur
atoms should be near the steel surface.
Watkins [3] suggested that FeS reacts with the oxide layer to form a eutectic sys-
tem, which its phase and potential diagrams are depicted in Fig. 6.3. This system
has a melting temperature of 900 °C and was postulated to form a viscous film at the
contacting surfaces under extreme conditions. The results of Glaeser et al. [5] sup-
ported the idea of the formation of iron-oxide-sulphide complex. On the other hand,
Barcroft et al. [43] could not detect any complex mixture of zinc oxide, phosphate,
and sulphide as the one proposed by Watkins [3]. Nevertheless, they suggested that
this system can still be formed at the asperity contacts where temperature can be
high. Bell et al. [6] suggested that the iron sulphide and iron oxide can enter the
phosphate layer as cations or fragments from the worn surface. They also suggested
that the replenishment of sulphur and oxide to the layer on the steel surface occurs
continuously either by entrainment, mixing, or diffusion. Similar to these findings,
174 A. Dorgham et al.

Fig. 6.3 The Fe_O_S

ternary system that can be
formed on the steel surface
under rubbing. a Is its Phase
diagram. Reprinted from
Watkins [3]. And b is its
potential diagram at 673 K.
Reprinted from Watanabe
et al. [106]

Glaeser et al. [5] suggested that the iron sulphide forms a chemisorbed layer of
iron-sulphide-iron oxide that prevents direct contact between asperities.
Nevertheless, it should be noted that other studies [51, 58, 68, 71] showed that iron
is absent in the formed tribofilm, i.e. iron detected was mainly iron oxide rather than
iron sulphide. These results led to the conclusion that the tribofilm is deposited over
the substrate rather than being a result of a chemical reaction with the rubbing surface.
This also suggests that the formation of the tribofilm occurs through a thermal route
rather than an oxidative one.
6 Tribochemistry and Morphology of P-Based Antiwear Films 175

6.4.3 Factors Affecting the Tribofilms’ Composition

and Formation

6.4.3.1 Material of Counterbodies

As the additive molecules adsorb to the substrate before decomposing to form a

tenacious tribofilm, it is expected that the type of substrate, e.g. steel, DLC-coated,
ceramic and etc., should have an important effect on the tribofilm formation and its
tenacity. There are two views related to the effect of substrate. The first considers the
substrate effect to be mainly related to its mechanical properties such as hardness.
On the other hand, the second view considers that the substrate effect is related to a
more complex chemo-mechanical rather than mechanical properties of the rubbing
interfaces, such as the chemical, electronic and structural properties of the substrate.
These two views are discussed in the following sections with much of the emphasis
is put on the second.

6.4.3.2 Mechanical Properties

Several studies investigated the effect of the mechanical properties of the substrate,
especially hardness and elastic modulus, on the tribofilm formation. The overall
conclusion is that the tribofilm formation, friction and wear are affected by not only
the mechanical properties of the substrate but also the compatibility between the
substrate and additive [38]. For instance, Li et al. [107] used XANES to investigate
the effect of steel hardness on the composition of the formed ZDDP antiwear films.
The results indicated that longer phosphate chains are formed on the hard substrates.
The results also indicated that the softer the substrate the larger the surface rough-
ness, tribofilm thickness, wear scar width and friction. Furthermore, more uniform
but thinner tribofilms were formed on the hard substrate but less uniform but thick
tribofilms were formed on the soft ones. FIB/SEM was used to examine the elements
present in the formed tribofilms and to assess the substrate damage based on the
dimensions of the wear scar. It was found that mainly sulphides and few sulphates
were formed on the hard substrates whereas the soft ones contained more sulphates
and fewer sulphides.
Sheasby et al. [108] also studied the effect of steel hardness on the formation
of protective tribofilm and wear performance. The study found that the wear per-
formance was enhanced with increasing the substrate hardness, although wear was
in general small, even for the soft substrates. The results showed that in several
occasions, medium softened substrates resulted in less wear than harder substrates.
They argued that mechanical mixing can occur between the formed tribofilm and
the soft substrate, which can result in improving the overall mechanical properties
of the interface and thus better resisting wear. However, the results of Vengudusamy
et al. [109], using six different types of DLCs of hardness ranged from about 760 to
6800 HV, suggested that there is no direct correlation between wear resistance of the
176 A. Dorgham et al.

contacting surfaces and the surface hardness. This further highlights the significance
of tailoring the tribological and mechanical properties of the interface by taking into
account the compatibility between the substrate and used additives [38].

6.4.3.3 Chemo-Mechanical Properties

Apart from hardness, the substrate chemical properties are expected to affect the
decomposition of the antiwear additives and the formation of their protective tri-
bofilms. This effect is manifested in the fundamental question of whether there is a
substantive requirement for the presence of metal cations, e.g. Fe, W and Ti, for the
tribofilm to be formed or such metallic cations are dispensable. There is a definite
consensus that the ZDDP molecules can adsorb to steel surfaces and decompose
to form protective tribofilms of excellent tenacity [21, 27, 28]. Furthermore, other
studies showed that ZDDP tribofilms can also be formed on surfaces other than
steel. For instance, Zhang and Spikes [41] were able to generate ZDDP tribofilms
on a WC substrate. The results showed that the rate of formation ranges from 0.2
to 0.7 nm/min depending on the interfacial shear stresses. Similarly, Gosvami et al.
[40] using elaborate in situ AFM tribotests showed that the ZDDP can adsorb and
decompose to form tribofilms on both Fe-coated and uncoated Si substrates with
a similar rate depending on the temperature and contact pressure. It is not clear
whether the ability to form tribofilms in these cases was due to the presence of W
and Si in particular or due to the operating conditions. This can be better understood
by studying some cases of coated surfaces, e.g. DLC, both non-doped and doped
with various metallic cations. For such surfaces, the exact chemo-mechanical nature
of the coating is expected to play a vital role in the adsorption and decomposition of
the P-based additives.
The wear and friction performance of DLC coatings and the properties of any
formed tribofilms on them were reviewed extensively in the literature [38, 110–114].
Several previous studies reported that the P-based additives such as ZDDP can react
and form protective tribofilms on DLC coatings even without containing any doped
metallic cations [109, 115–120]. In contrast, other studies found that no tribofilms can
be formed on non-doped DLC without metallic cations [121–126]. We will review
some of these reports and try to identify the reasons behind this apparent discrepancy.
The available literature will be divided into three main themes: (i) formation of
tribofilms on DLC, (ii) structure of the formed tribofilms on DLC and (iii) effect of
the formed tribofilms on friction and wear.
(i) Formation of tribofilms on DLC coatings
Regarding the formation of tribofilms on DLC coatings, it is interesting to note that
the literature is somehow divided between two contrasting views. The first suggests
that no tribofilm formation is possible on non-doped DLC whereas the other suggests
that the formation is possible on any type of DLC coatings.
In support of the first view, Haque et al. [121] found, using XPS, that no ZDDP-
derived tribofilm was formed on a-C:H DLC even in the case of one of the counter-
6 Tribochemistry and Morphology of P-Based Antiwear Films 177

bodies is made of cast iron. The role of iron cations was suggested to be deactivated
by the presence of a carbon transfer layer, which might be graphitic, on the cast
iron counterbody. This is also in line with the results of Bouchet et al. [125] regard-
ing the interaction of the ZDDP and MoDTC additives with hydrogen-free DLC
(a-C), hydrogenated (a-C:H) and Ti-doped (Ti–C:H) DLC coating s. The XPS and
TEM/EELS analysis revealed that no P-derived tribofilm was formed and no iron
was present, which indicated that the pin surface was protected by a transfer layer
from the DLC. Similarly, Podgornik et al. [122] investigated the effect of coating one
contacting surface or both with WC doped hydrogenated DLC coatings (Me–C:H) in
the presence of ZDDP antiwear additive. Based on the SEM/EDX analysis, the study
found that no P-based antiwear tribofilm was formed on the DLC coating. Further-
more, Podgornik et al. [124, 127] using SEM/EDX confirmed that for (a-C:H) DLC
no tribofilm was formed for both S-based and P-based additives. The results also
suggested that probably W from the coating can combine with S to form a protective
layer. Furthermore, Ban et al. [126] tested the reaction of Si-doped (a-Si:H) and non-
doped (a-C:H) DLC coatings with ZDDP. Using XPS, no tribofilm was detected on
the (a-C:H) DLC coating whereas a P-based tribofilm was formed on the Si-doped
DLC. Kalin et al. [128] also found that for the self-mated (DLC/DLC) nondoped
and doped Ti-, W-, and Si-DLC coatings, the EDX and FTIR results did not provide
evidence to support a reaction between the DLC coatings and the extreme pressure
(EP) additives. In another study, Kalin et al. [118] found that the tribo- and thermal
films can be formed on the Ti-doped DLC, which was found to have 10 times higher
activity than the W-doped (P/S > 25 folds). They also found some reactivity with a-
C:H DLC but much slower. They concluded that the metal doped DLCs behave more
like metal steel that catalyses the decomposition of the additive and the formation of
tribo- or thermal films.
In contrast to the previous view, Akbari et al. [120], using ATR-FTIR and XPS,
found that the thermal films can be formed equally on bare steel surfaces and Si-doped
(a-Si:H) and non-doped (a-C:H) DLC coatings. The thickness of the thermal films as
estimated from XPS sputtering was thicker in the case of bare steel as compared to the
coated ones, which suggested that steel can still catalyses the thermal decomposition
of the ZDDP even without rubbing. Apart from this, the decomposition products of the
ZDDP does not seem to be affected by the substrate as the short chain pyrophosphate
and zinc oxide was detected on all the tested surfaces. Nonetheless, ATR-FTIR
showed that organic sulphides (R–S) were present on all the surfaces but sulfhydryl
(R–SH) groups were detected on the steel surface only. These species were confirmed
by their XPS analysis except for the H–S, which could not be resolved. Nevertheless,
it should be noted that these results apply only to thermal films, which could indicate
that in the case of tribofilms any formed species on the DLC coatings, especially if
non-doped, are of low tenacity and thus under rubbing can be easily removed. This
can partly explain the discrepancy in the literature regarding whether a tribofilm can
be formed on non-doped DLC coatings or not. However, other studies, e.g. see [109,
115, 116], suggested that tribofilms can indeed be formed on the non-doped DLCs.
However, despite the assertion, the data provided do not seem to be conclusive for
178 A. Dorgham et al.

the a-C:H non-doped DLC, as will be discussed in the next paragraph describing the
structure of what was perceived to be a tribofilm on DLC coatings.
(ii) Structure of the tribofilms formed on DLC coatings
The studies that showed a formation of P-based tribofilms on DLC coatings indi-
cated unique tribofilms’ structures with some similarities to the ones formed on
bare steel surfaces. For instance, Vengudusamy et al. [109] analysed the ZDDP tri-
bofilms possibly formed on six different types of DLCs using various experimental
techniques including AFM, SLIM, ToF-SEM and EDX. The DLCs had a hardness
ranging from about 760 to 6800 HV. The structure of the tribofilm was analysed
using the AFM. The study found that a tribofilm with a pad-like structure similar to
one formed on steel surfaces can only be observed in the case of a DLC containing
metallic elements, e.g. W-doped DLC. For the other types of DLC, tribofilms of
minuscule amount with scattered patches were formed instead. The thickness of the
pads formed on the W-DLC was <30 nm, which was much larger than the one formed
on the other types of DLC. This may suggest that the metallic cations can catalyse
the decomposition of the additive and the formation of a protective tribofilm. SLIM
interferometry images after different rubbing times showed an interesting behaviour
that the tribofilms formed on DLCDLC contacting surfaces are less tenacious than
the ones formed on metal surfaces, which was also reported by other studies as well
[115, 116]. The low tenacity was evidenced based on the decrease in the concentra-
tions of the decomposition products over rubbing cycles. The ToF-SIMS analysis of
Equey et al. [115] and the XPS and TEM/EELS analysis of Haque et al. [121] showed
that no iron is present in the formed tribofilms on DLC surfaces, which suggests that
the low tenacity to the surface can be related to the absence of mixed oxide/sulphide
base layer.
In agreement with the previous discussion, Equey et al. [115] using AFM found
that the tribofilm seemed to lack the pad-like structure observed in the tribofilms
formed on steel surfaces but instead it appeared to have a similar structure like the one
of DLC coating, i.e. rounded nodular structure. The AFM images indicated that the
tribofilm does not cover the metal surface completely but forms more like scattered
islands of thickness up to 100 nm. In a different study, Equey et al. [116] tested the
decomposition of ZDDP, butylated triphenyl phosphorothionate (b-TPPT) and amine
phosphate (AP) additives on (a-C:H) DLC coating. Using ToF-SIMS and AFM, it
was found that the ZDDP and b-TPPT can form thin tribofilms, although AFM was
unable to quantify their thicknesses as they seemed to be within the roughness of the
DLC coating. This is in line with the conclusions of Topolovec et al. [117] regarding
(a-C:H) DLC and Cr-doped non-hydrogenated and graphitic DLC coatings.
The above discussion indicates that the main factor behind the wide disparity
between the observations of the tribofilm formation on DLC coatings is the tribofilm
tenacity. Any formed tribofilm does not seem to strongly adsorb to the DLC surface,
especially if non-doped, leading to its effortless removal under rubbing once formed.
Another explanation was provided by Kalin and Vižintin [129] who suggested the
presence of a different thermal activation barrier for the additive reaction with doped
as opposed to nondoped DLC, which is also different from the one of steel. The
6 Tribochemistry and Morphology of P-Based Antiwear Films 179

difference was attributed to the low thermal conductivity of the non-doped DLC in
comparison to steel or metal doped-DLC. This can lead to higher contact temperature
that helps in decomposing the additive and forming a tribofilm.
(iii) Effect of the formed tribofilms on friction and wear
Apart from the additive decomposition to form a tribofilm on the DLC surface, its
role in friction and wear is contentious. For instance, Haque et al. [130] suggested
that the decomposition of the ZDDP additive seems to be essential in protecting
the surface from polishing wear. This was based on the observation that outside
the boundary lubrication regime where the conditions might not be suitable for the
decomposition of the additive, the coating surface undergoes sp3 to sp2 conversion
(possibly graphitisation), which reduced friction and the DLC hardness leading to
more wear. Yang et al. [131] showed that the extent of this is additive dependent
but could not establish a direct relation between the friction reduction and sp2 /sp3
ratio in the coating or between the hardness reduction and wear volume. Contrary
to these results, Vengudusamy et al. [109] suggested that no graphitisation takes
place because of rubbing and that for most of the tested DLCs including a similar
a-C:H DLC coating used by Haque et al. [130], wear was less in the absence of
ZDDP than in its presence. In contrast, the results of Bouchet et al. [125] suggested
that ZDDP additives can improve wear properties of the hydrogen-free DLC (a-C),
hydrogenated (a-C:H) and Ti-doped (Ti–C:H) DLC coatings. Similarly, Equey et al.
[115] tested (a-C:H) DLC coating and found that although wear was insignificant in
the presence or absence of the ZDDP, its absence caused some abrasive wear scars
to appear on the surface of the coating.
Kalin and Vižintin [129] investigated the interactions occurring between the con-
tacting surfaces when only one of them is coated, i.e. DLC versus steel. They studied
a-C:H, Ti–C:H, a-C:H/a-C:H–W multilayer and a-C:H/a-Si:O single layer when
interacting with antiwear additive (mixture of amine phosphates) and EP additive
(dialkyl dithiophosphate). Wear was found to be dependent on the coating and addi-
tive types. Metal doped DLCs and steel were found to have similar tribological
behaviour except for steel/W-DLC, which shows large friction and wear. The EDX
surface analysis shows no P derived tribofilm was formed on the DLC coatings. In
the case of non-doped DLC, the additive presence seems to increase wear whereas
the opposite was found in the case of doped DLCs. They related this to the formation
of a soft tribofilm on the metal surface or metal doped DLC coating instead of a
transfer layer from the coating. The soft tribofilm can decrease wear of the substrate
more than in the case where the additive is not present. In line with these results,
Podgornik et al. [122] tested the effect of coating one contacting surface or both with
WC doped hydrogenated DLC coatings (Me–C:H) in the presence of ZDDP antiwear
additive and found that although no P-based antiwear tribofilm was formed on the
DLC coating, the additive type seems to have a great effect on the running-in period.
Interestingly, the results showed that the DLC/iron combination gives the best tri-
bological performance in terms of low friction and wear, which is in line with other
studies [123]. The EDX analysis revealed that in this case a mixed material from
the DLC coating and decomposition material from the additive can form a protective
180 A. Dorgham et al.

tribofilm on the uncoated or exposed steel surfaces, which seems to exhibit a superior
protective properties than the individual components. Donnet and Grill [132] indi-
cated that the low shear stress of the carbon-rich transfer layer formed on the DLC is
responsible for the improved tribological properties in comparison to steel surfaces.
Ban et al. [126] tested the reaction of Si-doped (a-Si:H) and non-doped (a-C:H) DLC
coatings with ZDDP. Using XPS, a P-based tribofilm was formed on the Si-doped
DLC, which seemed to have low shear stress that helped lower friction force. In
contrast, Kalin et al. [128] found that for the self-mated (DLC/DLC) non-doped and
doped Ti-, W-, and Si-DLC coatings, the use of EP additives mitigated wear but
friction relatively increased. The increases in friction suggested a formation of a
high shear strength interface, despite the observed reduction in surface roughness.
However, the EDX and FTIR results did not provide evidence to support a reaction
between the DLC coatings and additives.
Based on the above discussion, any tribofilm formed on the DLC coating can
generally serve two purposes [119]. First, it can form a lubricating layer that helps
separating the contacting surfaces. Second, the formed layer is soft and thus can
reduce the peak shear stresses and strains at the surface of the coating [133, 134]. This
can result in possibly graphitisation suppression and wear reduction [119] though
the effect on friction can be higher or lower depending on the type of the formed
interface, i.e. friction increases in the case of ZDDP tribofilms but decreases in the
case of carbon-rich layers.

6.4.3.4 Type of Base Oil

P-based additives, such as ZDDP and DDP, can be represented typically as a polar
moiety, attached to P or S atom, and a non-polar tail, which is typically an alkyl
moiety [135]. The affinity of such a molecule to the steel surface originates from
the molecules’ polarity due to the electron charge difference, i.e. asymmetric charge
distribution, between the different parts of the molecule due to the difference in
electronegativity between the bonded atoms [135]. Tomala et al. [136] used the AFM
to study the effect of base oil polarity on the decomposition of the ZDDP. The study
found that in the case of non-polar oils, the tribofilms were formed faster, thicker and
caused less surface roughening than the ones formed in the case of polar oils. The
explanation behind this is that polar base oils, as expected, have a large affinity to
the steel surface and thus compete with the additive and hinder its accessibility to the
metal surface. On the other hand, non-polar oils was found to improve the additive
accessibility to the steel surface, which results in accelerated formation of thicker
tribofilms than in the case of polar oils [135, 137]. The composition of the tribofilms
formed in polar and non-polar oils were similar consisting mainly of phosphate but
a difference was observed in the formed sulphur species. For the ZDDP in polar oils,
sulphide species were formed as opposed to the mixture of sulphide and sulphate
species in the case of non-polar oils [135].
Kar et al. [138] used TEM, SEM, XPS, AFM and microhardness tests to investigate
the effect of base oil polarity on the decomposition of additives. The results showed
6 Tribochemistry and Morphology of P-Based Antiwear Films 181

that polar oils provided smoother tribofilms than the ones formed in non-polar oils.
The results indicated that the coverage of polar oils on the metal surface enhanced the
effective lubrication and protected the metal surface. In contrast, other studies [139,
140] found that the P-containing additives reduce wear more efficiently when used in
non-polar oils due to the ease of access of the additive to the steel surface and thus the
better formation of the protective antiwear tribofilm. Suarez et al. [137] used SLIM,
SEM-EDX XPS, AFM and nanoindentation to study the effect of oil polarity on the
interplay between the tribological properties and composition of the formed ZDDP
antiwear tribofilms. The study showed that the tribofilm thickness is not the only
parameter determining the wear and friction performance but the composition can
play a vital role as well. The results showed that although the polar base oil suppresses
the formation of thick tribofilms and decreases their formation rate, it reduces friction
more efficiently than the thicker tribofilms formed in the non-polar oils. The tribofilm
structure in the polar oil seemed smoother and more homogeneous and continuous
than the one formed in non-polar oils. This smoother tribofilm provided better wear
properties, which was proposed to be related to the harder tribofilm formed in the
case of polar oils as opposed to the softer tribofilm in the case of non-polar oil. The
study suggested that this softer tribofilm leads to a larger contact area and thus larger
wear. However, larger contact area should necessarily mean lower contact pressure
and thus wear is expected to be less. This discrepancy cannot be explained based on
hardness point of view only. In addition, measuring the hardness of a tribofilm of
thickness less than 40 nm with the NI technique using an indentation depth of 20 nm,
will inevitably include a significant effect from the hard substrate underneath.

6.4.3.5 Duration of Rubbing and Heating

The duration of rubbing and heating can greatly affect the composition of the formed
tribofilms and therefore their antiwear properties. Yin et al. [1] used XANES to
study the decomposition species of the ZDDP after different rubbing times, i.e. 5 min,
30 min, 6 h and 12 h. They also investigated the effect of rubbing a mature film formed
after 30 min in base oil for another 5.5 h to examine the durability of the tribofilm
and the effect of rubbing without additives. The results indicated that the tribofilms
formed after short rubbing times exhibit different fingerprints than the ones formed
after longer rubbing times, which suggested that different species are formed initially
before the phosphate chains. They also found that initially a large concentration of
unreacted ZDDP adsorbs to the metal surface, which can coexist with the reacted
ZDDP. Rubbing the ZDDP tribofilm in base oil led to the depolymerisation of the
long chains into shorter ones.
Similarly, Gosvami et al. [40] used the in situ AFM liquid cell shown in Fig. 6.4
to study the evolution of ZDDP tribofilms over time. The study showed that initially
the tribofilm volume increased linearly over rubbing time. However, after a certain
threshold the volume started to have exponential growth. The EDX and Auger elec-
tron spectroscopy (AES) analysis of the tribofilm revealed that it consisted mainly
of Zn, P and S. Furthermore, the distribution of Fe was uniform inside and outside
182 A. Dorgham et al.

Fig. 6.4 In-situ AFM liquid cell used by Gosvami et al. [40]

the wear scar indicating that the tribofilm does not have any significant amount of
Fe, if any.
Different other studies also used in situ XANES to follow the change in com-
position of ZDDP thermal films over heating time. For instance, Dorgham et al.
[141] used the in situ XANES tribotester shown in Fig. 6.5 to follow the change in
composition of ZDDP tribo- and thermal films over time. The results showed that
the formation rate of the tribofilm is much faster than the thermal one but the two
eventually have a similar composition. It was also found that the decomposition of
ZDDP starts with the formation of sulphate species, which over time is reduced into
sulphide along with the formation of zinc phosphate. In agreement with these results,
Morina et al. [64] using the in situ XANES heating cell shown in Fig. 6.6 showed that
in the beginning of the thermal decomposition process, sulphide and sulphate species
are formed. In order to have a surface sensitive signal, the in situ experiments were
performed in the total external reflectance mode by tilting the samples to an angle
less than the glancing angle. The XANES results showed that in the beginning of
the thermal decomposition reaction, sulphide and sulphate species are formed. The
6 Tribochemistry and Morphology of P-Based Antiwear Films 183

Fig. 6.5 Schematic of the assembly of the tribological apparatus used by Dorgham et al. [141] for
in-situ XAS experiments

Fig. 6.6 Schematic of the assembly of the thermal liquid cell used by Morina et al. [64] in-situ
XAS experiments

sulphate formation was related to the effect of relative humidity [64, 143, 144], i.e.
water contamination in the oil. The presence of sulphate species was also detected
by Ferrari et al. [142, 145] who used the heating cell shown in Fig. 6.7 to follow the
ZDDP thermal films over heating time.
For the DDP antiwear additives, Kim et al. [35] tested two types of DDPs and
compared them with ZDDP. They performed these experiments at 170 °C at which
the additive decomposes in the oil and produces solid precipitates, which are subse-
quently deposited on steel coupons underneath. The results indicated that the pro-
gression of heating makes the ZDDP thermal films richer in sulphur whereas the
DDP films richer in phosphorus. Zhang et al. [26] found that the evolution of P/S
184 A. Dorgham et al.

Fig. 6.7 Schematic of the

assembly of the thermal
liquid cell used by Ferrari
et al. [142] for in-situ XAS
experiments

ratio was nearly constant over time for the ZDDP tribofilms, whereas it showed a
gradual increase for the case of DDP tribofilms. This is despite the general increase
in the tribofilm thickness over rubbing time, which indicates that the composition
evolves mainly during the initial stage of rubbing [25]. Najman et al. [25] suggested
that initially the DDP additive reacts rapidly with the substrate covered with oxides,
which leads to the oxidation of the sulphur species into iron sulphate. These species
do not help in protecting the contacting surfaces but only does the subsequently
formed phosphate [85].
The DDP additive forms Fe phosphate of short chains whereas the ZDDP forms
initially zinc phosphate of short chains that grow into longer ones away from the
metal surfaces whereas the layers near the substrate remain of shorter chains due to
the presence of Fe cations near the steel substrate [25, 66].

6.4.3.6 Concentration of Additive

The concentration of the ZDDP affects its adsorption and coverage on the metal sur-
face [105, 146], i.e. the larger the concentration of the ZDDP, the larger its adsorption
and coverage. Yin et al. [1] used XANES to study the effect of the ZDDP concentra-
tion, i.e. 0.25%, 0.5%, 1.0% and 2.0 wt%, on the decomposition species. The results
showed that a low concentration of ZDDP can lead to the formation of short phosphate
chains whereas a high concentration leads to the formation of long chains in addition
to the short ones and unreacted ZDDP. Furthermore, Tomala et al. [136] found that
the larger the additive concentration (i.e. 2 and 5%) the thicker the tribofilm and the
larger its roughness. Similarly, Ghanbarzadeh et al. [147] indicated that increasing
the additive concentration increases the formation rate and terminal film thickness
and decreases the average wear depth. Thus, the additive concentration evidently has
a strong effect on the decomposition kinetics. This is further supported by the study
6 Tribochemistry and Morphology of P-Based Antiwear Films 185

of Akbari et al. [120] of the effect of ZDDP concentration, i.e. 1%, 2.0% and 20
wt%, on the composition of thermal films formed on bare steel, Si-doped (a-Si:H)
and non-doped (a-C:H) DLC coatings. The results showed that the thickness of the
thermal film increases linearly with the concentration with a rate about 2.75 nm/min
on bare steel and 1.5 nm/min on a-C:H DLC coating.

6.4.3.7 Temperature and Load

Temperature and load are expected to affect not only the composition of the formed
tribofilm but also its structure as well as its tribological and mechanical properties.
Zhang and Spikes [40] showed using the AFM that at low temperature of 60 °C the
tribofilm morphology was patchy of roughness similar to the large initial roughness
of the substrate. As temperature increased to 100 and 120 °C, the tribofilm thickness
increased and a more pad-like structure was formed with wide pads elongated in
the direction of rubbing. Similarly, Yin et al. [1] showed that the high temperature
accelerates the decomposition of the unreacted ZDDP and helps increase the length
of the formed polyphosphate chains. However, the study found that after a certain
threshold, the high temperature can be detrimental to the chains length. Short chains
were formed at 200 °C as opposed to the long chains formed at 100 and 150 °C.
The study also found that high temperatures promote sulphate species formation.
At 100 °C, sulphide species were detected whereas sulphate species were found at
200 °C. Other studies observed a similar effect for the load [84]. For instance, under
high contact pressure, the DDP additive forms mainly sulphate, e.g. FeSO4 , whereas
under low pressure mainly disulphide (FeS2 ). Contact pressure did not seem to affect
the already short chains composing the DDP tribofilms but reduced the chain length
of the ZDDP tribofilms from long to medium.
Similar to load, temperature can notably increase the formation rate and terminal
film thickness of the formed tribofilms but decrease the average wear depth [147].
The tribofilm thickness during the early stages of rubbing, e.g. running-in period,
and not only the steady state thickness, seems to be responsible for the antiwear
protection [147]. Similarly, Parsaeian et al. [148, 149] found that by reducing the
temperature from 100 to 80 °C, the tribofilm thickness decreases accompanied by a
conspicuous increase in wear.
The effect of temperature on the mechanical properties of postmortem tribofilms
was studied by Pereira et al. [150] using nanoindentation. The tribofilms consisted
mainly of polyphosphate of medium chain length and sulphides as indicated by the
XANES analysis. The tribofilm thickness measurement based on FIB-SEM cross-
section suggested a thickness of 180 ± 60 nm, which is consistent with the XANES
P k-edge estimation of 105 nm. The evolution of the indentation modulus over tem-
perature, i.e. from 25 to 200 °C, showed that up to 200 °C the modulus was nearly
constant at around 100 GPa but dropped to about 70 GPa at 200 °C. The decrease
in the modulus might be responsible for the good antiwear properties of the ZDDP
tribofilm, i.e. the low modulus means a compliant sacrificial tribofilm that can be
easily worn instead of the substrate.
186 A. Dorgham et al.

The combined effect of load and temperature on the evolution of ZDDP tribofilms
was investigated by Gosvami et al. [40] using in situ AFM tribotests. The study
showed that temperature exponentially increases the growth rate of the tribofilm.
Similar to temperature, load ranging from 2 to 7 GPa, appears to increase the growth
rate of the tribofilm exponentially until reaching steady state. Therefore, the study
concluded that load and temperature have the same catalytic effect on the tribofilm
formation. This is in agreement with the results of Yin et al. [1], which found that
load has the same effect as temperature, i.e. accelerating the decomposition of the
unreacted ZDDP and the formation of long phosphate chains. The study also found
that load does not affect the sulphur species type or concentration, i.e. sulphides
were the only sulphur species to be formed and their intensity did not change much
with load. Zhang and Spikes [41] showed using SLIM that in the EHL regime the
larger the load (50–75 N) the faster the tribofilm formation and thicker the terminal
thickness of the formed tribofilms on WC. The formation rate ranges from 0.2 to
0.7 nm/min depending on the maximum shear stress (220–250 MPa) at the edge of
the tribofilm. Similarly, Tomala et al. [136] found that the larger the load, ranging
from 1.3 to 2.4 GPa, the thicker the tribofilm and the larger its roughness.
The effect of load on the ZDDP decomposition was also studied by Ji et al. [151]
using XANES and AFM. The results indicated that the formed tribofilms consisted
of polyphosphate of thickness between 10 and 100 nm. The results also showed that
the larger the load the thicker the tribofilm. Additionally, based on the correlation
between the friction force behaviour and electrical contact resistance measurements,
they concluded that the ZDDP decomposition and tribofilm formation undergo three
different stages:
1. Induction period, which increases with load and generally decreases with sliding
speed. During this period, ZDDP molecules adsorb to the substrate and friction
increases.
2. Tribofilm growth period during which tribofilm thickness increases and friction
stays constant.
3. Tribofilm growth and removal period during which tribofilm thickness and fric-
tion stay constant.
The effect of load and temperature on DDP additives seems to be similar to the
one of ZDDP. It was reported that the higher the temperature the thicker the thermal
film [82]. Furthermore, the tribofilm thickness was found to be at least twice the
thickness of the thermal film [82] and in general increases gradually over rubbing
time [25], which suggests that load and temperature can both help accelerate the
decomposition reaction.
In summary, the above discussion suggests a similar effect of temperature and
contact pressure on the growth rate of ZDDP and DDP tribofilms, which indicates
they have a catalytic effect on the additive decomposition process. Thus, it can be
stated that the decomposition of P-based additives is a thermally and mechanically
assisted reaction, which can be activated by the availability of either shear or heat.
6 Tribochemistry and Morphology of P-Based Antiwear Films 187

6.4.3.8 Sliding and Rolling Speeds

It is understood that the ratio of sliding to rolling speeds can affect the contact severity
and thus the lubrication regime. To further understand this effect on the decomposi-
tion of the ZDDP, Suarez et al. [135] followed the ZDDP tribofilm formation under
different levels of slide-to-roll ratio (SRR), i.e. 0, 2, 5 and 10%. The study found that
the larger the SRR, the faster the tribofilm formation. However, the study showed that
the exact sliding percentage, as long as >0, does not have a significant effect on the
limiting tribofilm thickness. On the other hand, the SRR appeared to have a different
effect on the topography of the tribofilm depending on the base oil. In the case of
polar oil, the tribofilm appeared homogeneous throughout the different levels of the
SRR. When nonpolar oil was used, low SRR generated smooth tribofilms whereas
large SRR (≥5%) generated a rougher pad-like structure. That is the larger the SRR
is, the rougher the tribofilm topography becomes and the more the structure appears
patchy.
The effect of sliding speed on the ZDDP decomposition was also investigated by
Ji et al. [151] using XANES and AFM. The results showed that lower velocities and
higher load increase the tribofilm thickness whereas higher speeds and lower loads
decrease the tribofilm thickness. In terms of friction relation to velocity, there was
no simple relation. Under all the tested loads there seems to be a certain threshold
of sliding speed below which the friction appears to decrease with increasing the
sliding speeds whereas the opposite trend was observed above this threshold.
The effect of high SRR, i.e. 50–230%, on the ZDDP tribofilm formation was
examined by Shimizu and Spikes [152] using SLIM, EDX and AFM. The study found
that in the mixed sliding and rolling condition, the formation rate of the tribofilm
is less sensitive to the exact SRR, which indicates that the rubbing time is far more
important than the sliding distance.

6.5 Mechanical Properties of Antiwear Films

6.5.1 Structure

The first report on the structure of the P-based antiwear films goes back to the study
of Bird and Galvin [51] who suggested that the ZDDP tribofilm is polymeric and has
a patchy structure. The authors conjectured that the first layer of this structure is a
discontinuous sulphide, possibly zinc sulphide, or sulphate layer covering the metal
surface. This layer in turn is covered with islands of unknown compounds containing
Zn, P, and S. They also found that the ratio of Zn:P:S changes considerably along
and across the wear scar. This heterogeneity can be related to the different local
conditions, e.g. temperature, pressure and load. For instance, Sheasby et al. [67] and
Palacios [7] observed that at low temperatures, e.g. 35 °C, a thin but uniform film
was formed as compared to a thick but less uniform film at higher temperatures, e.g.
188 A. Dorgham et al.

Fig. 6.8 Arrangements of

atoms in the amorphous
tribofilm. Reprinted from
Martin et al. [81]

150 °C. The tribofilm formed at high temperature was patchy in nature suggesting
that it was only formed at the asperity-asperity contacts where temperature was high.
In addition, the thickness was not found uniform amongst the different patches.
Martin et al. [81] used X-ray absorption fine structure (XAFS) spectroscopy to show
that the ZDDP wear particles have a continuously random amorphous structure, as
depicted in Fig. 6.8. This structure was hypothesised to be formed due to friction-
and shear-induced atomic-scale mixing processes at the interface [81] in addition to
the digestion of the iron oxide by the phosphate layers near the metal surface [8].
Furthermore, it could be formed due to thermal effects by the interface quenching
[81].
Later on, the tribofilm was identified of having a multilayer structure [1, 6, 9,
153–155]. The layers close the substrate appear to be solid and adhere strongly to the
rubbed surface. However, the outer layers are viscous or semi-solid of hydrocarbons
and organic radicals, which adhere weakly to the lower layers and hence they can be
removed easily by rinsing with a solvent [6, 153, 155]. The concentrations of these
products decrease along the depth whereas the concentrations of iron, iron oxide and
iron sulphide increase. Above the iron sulphide/oxide layer, a layer of amorphous
polyphosphate glass was identified, which can contain zinc oxide and zinc sulphide,
as depicted in Fig. 6.9. The mechanical and physical properties of this amorphous
3D network are predetermined by the Zn2+ cations and organic radicals. However,
the layers near the metal surface are mainly affected by Fe2+ and Fe3+ cations. Bell
et al. [6] reported that the detected phosphate was similar to P2 O5 glass. Therefore,
they proposed that the ZDDP tribofilm consists mainly of glassy polyphosphate that
has different physical and mechanical properties along the depth of the tribofilm. In
addition, they noticed that the molar ratio of M2 O/P2 O5 increased from 1.2 at the
6 Tribochemistry and Morphology of P-Based Antiwear Films 189

Fig. 6.9 Schematic of the structural model of the ZDDP tribofilm elucidating its multilayer char-
acteristic. Reprinted from Bell et al. [6]

Fig. 6.10 TEM Cross section image of the ZDDP Tribofilm layers and EDX semi-quantitative
analysis of the elements constituting the different layers. Reprinted from Ito et al. [103]

outer surface to about 2 in the bulk of the tribofilm. This suggested that the outer layers
have longer chain length than those in the bulk. This also indicated that the outer
layers have less concentration of zinc. Furthermore, recent experiments conducted
by Ito et al. [103, 104] showed that the concentration of zinc is much higher near the
metal surface, as shown in Fig. 6.10. This indicates that the cation exchange reaction
between zinc and iron at high temperature should have occurred during the initial
stage of the ZDDP decomposition. This also suggests that the presence of zinc near
the metal surface plays a significant role in the observed short phosphate chains near
the substrate.
Yin et al. [1] noticed that the XANES results of the Total Electron Yield (TEY),
which is surface sensitive, and Fluorescence Yield (FY), which is bulk sensitive,
were different. This suggested that the tribofilm has a layered structure consisting
of a layer of short phosphate chains laid with a layer of long phosphate chains. Bec
et al. [155] used the Surface-Force Apparatus (SFA) to determine the structure of
the ZDDP tribofilm. The authors reported a full schematic of the ZDDP tribofilm
190 A. Dorgham et al.

before and after washing with solvents as shown in Fig. 6.11. The tribofilm appeared
to be patchy and every patch has several layers. Using the AFM, Warren et al.
[156] and Graham et al. [157] further confirmed that the tribofilm has a patchy-like
structure. In addition, they observed stripes of discontinuous ridges along the wear
scar. Pidduck and Smith [158] also observed that these patches were elongated in the
direction of sliding. The estimated film thickness was in the range of 100–140 nm,
which is in agreement with the recent measurements using the Spacer Layer Imaging
Method (SLIM) of the Mini-Traction Machine (MTM) [159]. In addition, the results
of Bec et al. [155] indicated that the tribofilm thickness and friction coefficient change
substantially from one position to another. This suggests that these parameters should
be considered as local properties rather than as averaged values.
Aktary et al. [160] followed the topography evolution of ZDDP thermal films
using Atomic Force Microscopy (AFM). During the first two hours, the structure
appeared as isolated islands of phosphate precipitates, as confirmed by the Fourier
Transform Infrared spectroscopy (FTIR). However, after 3–6 h, these islands coa-
lesced and the thermal film became smooth and continuous. The thickness of the film
increased linearly with immersion time and continued to increase even after 6 h dur-
ing which it reached an average thickness of 420 nm. The results indicated that there
is some correlation between the structure, thickness and morphology of the thermal
film. The FTIR results showed that the early formed thin film of isolated islands
consists of shorter phosphate chains than the mature thick and continuous film. The
follow-up study of Aktary et al. [70] showed that the ZDDP tribofilm evolves in the
same way as the thermal film. The only difference found was that after long rubbing
time, the continuous film disintegrates to form very small pads. The results of Can-
ning et al. [161] confirmed that the tribofilm is largely heterogeneous in the lateral
direction. The tribofilm was found to have ridges of mainly long phosphate chains
and troughs of mainly short phosphate chains. Same conclusions were also reached
by Nicholls et al. [20, 28, 80, 162] using X-ray Photoemission Electron Microscopy
(XPEEM) and AFM. The results of these studies showed that the ZDDP tribofilm
is also heterogeneous along the depth. The large pads of the tribofilm were found to
have a multilayer system where long polyphosphate chains lay on the top of short
polyphosphate chains. Based on these results amongst others, Spikes [21] proposed
that the ZDDP tribofilm has a structure similar to one depicted in Fig. 6.12. This
structure is in line with the findings of the previous studies that the tribofilm consists
of multilayer structure, which is largely heterogeneous in the lateral direction and
along its depth. This is also similar to the structure of DDP tribofilms, which initially
appears less patchy than the ones of ZDDP but eventually the two provide a similar
uniform pad-like structure [25, 26]. Distinctively though, the DDP additive forms Fe
phosphate of predominantly short chains as opposed to the initially zinc phosphate
of short chains that grow into longer ones away from the metal surface in the case
of ZDDP [25, 66]. Furthermore, the formation of iron sulphides or sulphates in the
DDP tribofilms near the metal surface depends greatly on the operating conditions
[84], as discussed in detail in the previous sections.
6 Tribochemistry and Morphology of P-Based Antiwear Films 191

Fig. 6.11 Schematic of the multilayer structure of the ZDDP tribofilm a before and b after washing
with solvent. Reprinted from Bec et al. [155]

Fig. 6.12 Schematic of the patchy structure of the ZDDP tribofilm. Reprinted from Spikes [21]
192 A. Dorgham et al.

The structures proposed by Bec et al. [155], Martin et al. [9] and Spikes [21] seem
to be the most accepted picture of the structure of P-based antiwear tribofilms up to
date.

6.5.2 Hardness and Elastic Modulus

Due to the largely heterogeneous nature of the ZDDP or DDP antiwear films, whether
tribofilms or thermal films, probing their mechanical properties is a complex and
intricate task. The main difficulties are attributed to the error brought by averaging
over a large area and propagated by the different uncertainties associated with every
experimental technique. This necessitates the need to analyse a large area of the
sample at a high lateral resolution and at different length scales. Measuring the
mechanical properties at a small length scale became feasible with the advent of the
AFM, interfacial force microscope (IFM) and nanoindentation.
Many studies have already used the AFM to study the mechanical properties
of the ZDDP tribofilms [70, 77, 156, 157, 160, 163–166] as well as tribofilms of
other additives [167, 168]. This enormous work concluded that the layers close to
the metal surface are most likely elastic solid whereas the outer layers are probably
viscous. This indicates that the layers in between can have a combination of the two.
Furthermore, Warren et al. [156] observed stripes of discontinuous ridges along the
wear scar. These stripes were speculated to originate from repetitive sliding, which
suggested that the ridges were the only part of the tribofilm that carry the load. In
agreement with these results, the experiments of Graham et al. [157] showed that
these ridges were elongated in the sliding direction. Aktary et al. [70] attributed this
heterogeneity in the tribofilm growth to the local variations in the contact pressure
between the contacting surfaces. However, as the tribofilm is not heterogeneous only
in the lateral direction but also along its depth, this indicates that the heterogeneity
of the tribofilm is not only due to the local variations in the contact pressure [77].
Warren et al. [156] proposed three possibilities for the patchy structure of the P-based
tribofilms, which are as follows:
1. Ridges are formed due to the high contact pressure between the asperities whereas
the troughs are formed due to the three body interaction between the wear debris
and the contacting surfaces.
2. Ridges are formed due to the repeated rubbing between the asperity contacts at
the same average locations whereas the troughs are formed due to occasional
contact between the asperities at the troughs regions.
3. Ridges are formed due to the repeated rubbing between the asperity contacts
whereas the troughs are originated from the flow of some material from the
ridges to troughs’ regions.
The authors suggested that the way the ridges and troughs appeared indicated that
the third mechanism is likely to be the case. Other possibilities to account for the
6 Tribochemistry and Morphology of P-Based Antiwear Films 193

heterogeneity in the tribofilm growth have also been proposed recently by Gosvami
et al. [40] to be due to the following:
1. Heterogeneous random nucleation sites due to the surface roughness or any sur-
face defects.
2. Instabilities in the decomposition and growth mechanisms due to any variations
in the operational conditions.
These possibilities are plausible and can be combined with the ones proposed by
Warren et al. [156] to give a wider understanding.
To quantify the mechanical properties of the ZDDP or DDP tribofilms, several
studies attempting at measuring the elastic modulus using nanoindentation, AFM and
SFA. Table 6.3 summarises the indentation modulus that was reported for the ZDDP
and the measurement conditions. Aktary et al. [70] measured the elastic modulus
of the ZDDP tribofilm after different rubbing times and found that the modulus and
hardness of the tribofilm are independent of the rubbing time. They explained this
result by pointing out that the majority of the bulk material of the tribofilm at any
time consists mainly of short polyphosphate chains. Therefore, the nanomechani-
cal properties are predetermined by the properties of this bulk material. However,
Nicolls et al. [162] reported that the large pads of the tribofilm consist mainly of long
polyphosphate chains whereas the small pads and troughs consist of short chains.
Graham et al. [157] showed that the pads have a great elastic response. The measured
indentation modulus was about 180–250 GPa at the centre of the pad compared with
50–110 GPa at the edge. Nicolls et al. [80] found that the indentation modulus was
120 GPa at the centre of the pad compared with 90 GPa at the edge. In a later study,
Nicolls et al. [162] reported an indentation modulus of 80 GPa for the large pads but
could not measure the modulus of the small pads due to the effect of the substrate
on the indentation measurements of these thin regions.
This wide range of values of the reported indentation modulus could be related
to the extent of the additive decomposition to form polyphosphates [20] in addition
to the intrinsic uncertainties associated with probing such thin heterogeneous films
using any indentation technique.
In contrast to the aforementioned studies, which suggested that the tribofilm is
highly elastic, Aktary et al. [70] found that the ZDDP tribofilm exhibits high plasticity.
They suggested that this is in line with the sacrificial nature of the tribofilm, which
is continuously formed and removed at the interface. Conversely, Warren et al. [156]
showed that the ridges of the tribofilm exhibit a great capacity of elastic deformation,
which was also confirmed by the results of Graham et al. [157] and Ye et al. [163]
showing that the ridges of the ZDDP tribofilm resist plastic deformation to a great
extent.
Ye et al. [77, 163, 164] studied various tribofilms formed in oils containing ZDDP
and ZDDP/MoDTC additives. In the two cases, the results showed a gradual increase
of hardness and modulus over depth until they reach the ones of the substrate. This was
considered as evidence that the tribofilm consists of a multilayer system of a hard layer
covered by a softer one. Similarly, Bec et al. [155] showed that the ZDDP tribofilm
resists indentation by increasing hardness and elastic modulus with the indentation
194 A. Dorgham et al.

Table 6.3 Elastic modulus and hardness of the ZDDP thermal and tribo-film
Material Elastic Method Lubricant Test References
modulus conditions
(GPa)
5 min 88.5 ± 23.7 Nanoindentation 1.49 wt% Plint [70]
tribofilm
10 min 92.8 ± 18.6 ZDDP 100 C
tribofilm
40 min 88.6 ± 29.9 In MCT-10 225 N
tribofilm
1 h tribofilm 88.9 ± 12.1 25 Hz
2 h Tribofilm 96.1 ± 25.7
3 h thermal 34.8 ± 9.7
film
1 h tribofilm 87.8 ± 3.9 Nanoindentation 1.2 wt% Plint [80]
Edge of pad ZDDP 100 C
1 h tribofilm 119.5 ± 5.8 In MCT-10 220 N
Centre of pad 25 Hz
Sulphide 90 SFA Amsler [155]
oxide
t ≤ 80 nm 100 C
Polyphosphate 15 400 N
20 < t < 30 nm 5h
Polyphosphate 27–30
70 < t 40
< 100 nm
Polyphosphate
t > 140 nm
6 h tribofilm 25 IFM 1.2 wt% Plint [156]
Trough ZDDP 100 C
6 h tribofilm 81 In paraffinic 220 N
Ridge Base oil 25 Hz
1 h thermal 36 ± 9 200 C
film
Large pads 209 ± 38 IFM 5 mM/kg Plint [157]
Ridges ZDDP 100 C
Large pads 87 ± 23 In paraffinic 225 N
Troughs Base oil 25 Hz
Small pads 74 ± 20
Off pads 37 ± 7.3
1 h tribofilm 85.1 ± 11.1 Nanoindentation 1.2 wt% Plint [169]
And IFM ZDDP 100 C
In MCT-10 220 N 25 Hz
1 h tribofilm 80.5 ± 4.5 Nanoindentation 1.2 wt% Plint [80]
(continued)
6 Tribochemistry and Morphology of P-Based Antiwear Films 195

Table 6.3 (continued)

Material Elastic Method Lubricant Test References
modulus conditions
(GPa)
Ridges ZDDP 100 C
1 h tribofilm 30 In MCT-10 220 N
Troughs 25 Hz
10 min 78–110 ± 8 Nanoindentation 1.2 wt% Plint [162]
tribofilm
And IFM ZDDP 60 C
1 h tribofilm 97–110 ± 8 In MCT-10 60 N
25 Hz
1 h tribofilm 80–215 Nanoindentation Pin-on- [163]
Depends on ZDDP in Disc
Contact 5 W–30 SG 490 N
Depth Engine oil 0.03 m/s

depth. Therefore, they suggested that the heterogeneity of the mechanical properties
of the tribofilm along its depth is due to work hardening but not a real intrinsic
property. However, in contrast to these results, the loading-unloading experiments of
Warren et al. [156] using AFM indicated that the tribofilm ridges have two layers of
which the one that is more compliant lay beneath the stiffer surface layer. The base
layer seemed to exhibit a significant tenacity, which was manifested in the apparent
adhesion during the retraction of the AFM tip from the sample. These features were
neither observed in the troughs of the tribofilm nor in the thermal film. The mechanical
properties of the trough region of the tribofilm seemed identical to the ones of the
thermal film. Based on these results, Warren et al. [156] suggested that the thermal
film is formed first as a precursor to the tribofilm. In line with these results, Kim et al.
[22] showed that the surface layer of the ZDDP tribofilms, away from the substrate,
is harder than the bulk layers. In contrast, DDP tribofilms showed a more compliant
tribofilm without the presence of a hard surface layer. They suggested that the hard
crust protects against wear while the compliant bulk helps dissipate energy. This is
despite the fact that they observed that the more compliant but thicker DDP tribofilm
showed a much better tribological performance than the one of ZDDP. To resolve
this discrepancy, they suggested that the effective coverage of the DDP tribofilm is
higher than the ZDDP one leading to better protection against wear.
Based on the discussion above, it can be concluded that the ZDDP and DDP
antiwear films have rich mechanical properties. These properties can be affected by
the decomposition mechanism, load and temperature. In addition, the mechanical
properties might also be affected by the rheological properties of the tribofilm. For
instance, in case of the tribofilm is viscoelastic then its mechanical response can
look similar to the plastic behaviour of a compliant material [170]. Moreover, a
viscoelastic tribofilm may suggest that its mechanical properties are rate dependent.
196 A. Dorgham et al.

Hence, the history and rate of measurement can play a major role in the measured
properties. Therefore, studying the rheological properties of the antiwear tribofilm is
necessary to give insight into its mechanical as well as tribological properties. These
rheological properties will be discussed in detail in the subsequent section after the
tenacity and durability of the tribofilm is reviewed.

6.5.3 Tenacity and Durability

In order to study the tenacity and durability of ZDDP tribofilms, Bancroft et al. [2]
examined the effect of the ZDDP concentration in oil on the tribofilm that has already
been formed. After the formation of the tribofilm, the base oil containing ZDDP was
replaced with oil without ZDDP and then the rubbing continued for extended periods
of 6–24 h. The tribofilm showed a great thermal and mechanical stability even after
rubbing for periods as long as 24 h. In addition, the results showed that rubbing
the tribofilm in base oil has two main effects. The first one is that the tribofilm
maintains a certain thickness above 30 nm without being completely removed by
rubbing. The second effect is that the polyphosphate chains become shorter in the
form of orthophosphate and pyrophosphate. Ancillary experiments of Suominen
Fuller et al. [79] showed that when the oil of mature ZDDP tribofilm was replaced
by base oil without ZDDP, no change occurred in the tribofilm thickness with further
rubbing in the base oil. However, Parsaeian et al. [171] found that rubbing premature
tribofilm (after 25 min rubbing: before reaching steady state thickness) in base oil
without ZDDP causes an initial sharp decrease in the tribofilm thickness after the
first few minutes of rubbing. The reduction was about 20–70 nm depending on load
and temperature, i.e. the higher the load or temperature the higher the removal.
The sharp decrease in the tribofilm thickness was followed by a steady state period
during which the tribofilms maintained its thickness without any further removal.
Adding fresh oil containing ZDDP again after 60 min of total rubbing time (35 min
of rubbing in base oil) results in a fast growth of the tribofilm thickness similar
to the initial growth rate before replacing the oil with base oil. On the other hand,
rubbing mature tribofilm in base oil (after 180 min rubbing: after reaching steady
state thickness) results in just a small decrease in the tribofilm thickness. As indicated
before, increasing temperature or load after replacing oil with base oil resulted in
more immediate removal of the tribofilm. However, temperature and load did not
show any monotonic trend on the steady state thickness. The drop in the tribofilm
thickness was investigated using XPS, which showed that before replacing the oil
with base oil the top layer of the tribofilm contains longer phosphate chains. This
layer seems to be the one removed after replacing the oil, which suggests that it is
softer and less tenacious than the layers underneath it. It was also found using XPS
analysis that if rubbing is continued again by adding fresh oil containing ZDDP, the
removed long phosphate chains can be formed again.
6 Tribochemistry and Morphology of P-Based Antiwear Films 197

6.6 Rheological Properties of Antiwear Films

The friction, lubrication and adhesion properties of any tribological surface are
greatly affected by the rheological properties of the interface [172]. These prop-
erties can undergo changes when using additives such as ZDDP or DDP, which can
form a protective tribofilm of transient thickness covering the interface. In order to
quantify these changes, the rheological properties of the thin antiwear film should
be measured. This can be mainly achieved using two methods. The first one is by
generating a thick tribo- or thermal film that can be scratched and removed for ex
situ analysis using the bulk or interfacial rheometry. The second possible method is
by measuring these properties in situ. The in situ rheological measurements of the
antiwear tribo- or thermal films or generally speaking any thin film on a substrate are
performed, in essence, in the same way as the bulk rheological measurements, e.g.
creep and shear, but using different experimental techniques such as SFA and AFM.
For instance, Georges et al. [173] used SFA to measure the rheological and mechan-
ical properties of the ZDDP physisorbed films by imposing an oscillatory motion in
three different directions, i.e. x, y and z axes. This enabled them to measure the damp-
ing coefficient, normal stiffness and lateral stiffness at different separation distances
between the ball and disc. They could then relate the measured damping coefficient
(Aω) at a certain oscillatory frequency (ω) and a separation distance (D) to viscosity
(η) using the following relation:

6π η R 2 ω
Aω  (6.30)
D − 2L h

where R is the ball radius and L h is the total thickness of any adsorbed layers on each
surface. Good agreement between the experimental data and the damping coefficient
given by the previous equation was found as shown in Fig. 6.13. One interesting
result is the approximately exponential increase in the viscosity at small separa-
tion distances, which has the effect of increasing the viscous resistance to sliding.
Another important finding is that the elastic compressive and shear moduli appeared
to increase with reducing the separation distance. The authors related this observa-
tion to a possible compaction of the heterogeneous physisorbed layers. In agreement
with these results, Bec et al. [155] suggested that the contact pressure could compact
the loose layers of the ZDDP tribofilm and transfer them into a solid polyphosphate.
Bec et al. [155] used SFA to measure the mechanical and rheological properties of
the ZDDP tribofilm. The results showed that the sulphide and phosphate layers exhibit
elastoplastic properties. This suggests that the ZDDP tribofilm can be polymeric in
nature [51] and its rheological properties, similar to the mechanical properties, vary
along the depth of the tribofilm [6]. This was confirmed by Pidduck and Smith [158]
and Bec et al. [155] who reported that the base layer of sulphide resists flow signif-
icantly whereas the top layer of alkyl phosphate precipitates behaves as a viscous
polymer, which does not resist flow and could be easily removed. Similarly, the bulk
layers of polyphosphates forming the ridges of the tribofilm can also be removed
198 A. Dorgham et al.

Fig. 6.13 Comparison between the measured and predicted damping coefficient of the basic and
neutral ZDDP. Reprinted from Georges et al. [173]

easily leaving trough regions of bare sulphide/oxide layers. This suggested that the
polyphosphate layers in the trough originated from the ridges material, i.e. alkyl
phosphate precipitates and polyphosphates, that flowed into trough regions by shear
flow. Bec et al. [155] reported a viscosity of 5 × 104 to 3 × 105 Pa s for the alkyl
phosphate layer and 108 Pa s for the polyphosphate layer.
Based on the aforementioned discussion, it can be concluded that the P-based
antiwear films have rich rheological properties. However, no elaborate study was
conducted so far to unravel all these properties and to provide conclusive insights
into the relation between the rheological, mechanical and tribological properties of
the ZDDP or DDP tribo- and thermal films.

6.7 Tribological Properties of Antiwear Films

6.7.1 Friction Mechanism

Many studies aimed at understanding the genesis of friction of the P-based tribofilms
using different techniques. For example, Ye et al. [163] used the nanoscratch method
combined with AFM imaging and observed that the tribofilm exhibits different fric-
tion coefficients depending on the contact depth, which indicated that different levels
of shear strength exist within the tribofilm. The lowest friction coefficient found was
at few nanometres beneath the surface, which was attributed to the presence of an
ultra-low friction inner skin layer that can act as a type of solid lubricant to reduce
friction. In another study, Ye et al. [166] showed that the inner skin layer exhibits
a low shear modulus and can yield easily, which can explain the capability of this
layer to reduce friction. The heterogeneous friction behaviour along the depth is also
accompanied by a heterogeneity along the surface as reported by Neitzel et al. [174]
who observed that friction is different at different length scales. For instance, within
a material exhibiting a high friction coefficient, areas of low friction at the nano- or
6 Tribochemistry and Morphology of P-Based Antiwear Films 199

Fig. 6.14 Effect of the

ZDDP tribofilm on the
Stribeck curve after different
rubbing periods. Reprinted
from Taylor and Spikes [175]

microscopic scale might exist. These small domains of low friction force can affect
the macroscopic friction or induce slip.
Taylor et al. [159] reported that ZDDP forms a thick solid-like film on the wear
scar that produces an effective surface roughening. This increased roughness can
inhibit the entrainment of the fluid film between the contacting surfaces and thus
results in a higher friction than in the case of surfaces not covered with the ZDDP
tribofilm, as shown in Fig. 6.14. However, the results of Taylor and Spikes [175]
showed that even smooth ZDDP tribofilms can increase friction. This suggested that
whether the ZDDP tribofilm is rough or smooth, it can inhibit the entrainment of
the lubricating film between the rubbing contacts and hence increases friction. This
apparent increase in friction was also suggested to represent a shift in the Stribeck
curve to a higher speed, which means that the tribofilm is capable of maintaining the
boundary lubrication condition up to a higher speed than in the case of bare contacts
without a tribofilm.
A different explanation for the increase in friction force when the ZDDP tribofilm
is present on the steel surface was provided by Suarez et al. [137]. Assuming that
the friction force can be given by the following equation:

FF  τZDDP AZDDP + τsteel Asteel (6.31)

where τ ZDDP and τ steel are the mean shear strength of the ZDDP tribofilm and steel
surface, respectively, and AZDDP and Asteel are the real contact area of the ZDDP
tribofilm and steel surface, respectively. Therefore, increasing the tribofilm thickness
increases the contact area (AZDDP ) and thus increases the friction force.
The increased friction of the antiwear tribofilms over time can also be explained
based on the observations of Mazuyer et al. [176] that after a certain threshold of con-
tact time the layers covering the rubbing surfaces can interact and hence the mechan-
ical properties, e.g. shear elastic modulus and interfacial shear strength, become a
function of the contact time. In the case of the ZDDP tribofilm, this means that the
longer the rubbing time, the thicker the tribofilm will be and thus the stronger the
200 A. Dorgham et al.

Fig. 6.15 Schematic of the ordered structure of the compressed hydrodynamic layer. Reprinted
from Georges et al. [177]

interactions can occur between the tribofilms covering the contacting surfaces, which
leads to higher friction.
Georges et al. [177] highlighted the steric action of the adsorbed layers on friction.
When the separation distance is small, the pressure is high. In this case, the fluid starts
to have an ordered structure in the direction of shear, as shown in Fig. 6.15. This was
inferred from the observation that the stabilised friction coefficient was found to be
inversely proportional to the sliding speed. In addition, Li et al. [178] showed that
the strength of adhesion of the interfacial layers adsorbed on a substrate has great
influence on friction. In the case of a single layer that adhered loosely, friction was
maximum due to the puckering effect of the layer when probed by the AFM tip.
On the other hand, in the case of a bulk material with strongly adhered layers, this
puckering effect is suppressed and the thickness does not have any effect on friction
after a certain threshold value. Ancillary results of Lee et al. [179, 180] suggested
the same trend for different materials including MoS2 .
The fact that sulphur might be the first to reach the metal surface is very crucial
in order to relate the mechanism of additive decomposition and tribofilm formation
to friction and wear. In the case of sulphur containing deposits are formed initially
at the metal surface, then we can infer two main points. Firstly, during the running-
in period, the sulphur rich layer is the one that predetermines the observed friction
and wear. Therefore, if the porous FeS2 is the main component of this layer, which
is suggested in the case of surfaces covered with Fe2 O3 or Fe3 O4 as shown in the
potential diagram in Fig. 6.3, friction will be small. However, a larger friction is
expected in the case of a base layer consisting of a dense and uniform FeS, which is
likely to be formed on bare iron [106]. Furthermore, the existence of a sulphur rich
layer as a base layer on the metal surface indicates that the subsequently formed layers
6 Tribochemistry and Morphology of P-Based Antiwear Films 201

should be deposited on this layer, whether fully or partially, rather than completely
on the metal surface itself.

6.7.2 Antiwear Mechanism

The ZDDP additive when decomposed under high contact pressure or heat can form
a superior antiwear tribofilm on contacting surfaces. The DDP additive can provide
better [22, 23], comparable [24] or worse [25, 26] antiwear properties to the ones
of ZDDP, which seem to depend greatly on the operating conditions, properties of
contacting surfaces and chemistries of base oil and additives.
Since the first inception of the ZDDP in the late 1930s, extensive studies on the
genesis of its antiwear mechanism have been carried out [21]. Table 6.4 summarises
the several theories and mechanisms that were proposed to explain the capability of
the formed antiwear tribofilms in protecting the contacting surfaces from severe wear.
These different mechanisms can be categorised depending on whether the cause of
protection is rheological, mechanical or chemical.
Initially, it was suggested that the antiwear mechanism of antiwear additives is
based on forming a thicker hydrodynamic film. This thick film can reduce the stress
at the contacting asperities and hence reduces wear [7]. The sulphides, phosphorus
compounds and oxides were also found to have good lubricating properties that help
mitigate wear [181]. Molina [45] suggested that the ratio between the crystalline and
amorphous regions of the antiwear film might be important in evaluating the antiwear
action of the tribofilm. Molina’s results indicated that the amorphous pyrophosphate
has better lubricating properties that the crystalline phosphate. In line with these
results, it was suggested that the polyphosphate, which has a low melting temper-
ature of 200–300 °C, melts and forms a viscous glass on the contacting surfaces
that helps reduce wear [3]. In addition, the interfacial iron-oxide-sulphide eutectic
system was also proposed to form a viscous film at the contacting surfaces under
extreme conditions and hence separates the contacting surfaces and reduces wear
[3, 5]. Same mechanism was also proposed for the FeS layer covering the asperity-
asperity contacts [6]. Interestingly, the amount of sulphur in the tribofilm was found
to increase with increasing load [7], which highlights the smart action of ZDDP to
reduce wear. This smart action was explored by So and Lin [154] who indicated that
the plastic deformation is responsible for increasing the temperature at the asperities
and hence the decomposition of the additive. Additionally, this plastic deformation
creates subsurface defects that enhances the mixing and reaction of the decomposi-
tion products, i.e. P, S, Zn and O, with the rubbing surface. Accordingly, So et al.
[44] suggested that the antiwear mechanism depends closely on the ratio between
the tribofilm formation and its removal. In addition, Habeeb and Stover [11] reported
that the ZDDP tribofilm can also reduce wear by decomposing the peroxides. On the
other hand, other studies [8–10] suggested that the antiwear action and the absence
of severe abrasive wear is due to the digestion of the iron oxide generated during
wear into the amorphous phosphate structure of the formed interfacial tribofilm.
202 A. Dorgham et al.

Table 6.4 Summary of the proposed antiwear mechanisms of ZDDP

Authors Mechanism Details
Watkins [3] Rheological and mechanical Formation of a viscous glass
on the contacting surfaces that
flows and acts as a barrier
Gansheimer [181] Rheological and mechanical Good lubricating properties of
sulphides, phosphorus
compounds and oxides
Rounds [4] Rheological and mechanical Formation of oriented
sacrificial layer that reduces
shear stresses and acts as a
barrier
Bec et al. [155] Rheological and mechanical Formation of a grease- or
gel-like hydrocarbon-rich
layer that lias low shear
strength
Watkins [3], Glaeser et al. [5] Mechanical Formation of an interfacial
iron-oxide-sulphide eutectic
system as a barrier
Bell et al. [6] Mechanical Formation of FeS on the
surface as a barrier
Bell et al. [6] Mechanical Formation of a low shear stress
layers that reduce fatigue wear
and delamination processes
So et al. [44] Mechanical Balance between the tribofilm
formation and its removal
So and Liu [154] Mechanical Formation of a compliant
protective tribofilm
Palacios [7] Mechanical Formation of a thick
hydrodynamic film as a barrier
Habeeb and Stover [11] Chemical Decomposition of peroxides
Belin et al. [10], Martin et al. Chemical Digestion of sharp wear
[8, 9] particles of iron oxide
Molina [45] Chemical Formation of an amorphous
glass rather than a crystalline
one

Bell et al. [6] suggested that a layer exists in the tribofilm with a low shear stress
that helps reduce fatigue wear and delamination processes. In addition, they proposed
that the adhesion between the different layers is expected to be crucial for the load
carrying capability of the tribofilm and other wear and friction performance. On the
other hand, Rounds [4] proposed that the P-based additives such as ZDDP can act by a
mechanism similar to the one of the fatty acid by forming an oriented sacrificial layer
covering the surface. This antiwear capability depends on the formed hydrocarbon
chain length and chain branching. Accordingly, the hydrocarbon-rich layer at the
outermost of the tribofilm seems to play a role in reducing wear [155]. This layer
6 Tribochemistry and Morphology of P-Based Antiwear Films 203

Fig. 6.16 Effect of water on the life of bearing based on 100% life at 0.01% water in addition to the
available techniques to detect water and their range are indicated. Reprinted from Sheehan [185]

seemed to have a grease or gel-like nature that has a low shear strength, which helps
in redistributing the concentrated loads at the asperities and therefore reduces the
high shear stresses at the asperity-asperity contacts and thus mitigates wear.
In summary, the unique antiwear properties of the ZDDP or DDP tribofilms seem
to be related not only to its adhesion, hardness and elasticity but also to the capability
of the formed tribofilm to maintain local order on the molecular scale through the
flow, rearrangement and change in composition of the interfacial layers.

6.8 Antiwear Films in Humid Environments

Corrosion phenomena typically involves water and oxygen [182], which are abundant
at any interface. This makes water one the main elements in the corrosive environ-
ment. Water can affect the bearing performance in different ways [183]. On the one
hand, it can shorten the bearing life, as shown in Fig. 6.16, due to rust, hydrogen
embrittlement and oxidation. On the other hand, water can accelerate wear due to the
oxidation of the lubricating film and the destabilisation of the protective tribofilm.
Therefore, the relative degree of saturation of water in the oil is one of the important
factors to consider. Water can enter the oil from the humid air or directly as a free
water. The temperature of the oil determines whether this water will mix with the
oil as a dissolved water, i.e. below the saturation temperature, or as a free water, i.e.
above the saturation temperature [184].
In the following subsections, a review is provided on the effect of water on the
composition and tribological properties of the P-based antiwear film.
204 A. Dorgham et al.

6.8.1 Effect of Water on the Composition of Tribofilms

Rounds [72] showed that water accelerates the ZDDP decomposition and reduces S,
P and Zn contents in the antiwear film. These changes are additive dependent and
can be related to the formation of acidic hydrolysis products. In contrast, the results
of Nedelcu et al. [144] showed that water inhibits the growth of the ZDDP tribofilm.
This effect was manifested in the formation of shorter chains of polyphosphates.
The authors attributed these effects to the depolymerisation reactions of the long
polyphosphate chains and to the increased surface distress in the presence of water.
Ancillary experiments carried out by Cen et al. [143] showed that water indeed
inhibits the formation of the protective tribofilm. In addition, they noticed that shorter
phosphate chains are formed with increasing the amount of water in the oil.
In line with these results, Parsaeian et al. [186] found that the smaller the relative
humidity the thicker the tribofilm and the longer the formed phosphate chains. The
trend holds true for the two tested temperatures, i.e. 80 and 98 °C. This was explained
by the difficulty of ZDDP molecules to reach and react with the steel surface in an oil
containing higher amount of free or dissolved water. In another study, Parsaeian et al.
[148] studied the effect of water contamination in oil, i.e. 0, 0.5, 1.5 and 3 wt%, on
wear and tribofilm thickness. The results showed that the larger the water content is
present in oil the thinner the tribofilm thickness. Similar findings were also reported
by Faut and Wheeler [187] for tricresylphosphate (TCP) additive.
All these results highlight the fact that the presence of water affects not only the
rate of the additive decomposition but also the final composition of the protective
film. Owing to the high affinity of phosphates to water [6], when water contamination
occurs, some water molecules will react with the phosphate and replace ZnO, whereas
some amount of water will stay unreacted. This will affect the ratio of ZnO/ P2 O5
and thus the phosphate chain length. Furthermore, the presence of unreacted water is
expected to affect this ratio up or down depending on the localisation of water, i.e. at
the metal surface or in the bulk. Therefore, to better understand the effect of water on
the composition of ZDDP or DDP antiwear film, the focus should be given to trace
the composition without water and study how water can change this morphology.

6.8.2 Effect of Water on the Tribological Properties

of Tribofilms

There are only few studies [143, 144, 148, 149, 186] on the effect of water on the fric-
tion and wear performance of the P-based antiwear film, i.e. mainly of ZDDP. These
studies mainly showed that water inhibits the formation of the protective tribofilm
and increases wear. The authors attributed these effects to the depolymerisation reac-
tions of the long polyphosphate chains and to the increased surface distress in the
presence of water. Other reasons for the increased wear in the presence of water in the
6 Tribochemistry and Morphology of P-Based Antiwear Films 205

oil can also be related to corrosion, hydrogen embrittlement and accelerated fatigue
due to the possible condensation of small amount of water in the microcracks.
Cen et al. [143] found that the effect of water on friction is less prominent than the
case of wear. However, although their results showed that up to a moderate amount of
water, friction is not affected, the experiments conducted in a humid atmosphere of
a relative humidity 90%, showed that the friction coefficient drops by about 40%. It
should be noted that this behaviour of friction can be related to the one of wear. Due
to the larger wear in the case of large water contamination, surface smoothing can
occur during the induction period that can lead to a reduction in friction coefficient.
In addition, this reduction can also be related to the formation of thinner tribofilm of
short polyphosphate chains when high water concentration is present in the oil.
These scarce results highlight the importance to study the effect of water on the
decomposition of antiwear additive and the final composition. Understanding the
effect of water on these processes helps control the wear and friction performance by
just controlling the amount of water in the oil and its evaporation rate. In addition,
understanding the effect of water provides insight into the reaction kinetics of the
ZDDP or DDP tribofilm formation and its behaviour in different environments.

6.9 Concluding Remarks

Throughout this chapter an overview of the state-of-the-art knowledge of the dif-

ferent aspects concerning the P-based antiwear films of ZDDP and DDP additives
were presented and discussed. At first, a review was provided on the chemical and
physical nature of the antiwear film and the possible reactions during its formation.
It was shown that the P-based additives can decompose in different mechanisms,
i.e. thermally, hydrolytic or oxidative, to form a complex antiwear film consisting
mainly of Zn/Fe, P, S and O. The formation kinetics of these elements are still not
fully elucidated.
All the recent studies have confirmed that the formed P-based antiwear films have
a patchy layered structure of large and small pads. The large pads exhibit a great
elastic response. The elastic modulus, hardness and thickness of these pads are not
uniform but vary along the surface and depth. As these properties change substan-
tially from one position to another, this suggests that they should be considered as
local properties rather than averaged values. Thus, one of the future works should
focus on investigating the effect of the local parameters versus the average values
on the estimation of the friction and wear using the different models available in the
literature.
The P-based antiwear tribofilms help protect the contacting surfaces by different
mechanisms including the digestion of sharp particles and formation of sacrificial
rigid layers of good mechanical properties, e.g. elasticity and hardness. There is a
lack of studies, however, focusing on the role of the rheological properties of the
P-based tribofilms on their durability and antiwear capability.
206 A. Dorgham et al.

A systematic study taking into account the aforementioned factors seems to be

crucial, which should provide better insight into the origin of the superior antiwear
properties of ZDDP and DDP tribofilms in order to be able to optimise their properties
and eventually replace them with more environmentally friendly additives.

References

1. Z. Yin, M. Kasrai, M. Fuller, G.M. Bancroft, K. Fyfe, K.H. Tan, Application of soft x-ray
absorption spectroscopy in chemical characterization of antiwear films generated by ZDDP
part I: the effects of physical parameters. Wear 202(2), 172–191 (1997)
2. G.M. Bancroft, M. Kasrai, M. Fuller, Z. Yin, K. Fyfe, K.H. Tan, Mechanisms of tribochemical
film formation: stability of tribo- and thermally-generated ZDDP films. Tribol. Lett. 3, 47–51
(1997)
3. R.C. Watkins, The antiwear mechanism of ZDDP’s. Part II. Tribol. Int. 15(1), 13–15 (1982)
4. F.G. Rounds, Effects of additives on the friction of steel on steel I. Surface topography and
film composition studies. ASLE Trans. 7(1), 11–23 (1964)
5. W.A. Glaeser, D. Baer, M. Engelhardt. In situ wear experiments in the scanning auger spec-
trometer. Wear 162–164. Part A(0), 132–138 (1993). Wear of Materials: Proceedings of the
9th International Conference
6. J.C. Bell, K.M. Delargy, A.M. Seeney, Paper IX (ii) the removal of substrate material through
thick zinc dithiophosphate anti-wear films. Tribol. Ser. 21, 387–396 (1992)
7. J.M. Palacios, Thickness and chemical composition of films formed by antimony dithiocar-
bamate and zinc dithiophosphate. Tribol. Int. 19(1), 35–39 (1986)
8. J.M. Martin, Antiwear mechanisms of zinc dithiophosphate: a chemical hardness approach.
Tribol. Lett. 6(1), 1–8 (1999)
9. J.M. Martin, C. Grossiord, T. Le Mogne, S. Bec, A. Tonck, The two-layer structure of ZnDTP
tribofilms: part I: AES, XPS and XANES analyses. Tribol. Int. 34(8), 523–530 (2001)
10. M. Belin, J.M. Martin, J.L. Mansot, Role of iron in the amorphization process in friction-
induced phosphate glasses. Tribol. Trans. 32(3), 410–413 (1989)
11. J.J. Habeeb, W.H. Stover, The role of hydroperoxides in engine wear and the effect of zinc
dialkyldithiophosphates. ASLE Trans. 30(4), 419–426 (1986)
12. F. Rounds, Effects of hydroperoxides on wear as measured in four-ball wear tests. Tribol.
Trans. 36(2), 297–303 (1993)
13. P.A. Willermet, D.P. Dailey, R.O. Carter, P.J. Schmitz, W. Zhu, Mechanism of formation of
antiwear films from zinc dialkyldithiophosphates. Tribol. Int. 28(3), 177–187 (1995)
14. M.N. Webster, C.J.J. Norbart, An experimental investigation of micropitting using a roller
disk machine. Tribol. Trans. 38(4), 883–893 (1995)
15. C. Benyajati, A.V. Olver, C.J. Hamer, An experimental study of micropitting, using a new
miniature test-rig. Tribol. Ser. 43, 601–610 (2003)
16. V. Brizmer, H.R. Pasaribu, G.E. Morales-Espejel, Micropitting performance of oil additives
in lubricated rolling contacts. Tribol. Trans. 56(5), 739–748 (2013)
17. E. Lainé, A.V. Olver, T.A. Beveridge, Effect of lubricants on micropitting and wear. Tribol.
Int. 41(11), 1049–1055 (2008)
18. J. Andersson, M. Antonsson, L. Eurenius, E. Olsson, M. Skoglundh, Deactivation of diesel
oxidation catalysts: vehicle and synthetic aging correlations. Appl. Catal. B Environ. 72(1),
71–81 (2007)
19. C. Larese, F. Cabello Galisteo, M. López Granados, R. Mariscal, J.L.G. Fierro, M. Furió, R.
Fernández Ruiz, Deactivation of real three way catalysts by CePo4 formation. Appl. Catal. B
Environ. 40(4), 305–317 (2003)
6 Tribochemistry and Morphology of P-Based Antiwear Films 207

20. M.A. Nicholls, P.R. Norton, G.M. Bancroft, M. Kasrai, T. Do, B.H. Frazer, G. De Stasio,
Nanometer scale chemomechanical characterization of antiwear films. Tribol. Lett. 17(2),
205–216 (2004)
21. H. Spikes, The history and mechanisms of ZDDP. Tribol. Lett. 17(3), 469–489 (2004)
22. B.H. Kim, R. Mourhatch, P.B. Aswath, Properties of tribofilms formed with ashless dithio-
phosphate and zinc dialkyl dithiophosphate under extreme pressure conditions. Wear 268(3),
579–591 (2010)
23. X. Fu, W. Liu, Q. Xue, The application research on series of ashless P-containing EP and AW
additives. Ind. Lubr. Tribol. 57(2), 80–83 (2005)
24. R. Sarin, A.K. Gupta, D.K. Tuli, A.S. Verma, M.M. Rai, A.K. Bhatnagar, Synthesis and
performance evaluation of o, o-dialkylphosphorodithioic disulphides as potential antiwear,
extreme-pressure and antioxidant additives. Tribol. Int. 26(6), 389–394 (1993)
25. M.N. Najman, M. Kasrai, G.M. Bancroft, Chemistry of antiwear films from ashless thiophos-
phate oil additives. Tribol. Lett. 17(2), 217–229 (2004)
26. Z. Zhang, E.S. Yamaguchi, M. Kasrai, G.M. Bancroft, Tribofilms generated from ZDDP and
DDP on steel surfaces: part 1, growth, wear and morphology. Tribol. Lett. 19(3), 211–220
(2005)
27. A.M. Barnes, K.D. Bartle, V.R.A. Thibon, A review of zinc dialkyldithiophosphates (ZDDPS):
characterisation and role in the lubricating oil. Tribol. Int. 34(6), 389–395 (2001)
28. M.A. Nicholls, T. Do, P.R. Norton, M. Kasrai, G. Michael Bancroft, Review of the lubrication
of metallic surfaces by zinc dialkyl-dithiophosphates. Tribol. Int. 38(1), 15–39 (2005)
29. D.R. Armstrong, E.S. Ferrari, K.J. Roberts, D. Adams, An examination of the reactivity of
zinc di-alkyl-di-thiophosphate in relation to its use as an anti-wear and anti-corrosion additive
in lubricating oils. Wear 217(2), 276–287 (1998)
30. L.R. Rudnick, Lubricant Additives: Chemistry and Applications (CRC Press, Boca Raton,
2009)
31. P.G. Harrison, T. Kikabhai, Proton and phosphorus-31 nuclear magnetic resonance study of
zinc (II) o, o -dialkyl dithiophosphates in solution. J. Chem. Soc. Dalton Trans. (4), 807–814
(1987)
32. Z. Pawlak, Tribochemistry of Lubricating Oils, vol. 45 (Elsevier, Amsterdam, 2003)
33. D.R. Armstrong, E.S. Ferrari, K.J. Roberts, D. Adams, An investigation into the molecular
stability of zinc di-alkyl-di-thiophosphates (ZDDPs) in relation to their use as anti-wear and
anti-corrosion additives in lubricating oils. Wear 208(1–2), 138–146 (1997)
34. B.H. Kim, V. Sharma, P.B. Aswath, Chemical and mechanistic interpretation of thermal films
formed by dithiophosphates using XANES. Tribol. Int. 114, 15–26 (2017)
35. H. Spedding, R.C. Watkins, The antiwear mechanism of ZDDP’s. Part I. Tribol. Int. 15(1),
9–12 (1982)
36. T.H. Handley, J.A. Dean, O, O -dialkyl phosphorodithioic acids as extractants for metals.
Anal. Chem. 34(10), 1312–1315 (1962)
37. N.E. Gallopoulos, Thermal decomposition of metal dialkyldithiophosphate oil blends. ASLE
Trans. 7(1), 55–63 (1964)
38. P.A. Willermet, L.R. Mahoney, C.M. Bishop, Lubricant degradation and wear III. Antioxidant
reactions and wear behavior of a zinc dialkyldithiophosphate in a fully formulated lubricant.
ASLE Trans. 23(3), 225–231 (1980)
39. A. Neville, A. Morina, T. Haque, M. Voong, Compatibility between tribological surfaces
and lubricant additives—how friction and wear reduction can be controlled by surface/lube
synergies. Tribol. Int. 40(10), 1680–1695 (2007)
40. N.N. Gosvami, J.A. Bares, F. Mangolini, A.R. Konicek, D.G. Yablon, R.W. Carpick, Mecha-
nisms of antiwear tribofilm growth revealed in situ by single-asperity sliding contacts. Science
348(6230), 102–106 (2015)
41. J. Zhang, H. Spikes, On the mechanism of ZDDP antiwear film formation. Tribol. Lett. 63(2),
1–15 (2016)
42. W. Tysoe, On stress-induced tribochemical reaction rates. Tribol. Lett. 65(2), 48 (2017)
208 A. Dorgham et al.

43. F.T. Barcroft, R.J. Bird, J.F. Hutton, D. Park, The mechanism of action of zinc thiophosphates
as extreme pressure agents. Wear 77(3), 355–384 (1982)
44. H. So, Y.C. Lin, G.G.S. Huang, T.S.T. Chang, Antiwear mechanism of zinc dialkyl dithio-
phosphates added to a paraffinic oil in the boundary lubrication condition. Wear 166(1), 17–26
(1993)
45. A. Molina, Isolation and chemical characterization of a zinc dialkyldithiophosphate-derived
antiwear agent. ASLE Trans. 30(4), 479–485 (1986)
46. R.B. Jones, R.C. Coy, The chemistry of the thermal degradation of zinc dialkyldithiophosphate
additives. ASLE Trans. 24(1), 91–97 (1981)
47. D. Shakhvorostov, M.H. Müser, Y. Song, P.R. Norton, Smart materials behavior in phosphates:
role of hydroxyl groups and relevance to antiwear films. J. Chem. Phys. 131(4), 044704 (2009)
48. I.-M. Feng, Pyrolysis of zinc dialkyl phosphorodithioate and boundary lubrication. Wear 3(4),
309–311 (1960)
49. I.-M. Feng, W.L. Perilstein, M.R. Adams, Solid film deposition and non-sacrificial boundary
lubrication. ASLE Trans. 6(1), 60–66 (1963)
50. R.C. Coy, R.B. Jones, The thermal degradation and EP performance of zinc dialkyldithio-
phosphate additives in white oil. ASLE Trans. 24(1), 77–90 (1981)
51. R.J. Bird, G.D. Galvin, The application of photoelectron spectroscopy to the study of ep films
on lubricated surfaces. Wear 37(1), 143–167 (1976)
52. S.-H. Choa, K.C. Ludema, G.E. Potter, B.M. Dekoven, T.A. Morgan, K.K. Kar, A model of
the dynamics of boundary film formation. Wear 177(1), 33–45 (1994)
53. R.G. Pearson, Hard and soft acids and bases. J. Am. Chem. Soc. 85(22), 3533–3539 (1963)
54. R.G. Pearson, J. Songstad, Application of the principle of hard and soft acids and bases to
organic chemistry. J. Am. Chem. Soc. 89(8), 1827–1836 (1967)
55. R.G. Pearson, Hard and soft acids and bases, HSAB, part 1: fundamental principles. J. Chem.
Educ. 45(9), 581 (1968)
56. R.G. Pearson, Hard and soft acids and bases, HSAB, part II: underlying theories. J. Chem.
Educ. 45(10), 643 (1968)
57. R.G. Pearson, Recent advances in the concept of hard and soft acids and bases. J. Chem. Educ.
64(7), 561 (1987)
58. P.A. Willermet, R.O. Carter III, E.N. Boulos, Lubricant-derived tribochemical films—an infra-
red spectroscopic study. Tribol. Int. 25(6), 371–380 (1992)
59. A.J. Burn, The mechanism of the antioxidant action of zinc dialkyl dithiophosphates. Tetra-
hedron 22(7), 2153–2161 (1966)
60. J.A. Howard, Y. Ohkatsu, J.H.B. Chenier, K.U. Ingold, Metal complexes as antioxidants. I.
The reaction of zinc dialkyldithiophosphates and related compounds with peroxy radicals.
Can. J. Chem. 51(10), 1543–1553 (1973)
61. J.-M. Martin, C. Grossiord, T. Le Mogne, J. Igarashi, Transfer films and friction under bound-
ary lubrication. Wear 245(1–2), 107–115 (2000)
62. M.L.S. Fuller, M. Kasrai, G. Michael Bancroft, K. Fyfe, K.H. Tan, Solution decomposition
of zinc dialkyl dithiophosphate and its effect on antiwear and thermal film formation studied
by X-ray absorption spectroscopy. Tribol. Int. 31(10), 627–644 (1998)
63. Z. Yin, M. Kasrai, G.M. Bancroft, K. Fyfe, M.L. Colaianni, K.H. Tan, Application of soft
X-ray absorption spectroscopy in chemical characterization of antiwear films generated by
ZDDP part II: the effect of detergents and dispersants. Wear 202(2), 192–201 (1997)
64. A. Morina, H. Zhao, J.F.W. Mosselmans, In-situ reflection-XANES study of ZDDP and
MoDTC lubricant films formed on steel and diamond like carbon (DLC) surfaces. Appl.
Surf. Sci. 297, 167–175 (2014)
65. M. Nicholls, M.N. Najman, Z. Zhang, M. Kasrai, P.R. Norton, P.U.P.A. Gilbert, The contri-
bution of XANES spectroscopy to tribology. Can. J. Chem. 85(10), 816–830 (2007)
66. Z. Zhang, E.S. Yamaguchi, M. Kasrai, G.M. Bancroft, X. Liu, M.E. Fleet, Tribofilms generated
from ZDDP and DDP on steel surfaces: part 2, chemistry. Tribol. Lett. 19(3), 221–229 (2005)
67. J.S. Sheasby, T.A. Caughlin, A.G. Blahey, K.F. Laycock, A reciprocating wear test for eval-
uating boundary lubrication. Tribol. Int. 23(5), 301–307 (1990)
6 Tribochemistry and Morphology of P-Based Antiwear Films 209

68. P.A. Willermet, J.M. Pieprzak, D.P. Dailey, R.O. Carter, N.E. Lindsay, L.P. Haack et al., The
composition of surface layers formed in a lubricated cam/tappet contact. J. Tribol. 113(1),
38–47 (1991)
69. P.A. Willermet, D.P. Dailey, R.O. Carter III, P.J. Schmitz, W. Zhu, J.C. Bell, D. Park, The
composition of lubricant-derived surface layers formed in a lubricated cam/tappet contact ii.
effects of adding overbased detergent and dispersant to a simple ZDTP solution. Tribol. Int.
28(3), 163–175 (1995)
70. M. Aktary, M.T. McDermott, G.A. McAlpine, Morphology and nanomechanical properties of
ZDDP antiwear films as a function of tribological contact time. Tribol. Lett. 12(3), 155–162
(2002)
71. N.E. Lindsay, R.O. Carter III, P.J. Schmitz, L.P. Haack, R.E. Chase, J.E. deVries, P.A. Willer-
met, Characterization of films formed at a lubricated cam/tappet contact. Spectrochim. Acta
Part A Mol. Spectrosc. 49(13–14), 2057–2070 (1993)
72. F.G. Rounds, Some factors affecting the decomposition of three commercial zinc organ-
odithiophosphates. ASLE Trans. 18(2), 79–89 (1975)
73. R. McClintock, Effect of lubricants on rear axle pinion bearing breakin. ASLE Trans. 6(2),
154–160 (1963)
74. J.S. Sheasby, T.A. Caughlin, J.J. Habeeb, Observation of the antiwear activity of zinc
dialkyldithiophosphate additives. Wear 150(1–2), 247–257 (1991)
75. M. Kasrai, M. Puller, M. Scaini, Z. Yin, R.W. Brunner, G.M. Bancroft, M.E. Fleet, K. Fyfe,
K.H. Tan, Study of tribochemical film formation using x-ray absorption and photoelectron
spectroscopies. Tribol. Ser. 30, 659–669 (1995)
76. E.H. Loeser, R.C. Wiquist, S.B. Twiss, Cam and tappet lubrication. IV—radioactive study of
sulfur in the EP film. ASLE Trans. 2(2), 199–207 (1959)
77. J. Ye, S. Araki, M. Kano, Y. Yasuda, Nanometerscale mechanical/structural properties of
molybdenum dithiocarbamate and zinc dialkylsithiophosphate tribofilms and friction reduc-
tion mechanism. Jpn. J. Appl. Phys. 44(7B), 5358–5361 (2005)
78. M. Fuller, Z. Yin, M. Kasrai, G. Michael Bancroft, E.S. Yamaguchi, P. Ray Ryason, P.A.
Willermet, K.H. Tan, Chemical characterization of tribochemical and thermal films gener-
ated from neutral and basic ZDDPs using X-ray absorption spectroscopy. Tribol. Int. 30(4),
305–315 (1997)
79. M.L.S. Fuller, L.R. Fernandez, The use of X-ray absorption spectroscopy for monitoring the
thickness of antiwear films from ZDDP. Tribol. Lett. 8, 187–192 (2000)
80. M.A. Nicholls, G. Michael Bancroft, P.R. Norton, M. Kasrai, G. De Stasio, B.H. Frazer, L.M.
Wiese, Chemomechanical properties of antiwear films using X-ray absorption microscopy
and nanoindentation techniques. Tribol. Lett. 17(2), 245–259 (2004)
81. J.M. Martin, M. Belin, J.L. Mansot, H. Dexpert, P. Lagarde, Frictioninduced amorphization
with ZDDP—an EXAFS study. ASLE Trans. 29(4), 523–531 (1986)
82. M.N. Najman, M. Kasrai, G.M. Bancroft, A. Miller, Study of the chemistry of films generated
from phosphate ester additives on 52100 steel using X-ray absorption spectroscopy. Tribol.
Lett. 13(3), 209–218 (2002)
83. M.N. Najman, M. Kasrai, G.M. Bancroft, B.H. Frazer, G. De Stasio, The correlation of micro-
chemical properties to antiwear (AW) performance in ashless thiophosphate oil additives.
Tribol. Lett. 17(4), 811–822 (2004)
84. Z. Zhang, M. Najman, M. Kasrai, G.M. Bancroft, E.S. Yamaguchi, Study of interaction of
EP and AW additives with dispersants using XANES. Tribol. Lett. 18(1), 43–51 (2005)
85. M.N. Najman, M. Kasrai, G.M. Bancroft, Investigating binary oil additive systems containing
P and S using X-ray absorption near-edge structure spectroscopy. Wear 257(1), 32–40 (2004)
86. M. Crobu, A. Rossi, F. Mangolini, N.D. Spencer, Tribochemistry of bulk zinc metaphosphate
glasses. Tribol. Lett. 39(2), 121–134 (2010)
87. R. Heuberger, A. Rossi, N.D. Spencer, XPS study of the influence of temperature on ZnDTP
tribofilm composition. Tribol. Lett. 25(3), 185–196 (2006)
88. R. Heuberger, A. Rossi, N.D. Spencer, Reactivity of alkylated phosphorothionates with steel:
a tribological and surface analytical study. Lubr. Sci. 20, 79–102 (2008)
210 A. Dorgham et al.

89. R. Heuberger, A. Rossi, N.D. Spencer, Pressure dependence of znDTP tribochemical film
formation: a combinatorial approach. Tribol. Lett. 28(2), 209–222 (2007)
90. J.R. Van Wazer, K.A. Holst, Structure and properties of the condensed phosphates. I. Some
general considerations about phosphoric acids. J. Am. Chem. Soc. 72(2), 639–644 (1950)
91. J.R. Van Wazer, Structure and properties of the condensed phosphates. III. Solubility frac-
tionation and other solubility studies. J. Am. Chem. Soc. 72(2), 647–655 (1950)
92. C. Minfray, J.M. Martin, C. Esnouf, T. Le Mogne, R. Kersting, B. Hagenhoff, A multi-
technique approach of tribofilm characterisation. Thin Solid Films 447, 272–277 (2004)
93. R. Heuberger, A. Rossi, N.D. Spencer, XPS study of the influence of temperature on ZnDTP
tribofilm composition. Tribol. Lett. 25(3), 185–196 (2007)
94. M. Crobu, A. Rossi, N.D. Spencer, Effect of chain-length and countersurface on the tribo-
chemistry of bulk zinc polyphosphate glasses. Tribol. Lett. 48(3), 393–406 (2012)
95. M. Crobu, A. Rossi, F. Mangolini, N.D. Spencer, Chain-length-identification strategy in
zinc polyphosphate glasses by means of XPS and ToF-SIMS. Anal. Bioanal. Chem. 403(5),
1415–1432 (2012)
96. E. Liu, S.D. Kouame, An XPS study on the composition of zinc dialkyl dithiophosphate
tribofilms and their effect on camshaft lobe wear. Tribol. Trans. 57(1), 18–27 (2014)
97. E.C. Onyiriuka, Zinc phosphate glass surfaces studied by xps. J. Non-Cryst. Solids 163(3),
268–273 (1993)
98. J.W. Wiench, M. Pruski, B. Tischendorf, J.U. Otaigbe, B.C. Sales, Structural studies of zinc
polyphosphate glasses by nuclear magnetic resonance. J. Non-Cryst. Solids 263, 101–110
(2000)
99. B. Tischendorf, J.U. Otaigbe, J.W. Wiench, M. Pruski, B.C. Sales, A study of short and
intermediate range order in zinc phosphate glasses. J. Non-Cryst. Solids 282(2), 147–158
(2001)
100. G. Walter, U. Hoppe, J. Vogel, G. Carl, P. Hartmann, The structure of zinc polyphosphate
glass studied by diffraction methods and 31P NMR. J. Non-Cryst. Solids 333(3), 252–262
(2004)
101. M. Najman, M. Kasrai, G. Michael Bancroft, R. Davidson, Combination of ashless antiwear
additives with metallic detergents: interactions with neutral and overbased calcium sulfonates.
Tribol. Int. 39(4), 342–355 (2006)
102. B. Dacre, C.H. Bovington, The effect of metal composition on the adsorption of zinc di-
isopropyldithiophosphate. ASLE Trans. 26(3), 333–343 (1983)
103. K. Ito, J.-M. Martin, C. Minfray, K. Kato, Lowfriction tribofilm formed by the reaction of
ZDDP on iron oxide. Tribol. Int. 39(12), 1538–1544 (2006)
104. K. Ito, J.M. Martin, C. Minfray, K. Kato, Formation mechanism of a low friction ZDDP
tribofilm on iron oxide. Tribol. Trans. 50(2), 211–216 (2007)
105. S. Plaza, The adsorption of zinc dibutyldithiophosphates on iron and iron oxide powders.
ASLE Trans. 30(2), 233–240 (1987)
106. M. Watanabe, M. Sakuma, T. Inaba, Y. Iguchi, Formation and oxidation of sulfides on pure
iron and iron oxides. Mater. Trans. JIM 41(7), 865–872 (2000)
107. Y.-R. Li, G. Pereira, M. Kasrai, P.R. Norton, The effect of steel hardness on the performance
of ZDDP antiwear films: a multi-technique approach. Tribol. Lett. 29(3), 201–211 (2008)
108. J.S. Sheasby, T.A. Caughlin, W.A. Mackwood, The effect of steel hardness on the performance
of antiwear additives. Wear 201(1–2), 209–216 (1996)
109. B. Vengudusamy, J.H. Green, G.D. Lamb, H.A. Spikes, Tribological properties of tribofilms
formed from ZDDP in DLC/DLC and DLC/steel contacts. Tribol. Int. 44(2), 165–174 (2011)
110. A. Erdemir, C. Donnet, Tribology of diamond-like carbon films: recent progress and future
prospects. J. Phys. D Appl. Phys. 39(18), R311 (2006)
111. I. Velkavrh, M. Kalin, J. Vizintin, The performance and mechanisms of DLC-coated surfaces
in contact with steel in boundary-lubrication conditions: a review. Strojniški vestnik 54(3),
189–206 (2008)
112. M. Kalin, I. Velkavrh, J. Vižintin, L. Ožbolt, Review of boundary lubrication mechanisms of
DLC coatings used in mechanical applications. Meccanica 43(6), 623–637 (2008)
6 Tribochemistry and Morphology of P-Based Antiwear Films 211

113. D.M. Nuruzzaman, M.A. Chowdhury, A. Nakajima, M.L. Rahaman, S.M.I. Karim, Friction
and wear of diamond like carbon (DLC) coatings—a review. Recent Pat. Mech. Eng. 4(1),
55–78 (2011)
114. R. Zahid, M.B.H. Hassan, M. Varman, R.A. Mufti, M.A. Kalam, N.W.B.M. Zulkifli, M.
Gulzar, A review on effects of lubricant formulations on tribological performance and bound-
ary lubrication mechanisms of non-doped DLC/DLC contacts. Crit. Rev. Solid State Mater.
Sci. 42, 1–28 (2016)
115. S. Equey, S. Roos, U. Mueller, R. Hauert, N.D. Spencer, R. Crockett, Tribofilm formation
from ZnDTP on diamond-like carbon. Wear 264(3), 316–321 (2008)
116. S. Equey, S. Roos, U. Mueller, R. Hauert, N.D. Spencer, R. Crockett, Reactions of zinc-
free anti-wear additives in DLC/DLC and steel/steel contacts. Tribol. Int. 41(11), 1090–1096
(2008)
117. K. Topolovec-Miklozic, F. Lockwood, H. Spikes, Behaviour of boundary lubricating additives
on dlc coatings. Wear 265(11), 1893–1901 (2008)
118. M. Kalin, E. Roman, L. Ožbolt, J. Vižintin, Metal-doped (Ti, WC) diamond-like-carbon
coatings: reactions with extreme-pressure oil additives under tribological and static conditions.
Thin Solid Films 518(15), 4336–4344 (2010)
119. M. Kalin, E. Roman, J. Vižintin, The effect of temperature on the tribological mechanisms
and reactivity of hydrogenated, amorphous diamond-like carbon coatings under oil-lubricated
conditions. Thin Solid Films 515(7), 3644–3652 (2007)
120. S. Akbari, J. Kovač, M. Kalin, Effect of ZDDP concentration on the thermal film formation
on steel, hydrogenated non-doped and Si-doped DLC. Appl. Surf. Sci. 383, 191–199 (2016)
121. T. Haque, A. Morina, A. Neville, R. Kapadia, S. Arrowsmith, Nonferrous coating/lubricant
interactions in tribological contacts: assessment of tribofilms. Tribol. Int. 40(10), 1603–1612
(2007)
122. B. Podgornik, S. Jacobson, S. Hogmark, DLC coating of boundary lubricated compo-
nents—advantages of coating one of the contact surfaces rather than both or none. Tribol. Int.
36(11), 843–849 (2003)
123. B. Podgornik, J. Vižintin, S. Jacobson, S. Hogmark, Tribological behaviour of WC/C coatings
operating under different lubrication regimes. Surf. Coat. Technol. 177, 558–565 (2004)
124. B. Podgornik, J. Vižintin, Tribological reactions between oil additives and dlc coatings for
automotive applications. Surf. Coat. Technol. 200(5), 1982–1989 (2005)
125. M.I. de Barros’ Bouchet, J.M. Martin, T. Le-Mogne, B. Vacher, Boundary lubrication mecha-
nisms of carbon coatings by MoDTC and ZDDP additives. Tribol. Int. 38(3), 257–264 (2005)
126. M. Ban, M. Ryoji, S. Fujii, J. Fujioka, Tribological characteristics of Si-containing diamond-
like carbon films under oil-lubrication. Wear 253(3), 331–338 (2002)
127. B. Podgornik, S. Jacobson, S. Hogmark, Influence of ep additive concentration on the tribo-
logical behaviour of dlc-coated steel surfaces. Surf. Coat. Technol. 191(2), 357–366 (2005)
128. M. Kalin, J. Vižintin, J. Barriga, K. Vercammen, K. van Acker, A. Arnšek, The effect of
doping elements and oil additives on the tribological performance of boundary-lubricated
dlc/dlc contacts. Tribol. Lett. 17(4), 679–688 (2004)
129. M. Kalin, J. Vižintin, Differences in the tribological mechanisms when using non-doped,
metal-doped (Ti, WC), and non-metal-doped (Si) diamond-like carbon against steel under
boundary lubrication, with and without oil additives. Thin Solid Films 515(4), 2734–2747
(2006)
130. T. Haque, A. Morina, A. Neville, Tribological performance evaluation of a hydrogenated
diamond-like carbon coating in sliding/rolling contact—effect of lubricant additives. Proc.
Inst. Mech. Eng. Part J J. Eng. Tribol. 225(6), 393–405 (2011)
131. L. Yang, A. Neville, A. Brown, P. Ransom, A. Morina, Effect of lubricant additives on the
WDLC coating structure when tested in boundary lubrication regime. Tribol. Lett. 57(2), 14
(2015)
132. C. Donnet, A. Grill, Friction control of diamond-like carbon coatings. Surf. Coat. Technol.
94, 456–462 (1997)
212 A. Dorgham et al.

133. M. Kalin, S. Jahanmir, G. Dražič, Wear mechanisms of glass-infiltrated alumina sliding against
alumina in water. J. Am. Ceram. Soc. 88(2), 346–352 (2005)
134. L. Lazzarotto, L. Dubar, A. Dubois, P. Ravassard, J. Oudin, Three selection criteria for the
cold metal forming lubricating oils containing extreme pressure agents. J. Mater. Process.
Technol. 80, 245–250 (1998)
135. A. Naveira-Suarez, A. Tomala, M. Grahn, M. Zaccheddu, R. Pasaribu, R. Larsson, The influ-
ence of base oil polarity and slide–roll ratio on additive-derived reaction layer formation.
Proc. Inst. Mech. Eng. Part J J. Eng. Tribol. 225, 565–576 (2011)
136. A. Tomala, A. Naveira-Suarez, I.C. Gebeshuber, R. Pasaribu, Effect of base oil polarity on
micro and nanofriction behaviour of base oil+ ZDDP solutions. Tribol. Mater. Surf. Interfaces
3(4), 182–188 (2009)
137. A. Naveira Suarez, M. Grahn, R. Pasaribu, R. Larsson, The influence of base oil polarity
on the tribological performance of zinc dialkyl dithiophospate additives. Tribol. Int. 43(12),
2268–2278 (2010)
138. P. Kar, P. Asthana, H. Liang, Formation and characterization of tribofilms. J. Tribol. 130(4),
042301 (2008)
139. I. Minami, K. Hirao, M. Memita, S. Mori, Investigation of anti-wear additives for low viscous
synthetic esters: hydroxyalkyl phosphonates. Tribol. Int. 40(4), 626–631 (2007)
140. I. Minami, S. Mori, Anti-wear additives for ester oils. J. Synth. Lubr. 22(2), 105–121 (2005)
141. A. Dorgham, A. Neville, K. Ignatyev, F. Mosselmans, A. Morina, An in situ synchrotron
XAS methodology for surface analysis under high temperature, pressure, and shear. Rev. Sci.
Instrum. 88(1), 015101 (2017)
142. E.S. Ferrari, K.J. Roberts, D. Adams, A multi-edge X-ray absorption spectroscopy study of the
reactivity of zinc di-alkyl-di-thiophosphates (ZDDPS) anti-wear additives: 1. An examination
of representative model compounds. Wear 236(1), 246–258 (1999)
143. H. Cen, A. Morina, A. Neville, R. Pasaribu, I. Nedelcu, Effect of water on ZDDP anti-wear
performance and related tribochemistry in lubricated steel/steel pure sliding contacts. Tribol.
Int. 56, 47–57 (2012)
144. I. Nedelcu, E. Piras, A. Rossi, H.R. Pasaribu, XPS analysis on the influence of water on the
evolution of zinc dialkyldithiophosphate–derived reaction layer in lubricated rolling contacts.
Surf. Interface Anal. 44(8), 1219–1224 (2012)
145. E.S. Ferrari, K.J. Roberts, M. Sansone, D. Adams, A multi-edge xray absorption spectroscopy
study of the reactivity of zinc di-alkyl-dithiophosphates anti-wear additives: 2. In situ studies
of steel/oil interfaces. Wear 236(1), 259–275 (1999)
146. C.H. Bovington, B. Dacre, The adsorption and reaction of decomposition products of zinc
di-isopropyldiophosphate on steel. ASLE Trans. 27(3), 252–258 (1984)
147. A. Ghanbarzadeh, P. Parsaeian, A. Morina, M.C.T. Wilson, M.C.P. van Eijk, I. Nedelcu, D.
Dowson, A. Neville, A semi-deterministic wear model considering the effect of zinc dialkyl
dithiophosphate tribofilm. Tribol. Lett. 61(1), 12 (2016)
148. P. Parsaeian, A. Ghanbarzadeh, M. Wilson, M.C.P. Van Eijk, I. Nedelcu, D. Dowson, A.
Neville, A. Morina, An experimental and analytical study of the effect of water and its tribo-
chemistry on the tribocorrosive wear of boundary lubricated systems with ZDDP-containing
oil. Wear 358, 23–31 (2016)
149. P. Parsaeian, A. Ghanbarzadeh, M.C.P. Van Eijk, I. Nedelcu, A. Morina, A. Neville, Study
of the interfacial mechanism of ZDDP tribofilm in humid environment and its effect on
tribochemical wear; part II: numerical. Tribol. Int. 107, 33–38 (2017)
150. G. Pereira, D. Munoz-Paniagua, A. Lachenwitzer, M. Kasrai, P.R. Norton, T. Weston Cape-
hart, T.A. Perry, Y.-T. Cheng, A variable temperature mechanical analysis of ZDDP-derived
antiwear films formed on 52100 steel. Wear 262(3), 461–470 (2007)
151. H. Ji, M.A. Nicholls, P.R. Norton, M. Kasrai, T. Weston Capehart, T.A. Perry, Y.-T. Cheng,
Zinc-dialkyldithiophosphate antiwear films: dependence on contact pressure and sliding
speed. Wear 258(5), 789–799 (2005)
152. Y. Shimizu, H.A. Spikes, The influence of slide–roll ratio on ZDDP tribofilm formation.
Tribol. Lett. 64(2), 19 (2016)
6 Tribochemistry and Morphology of P-Based Antiwear Films 213

153. P.M.E. Cann, G.J. Johnston, H.A. Spikes, Formation of Thick Films by Phosphorus-Based
Anti-wear Additives (Mechanical Engineering Publication Ltd, 1987), pp. 543–554
154. H. So, Y.C. Lin, The theory of antiwear for ZDDP at elevated temperature in boundary
lubrication condition. Wear 177(2), 105–115 (1994)
155. S.A.J.M.R.C.J.C. Bec, A. Tonck, J.-M. Georges, R.C. Coy, J.C. Bell, G.W. Roper, Relationship
between mechanical properties and structures of zinc dithiophosphate anti–wear films, in
Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering
Sciences, vol. 455 (The Royal Society, 1999), pp. 4181–4203
156. O.L. Warren, J.F. Graham, P.R. Norton, J.E. Houston, T.A. Michalske, Nanomechanical prop-
erties of films derived from zinc dialkyldithiophosphate. Tribol. Lett. 4(2), 189–198 (1998)
157. J.F. Graham, C. McCague, P.R. Norton, Topography and nanomechanical properties of tribo-
chemical films derived from zinc dialkyl and diaryl dithiophosphates. Tribol. Lett. 6, 149–157
(1999)
158. A.J. Pidduck, G.C. Smith, Scanning probe microscopy of automotive anti-wear films. Wear
212(2), 254–264 (1997)
159. L. Taylor, A. Dratva, H.A. Spikes, Friction and wear behavior of zinc dialkyldithiophosphate
additive. Tribol. Trans. 43(3), 469–479 (2000)
160. M. Aktary, M.T. McDermott, J. Torkelson, Morphological evolution of films formed from
thermooxidative decomposition of ZDDP. Wear 247(2), 172–179 (2001)
161. G.W. Canning, M.L. SuominenFuller, G.M. Bancroft, M. Kasrai, J.N. Cutler, G. De Stasio, B.
Gilbert, Spectromicroscopy of tribological films from engine oil additives. Part I Films from
ZDDP’s. Tribol. Lett. 6(3–4), 159–169 (1999)
162. M.A. Nicholls, P.R. Norton, G.M. Bancroft, M. Kasrai, G. De Stasio, L.M. Wiese, Spatially
resolved nanoscale chemical and mechanical characterization of ZDDP antiwear films on
aluminum-silicon alloys under cylinder/bore wear conditions. Tribol. Lett. 18(3), 261–278
(2005)
163. J. Ye, M. Kano, Y. Yasuda, Evaluation of local mechanical properties in depth in
MoDTC/ZDDP and ZDDP tribochemical reacted films using nanoindentation. Tribol. Lett.
13(1), 41–47 (2002)
164. J. Ye, M. Kano, Y. Yasuda, Evaluation of nanoscale friction depth distribution in ZDDP and
MoDTC tribochemical reacted films using a nanoscratch method. Tribol. Lett. 16, 107–112
(2004)
165. J. Ye, M. Kano, Y. Yasuda, Friction property study of the surface of ZDDP and MoDTC
antiwear additive films using AFM/LFM and force curve methods. Tribotest 9(1), 13–21
(2002)
166. J. Ye, M. Kano, Y. Yasuda, Determination of nanostructures and mechanical properties on
the surface of molybdenum dithiocarbamate and zinc dialkyl-dithiophosphate tribochemical
reacted films using atomic force microscope phase imaging technique. J. Appl. Phys. 93(9),
5113–5117 (2003)
167. K.T. Miklozic, J. Graham, H. Spikes, Chemical and physical analysis of reaction films formed
by molybdenum dialkyl-dithiocarbamate friction modifier additive using raman and atomic
force microscopy. Tribol. Lett. 11(2), 71–81 (2001)
168. C.C. Chou, J.F. Lin, A new approach to the effect of EP additive and surface roughness on
the pitting fatigue of a line-contact system. J. Tribol. 124(2), 245–258 (2002)
169. M.A. Nicholls, T. Do, P.R. Norton, G. Michael Bancroft, M. Kasrai, T. Weston Capehart,
Y.-T. Cheng, T. Perry, Chemical and mechanical properties of ZDDP antiwear films on steel
and thermal spray coatings studied by XANES spectroscopy and nanoindentation techniques.
Tribol. Lett. 15(3), 241–248 (2003)
170. K.D. Costa, Single-cell elastography: probing for disease with the atomic force microscope.
Dis. Markers 19(2–3), 139–154 (2004)
171. P. Parsaeian, A. Ghanbarzadeh, M.C.P. Van Eijk, I. Nedelcu, A. Neville, A. Morina, A new
insight into the interfacial mechanisms of the tribofilm formed by zinc dialkyl dithiophosphate.
Appl. Surf. Sci. 403, 472–486 (2017)
214 A. Dorgham et al.

172. M.C. Friedenberg, C. Mathew Mate, Dynamic viscoelastic properties of liquid polymer films
studied by atomic force microscopy. Langmuir 12(25), 6138–6142 (1996)
173. J.-M. Georges, A. Tonck, S. Poletti, E.S. Yamaguchi, P.R. Ryason, Film thickness and mechan-
ical properties of adsorbed neutral and basic zinc diisobutyl dithiophosphates. Tribol. Trans.
41(4), 543–553 (1998)
174. I. Neitzel, V. Mochalin, J.A. Bares, R.W. Carpick, A. Erdemir, Y. Gogotsi, Tribological prop-
erties of nanodiamond-epoxy composites. Tribol. Lett. 47(2), 195–202 (2012)
175. L.J. Taylor, H.A. Spikes, Friction-enhancing properties of ZDDP antiwear additive: part
I—friction and morphology of ZDDP reaction films. Tribol. Trans. 46(3), 303–309 (2003)
176. D. Mazuyer, A. Tonck, S. Bec, J.L. Loubet, J.M. Georges, Nanoscale surface rheology in
tribology. Tribol. Ser. 39, 273–282 (2001)
177. J.-M. Georges, A. Tonck, J.-L. Loubet, D. Mazuyer, E. Georges, F. Sidoroff, Rheology and
friction of compressed polymer layers adsorbed on solid surfaces. J. Phys. II 6(1), 57–76
(1996)
178. Q. Li, C. Lee, R.W. Carpick, J. Hone, Substrate effect on thickness-dependent friction on
graphene. Physica Status Solidi (b) 247(11–12), 2909–2914 (2010)
179. C. Lee, X. Wei, Q. Li, R. Carpick, J.W. Kysar, J. Hone, Elastic and frictional properties of
graphene. Physica Status Solidi (b) 246(11–12), 2562–2567 (2009)
180. C. Lee, Q. Li, W. Kalb, X.-Z. Liu, H. Berger, R.W. Carpick, J. Hone, Frictional characteristics
of atomically thin sheets. Science 328(5974), 76–80 (2010)
181. J. Gansheimer, On the lubricating properties of mixtures of mineral oil with certain inorganic
phosphates, hydroxides, and sulfides. ASLE Trans. 15(3), 201–206 (1972)
182. D. Landolt, S. Mischler, Tribocorrosion of Passive Metals and Coatings (Elsevier, Amsterdam,
2011)
183. M. Duncanson, Detecting and controlling water in oil, in Proceedings of Noria Lubrication
Excellence 2005 Conference, San Antonio, Texas, 25–29 April 2005
184. A.C. Eachus, The trouble with water. Tribol. Lubr. Technol. 61(10), 32–38 (2005)
185. P.E. Sheehan, The wear kinetics of NaCl under dry nitrogen and at low humidities. Chem.
Phys. Lett. 410(1), 151–155 (2005)
186. P. Parsaeian, M.C.P. Van Eijk, I. Nedelcu, A. Neville, A. Morina, Study of the interfacial
mechanism of ZDDP tribofilm in humid environment and its effect on tribochemical wear;
part I: experimental. Tribol. Int. 107, 135–143 (2017)
187. O.D. Faut, D.R. Wheeler, On the mechanism of lubrication by tricresylphosphate (TCP)—the
coefficient of friction as a function of temperature for TCP on M-50 steel. ASLE Trans. 26(3),
344–350 (1983)
Chapter 7
In Situ Observation of Lubricating Films
by Micro-FTIR

Shigeyuki Mori

Abstract Tribological properties closely depend on the structures of lubricating

films between moving surfaces. Lubricating films are formed at lubricating contacts
under dynamic conditions, including high temperature, high pressure and high shear-
ing. Thus, structures of lubricating films are greatly affected by the conditions of the
environment. In situ observation of lubricating films by micro-FTIR is described in
order to characterize the film components and structure formed from lubricants under
dynamic conditions.

7.1 Background

7.1.1 Tribological Contact

Tribological properties closely depend on the structures of lubricating films between

moving surfaces. Lubricating films are formed at lubricating contacts under dynamic
conditions, including high temperature, high pressure and high shearing. Thus, struc-
tures of lubricating films are greatly affected by the conditions of the environment
(Fig. 7.1). A lubricant film under elasto-hydrodynamic lubrication (EHL) conditions
can be formed using a lubricant having proper viscosity only during lubricating con-
ditions. Under boundary lubrication conditions, lubricant films that are dynamically
altered by friction and wear are formed via the balance of formation and removal of
films during lubrication. Lubricating films are formed dynamically, and film struc-
tures should be altered during lubricating process. Therefore, in situ observation of
a lubricating film is needed to fundamentally understand tribological phenomena.
Because tribological phenomena occur under the effect of mechanical energy, in situ
observation is one of the most important techniques to solve tribological problems
[1].

S. Mori (B)
Department of Chemical Engineering, Faculty of Engineering, Iwate University, 4-3-5 Ueda,
Morioka, Iwate 020-8551, Japan
e-mail: mori@iwate-u.ac.jp
© Springer Nature Switzerland AG 2018 215
M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_7
216 S. Mori

Fig. 7.1 Dynamic structure of lubricant film at sliding contact

7.1.2 In Situ Observation

There are many techniques for in situ observation of lubricating films. The meth-
ods for in situ observation using electromagnetic waves, such as infrared, visible,
ultraviolet and X-ray light, are useful to investigate lubricating films in air, because
air is transparent for electromagnetic waves (Fig. 7.2). At least one tribomaterial
should be transparent for the electromagnetic wave to obtain the information of a
lubricant film between two surfaces. The method using interferometry and a space
layer disk makes it possible to determine the thickness of a lubricant film at nanoscale
under EHL conditions [2]. However, the information on chemical structures of the
film cannot be supplied, although tribological properties are closely dependent on the
chemical structure of the film. Spectroscopic methods are better able to obtain chem-
ical information from film spectra. For example, spectra of visible and ultraviolet
regions provide chemical bonds, while infrared spectra contain vibration modes of
chemical bonds. In this chapter, the application of infrared spectroscopy to chemical
analysis of lubrication films is summarized.
In situ observation of lubricating films has been performed with infrared emission
spectroscopy, and molecular orientation of a lubricant film has been observed [3].
However, the spectrum is not enough to analyze chemical structures of lubricant
films. Fourier-transform infrared spectroscopy (FTIR) is developed to supply clear
IR spectroscopic information. High spatial resolution of spectrum can be obtained
with an infrared microscope. Then, micro-FTIR has been used for in situ observation
of lubricating films under EHL conditions [4, 5].
7 In Situ Observation of Lubricating Films by Micro-FTIR 217

Fig. 7.2 In situ observation

of lubricant film with light

7.2 Principle and Method

7.2.1 Infrared Spectroscopy

Molecular vibration absorbs infrared radiation. There are typical vibration modes
such as C–H, C–O–C, C=O, O–H, and N–H at 3000–2840, 1260–1000, 1750–1735,
3650–3200, and 3400–3350 cm−1 , respectively. Their absorption peaks are observed
at different positions, for example C–H stretching vibration at 2930 cm−1 . Organic
liquid as a base oil can be detected from the peak of C–H stretching vibration.
Additives such as oiliness agents can be analyzed by the vibrations of C=O, O–H,
N–H and so on. Additives having organic and inorganic functional groups, such as
phosphate, are also detected by IR absorption peak. Therefore, chemical components
of a lubricating film can be assigned from IR absorption spectrum of the film (see
Table 7.1) [6].

7.2.2 Lubrication Tester

Figure 7.3a, b shows a typical ball-on-disk tribotester for in situ IR analysis [5, 7].
Ball and disk are made of bearing steel AISI52100 and IR transparent material,
respectively. Lubrication tests can be performed between ball and disk under EHL
conditions. The IR beam comes from a Cassegrain mirror of a microscope through
an IR transparent disk and is reflected on steel ball surface (Fig. 7.3b). The IR beam
218 S. Mori

Table 7.1 IR absorption peak positions of lubricant components

Lubricant component Chemical structure Peak position (cm-1 )
Base oil hydrocarbon C–H st 3000–2840
C–H bend 1600–1400
Ether C–O–C 1260–1000
Ester COOR 1750–1735
1300–1000
Oiliness agent O–H 3650–3580
3550–3200
COOH 1720–1700
N–H 3400–3350
3330–3250
Viscosity index improver COOR 1750–1735
1300–1000
EP additive sulfide S–S 500–400 w
S–C 700–600 w
Phosphate P=O 1350–1250
O–P–O 970–940
Thickener soap COO- 1610–1550
1420–1300
2nd-amide N–H assoc 3100–3070
2nd-amide C=O assoc 1650
st: stretching, bend: bending, assoc: associated, w: weak

crosses the lubricant film layer twice, and the IR spectrum of the oil film can be
obtained during lubrication tests. The sampling position for IR spectrum is scanned
as shown in Fig. 7.4a for mapping and (b) for rapid scan.

7.2.3 Materials

Disk window material can be selected from diamond, silicon, germanium, calcium
fluoride (CaF2 ), and sapphire (Al2 O3 ), depending on the IR transparent region [5],
strength, hardness, and cost. Because the mechanical property of silicon is similar
to bearing steel, it is a convenient material to investigate an EHL lubrication film.
Although CaF2 has a wide range of IR transparency, it is weak against moisture and
polar compounds such as soap. Silicon and germanium have a good transparency of
IR, but they are not transparent for visible light.
7 In Situ Observation of Lubricating Films by Micro-FTIR 219

Fig. 7.3 A typical ball-on-disk tester with micro-FTIR; a Photograph of the tester b schematic
diagram

Fig. 7.4 Aperture size and data sampling position a Scanning mode, b rapid scan mode

7.3 Structure of EHL Film

7.3.1 Film Thickness and Pressure

Figure 7.5a shows the IR spectrum of a lubricant film formed with synthetic hydro-
carbon oil (PAO:polyalphaolefin) under EHL conditions [8, 9]. The main peak at
2930 cm−1 is originated from C–H stretching vibration of PAO. According to the
Beer-Lambert law (5.1), the thickness (t F ) of an oil film can be estimated from the
220 S. Mori

Fig. 7.5 a IR spectra of PAO at different sampling positions (cf. Fig 7.3d), b the profile of film
thickness obtained from the absorbance. Conditions: 0.2 m/s, 10 N

absorbance (A) of the peak using a calibration curve obtained previously. Because
an aperture size for an IR beam is 20–50 µm, the distribution of oil film thickness
can be obtained by scanning the measuring position around the Hertzian contact,
as shown in Fig. 7.5b. The sensitivity of IR measurement for thickness is less than
50 nm and depends on the absorption coefficient (ε) of the characteristic peak. The
film thickness of hydrocarbon oil at even several microns, which is impossible to
measure with interferometry, can be measured from the absorbance of C–H bending
vibration peak. When the film is too thick, the IR peak intensity becomes saturated
and then film thickness cannot be precisely obtained.

A  −εc t F (5.1)

where ε and c are the absorption coefficient and the concentration of the component,
respectively.
The peak position of C–H vibration in the IR spectrum at the center of Hertzian
contact shifts slightly to a higher wavenumber as shown in Fig. 7.5a, which can be
explained by the configuration of alkyl chains under high pressure. Because the shift
rate has been reported as 15 cm−1 /GPa [8], the distribution of pressure around the
contact area can be estimated from the peak shift as shown in Sect. 4.2.

7.3.2 Component of Lubricant

The lubricant is a mixture of a base oil and additives such as oiliness agents, friction
modifiers, viscosity index improvers, extreme pressure additives, anti-wear additives
and so on. The lubricating property is closely dependent on the lubricant component
7 In Situ Observation of Lubricating Films by Micro-FTIR 221

Fig. 7.6 IR spectrum of oleic acid containing lubricant obtained during lubrication under EHL
condition

Fig. 7.7 IR spectra of a monomer and a dimer of oleic acid in polar and non-polar base oils obtained
under static conditions of 3.4 N load

and the concentration of additives. IR spectrum provides chemical structures and

concentrations of lubricant components. Figure 7.6 shows IR spectrum of a lubricant
film containing oleic acid as an oiliness agent in PAO under a dynamic condition [9,
10]. In addition to the peak of C–H from PAO, there is a small peak of C=O from
oleic acid. The concentration of oleic acid can be estimated from the intensity ratio
of C=O and C–H, according to a calibration curve obtained previously. The peak
position of C=O at 1720 cm−1 indicates that the carboxyl group of the oleic acid
forms a dimer which can dissolve in non-polar base oil. When oleic acid is dissolved
in a polar base oil such as polyalkylene glycol, the peak position of C=O is shifted
to a higher wavenumber at 1730 cm−1 . This result indicates that the carboxyl group
interacts with glycol as a polar molecule and then oleic acid exists as a monomer in
a polar base oil as shown in Fig. 7.7 [7].
222 S. Mori

Fig. 7.8 Effect of base oil

on concentration of oleic
acid at EHL contact region.
Conditions: load 10 N,
entrainment speed 0.13 m/s,
SRR 0%

A dynamic behavior of viscosity index improver (VII) can be observed by a

micro-FTIR spectrometer [11]. The concentration of lubricant components such as
alcohols, amines, amides, and water can also be estimated from their characteristic
absorption peaks in IR spectra as shown in the following section.

7.3.3 Chemical Information

The chemical composition of a lubricant affects the tribological property through

viscosity of the lubricant and the adsorption of additives on lubricating surfaces. The
distribution of additive concentration can be obtained by scanning sampling position
in IR spectra as shown in Fig. 7.4a, b. Figure 7.8 shows a concentration profile of
oleic acid in PAO observed around the Hertzian contact during EHL lubricating
tests [7]. The concentration change at the lubricating contact is dependent on the
combination of the additive and base oil, as shown in Fig. 7.8. The concentration
of oleic acid in PAO as a non-polar base oil is decreased to almost half of the
original concentration. When oleic acid dissolves in polypropylene glycol (PPG)
or polybutylene glycol (PBG) as a polar base oil, the concentration of oleic acid
is not decreased in the contact zone. It can be explained that oleic acid comes into
the contact zone with the base oil because the oleic acid molecule interacts with
a polar base oil, as shown in Fig. 7.7. In the case of polybutene as a base oil, the
concentration is not changed, although the base oil is non-polar. Polybutene has a
large pressure-viscosity coefficient and is known to solidify through high pressure at
the EHL contact. Then, the solidified lubricant with oleic acid comes into the contact
zone without changing the concentration of oleic acid.
7 In Situ Observation of Lubricating Films by Micro-FTIR 223

Fig. 7.9 Chemical shift of characteristic IR peak position of C–H (a) and O–H (b) vibrations.
Molecular interaction of PPG through hydrogen bonding of the end-group OH (c)

7.4 Molecular Interaction

7.4.1 Interaction of Components

The molecular interaction between lubricant components affects the physical prop-
erties of lubricants. For example, the solubility of an additive to a base oil is closely
dependent on the molecular interaction of the two components. The adsorption prop-
erty of additives can be controlled by the balance of the molecular interaction of the
additive with the surface and base oils. The viscosity of a lubricant is also affected
by molecular interaction of the base oil itself. The molecular interaction is expected
to be altered by high pressure at lubricating contacts.
Figure 7.10a shows IR spectra of polypropylene glycol inside (0 µm) and outside
(−150 µm) of the EHL contact region. A blue shift of C–H vibration due to high
pressure can be seen as described above (see Fig. 7.5a) [8]. On the other hand,
the peak of O–H vibration originated from the end-group of polypropylene glycol
shifts to a lower wavenumber as shown in Fig. 7.9b. It is opposite to the shift of
C–H vibration. The result can be explained by the molecular interaction of O–H
through hydrogen bonding at high pressure, as shown in Fig. 7.9c. The end-group
O–H interacts with other O–H groups through hydrogen bonding. The interaction
stabilizes at high pressure, resulting in a long bond distance of O–H bond. Because
bond strength decreases by increasing bond distance between O and H, the peak
position of O–H vibration shifts to a lower wavenumber. The molecular interaction
of carboxylic acid, amine, and amide will be shown in the following section.
The molecular interaction of the base oil at high pressure affects the traction
coefficient through the viscosity of a lubricant at high pressure. Two polypropy-
lene glycols are selected to compare the lubrication property under EHL conditions.
Polypropylene glycol has an O–H end-group, but the other has an O–CH3 group
instead. Their main chain structures are the same, and thus the viscosities of the
two oils at atmospheric pressure are similar. The former molecules interact through
224 S. Mori

Fig. 7.10 Effect of load on peak position of PPG lubricant component around EHL contact

hydrogen bonding between end groups, but the latter molecules interact with each
other without hydrogen bonding. The glycol with O–H exhibits a higher traction
coefficient (0.06) than that without O–H (0.026) obtained by a ball on disk tester
under the conditions of 16 N load, 0.27 m/s sliding speed and 40 °C [12]. One of
the explanations for the high traction coefficient is the molecular interaction through
hydrogen bonding by high pressure at contact area.

7.4.2 Effect of Load

Figure 7.10a shows the distribution of chemical shifts of C=O vibration of oleic acid
in PAO around the Hertzian contact zone [7]. The shift increases at the center of the
Hertzian contact under high load. The peak shift of O–H vibration of the end-group
in polypropylene glycol is also observed, as shown in Fig. 7.10b [8].
These results clearly indicate that the hydrogen bonding of COOH and O–H
groups can be stabilized by high pressure at the lubricating contact.
The peak shift at high pressure and temperature is confirmed through IR measure-
ment using Diamond Anvil Cell (DAC) under the conditions of high pressure (up to
5 GPa) and high temperature (up to 100 °C) [12]. A red shift of peaks is observed
at high pressure. On the other hand, a blue shift is observed at high temperature.
The rate of peak shift of O–H group is −25 cm−1 /GPa at 25 °C and 0.18 cm−1 /K
under atmospheric pressure. The pressure has more effective on the peak shift than
the temperature under these observation conditions.
7 In Situ Observation of Lubricating Films by Micro-FTIR 225

Fig. 7.11 Vibration modes

and their vectors of electric
vibration

7.5 Molecular Alignment

7.5.1 Effect of Shear

It is well-known that in non-Newtonian fluids, the viscosity of a lubricant decreases

at high shear conditions, which is explained by the molecular alignment of lubri-
cant molecules. The molecular alignment can be detected by IR or polarized IR
spectroscopy as shown in Fig. 7.11. There are two stretching vibration modes in a
carboxyl group, symmetric vs and anti-symmetric vas . Their electronic vectors are
at right angles to each other. The stretching vibration of the methylene group CH2
in the alkyl chain also has two vibrational modes as shown in Fig. 7.11. Polarized
IR spectra reveal the molecular orientation under shearing condition from a linear
dichroic ratio (D) of IR absorption as (5.2),

D  A(//)/A(⊥) (5.2)

where A(//) and A(⊥) are the absorbances for radiation polarized parallel and per-
pendicular, respectively (see Fig. 7.11).

7.5.2 Effect of Molecular Structure

Long alkyl chains of a lubricant can form a line along the sliding direction under
shearing conditions as mentioned above. The molecular orientation of fatty acids can
also be detected by a dichroic ratio of IR spectrum (5.2). Alkyl cyanobiphenyl (ACB)
is one of the liquid crystals and exhibits a low traction coefficient (0.035) when a
lubrication test with ACB is carried out under EHL conditions [13]. When alkyl
cyanocyclohexylphenyl (ACC) is used as a lubricant, the traction coefficient (0.045)
is slightly higher compared with ACB as shown in Fig. 7.12a. This result suggests
that a rigid and flat structure of biphenyl group in ACB molecules plays an important
226 S. Mori

Fig. 7.12 In situ observation of molecular alignment of liquid crystal (5CB) under shearing con-
dition by micro-FTIR a Molecular structure and traction coefficient, b subtracted IR spectrum, c
vibration modes and IR beam

role for the low traction coefficient. Therefore, the molecular orientation of ACB
is confirmed by in situ observation during the sliding condition. Figure 7.12b is the
subtracted IR spectrum of ACB obtained by in situ observation during sliding. The IR
spectrum obtained under a dynamic condition was subtracted by that obtained under
a static condition, so that the intensity of C–H vibration that mainly originated from
the alkyl chain disappears. There are three characteristic peaks at 1606 and 1494, 845
and 2227 cm−1 assigned as benzene ring breathing (in plane), benzene C–H bending
(out of plane), and stretching vibration of cyano-group C≡N. It is noteworthy that the
intensities of benzene (in plane) and cyano-group are increased, but that of benzene
(out of plane) is decreased. This result clearly shows that the plane of diphenyl plane
is oriented in the plane of shear during lubricating condition as shown in Fig. 7.12c.
Therefore, one of the important factors for the low traction coefficient of ACB is the
molecular orientation of a flat structure of the cyano-biphenyl group in the shearing
plane. A similar behavior of molecular orientation can be seen with a mixture of
benzoic acid derivatives and oleic acid which aggregate through hydrogen bonding
with carboxyl groups [14].
7 In Situ Observation of Lubricating Films by Micro-FTIR 227

Fig. 7.13 In situ observation

of lubricating film formed
during friction with Fe film
and steel rod using ATR
method: penetration depth,
d p [see (5.3)]

7.6 Boundary Lubrication

7.6.1 Adsorbed Layer

The structure of the adsorbed layer plays an important role in the tribological per-
formance in boundary lubrication. A lubricant film of 0.1–1 µm in thickness formed
under EHL conditions is suitable to be analyzed using IR spectroscopy. This method,
however, is not sensitive enough to detect nanoscale films, such as adsorption films
of lubricant additives. In the case of films at the nanoscale, a highly surfaces-sensitive
method, such as Sum Frequency Generation (SFG) spectroscopy, is needed, and the
results obtained with SFG will be mentioned below. Attenuated Total Reflection
(ATR) IR spectroscopy can be applied to in situ observation of the chemical struc-
tures of thin surface films. When the infrared beam reflects totally at the interface
between IR transparent materials such as Ge and lubricant oil, the IR beam penetrates
into the oil phase as shown in Fig. 7.13. As described in (5.3), the penetration depth
(d p ) is determined by the incident angle (θ ), the index of refraction of materials (n1 ,
n2 ), and the wavelength (λ) of IR. In the case of Ge, lubricant and incident angle
of 45°, the IR beam can penetrate into the lubricant phase with depth at submicron.
When the Ge surface is covered with a thin layer (around 10 nm) of Fe prepared by
sputter deposition, chemical interaction, including the chemical reaction of iron with
the lubricant components, can be analyzed by ATR method [15].
λ
dp    (5.3)
2π n 1 sin θ − (n 2 /n 1 )2
2

7.6.2 Reaction Layer

Boundary films formed through tribochemical reaction were also analyzed by ATR
method. A reciprocating friction test of a steel roller and an Fe-coated ATR disk with
ZnDTP-containing lubricant was carried out under ATR spectroscopic measurement.
The tribochemical reaction product from ZnDTP was detected during a lubrication
228 S. Mori

Fig. 7.14 In situ observation

of lubricant film formed
from ZnDTP at 150 °C under
(a) static and (b) dynamic
conditions at the sliding
interface between Fe film
and a steel rod

test by ATR method [16]. Although no spectral change is observed during heating at
150 °C for 90 h (Fig. 7.14a), phosphate as a tribochemical reaction product is detected
by a lubrication test at 150 °C (Fig. 7.14b). A lubricating film was investigated under
a boundary lubricating condition by ATR method [15].

7.7 Applications

IR spectroscopy has been applied to in situ observation of lubricant films mainly

under EHL conditions. Other applications are described below.

7.7.1 Water-Based Lubricant

There are several lubrication fluids with water for practical use, such as O/W (oil in
water) and W/O (water in oil) emulsions and water soluble lubricants. The behavior of
oil and water at the sliding contacts is important to understand tribological properties
7 In Situ Observation of Lubricating Films by Micro-FTIR 229

Fig. 7.15 Phase separation of o/w emulsion at lubricating contact under EHL condition

of water-based lubricants. For example, a base oil with the proper viscosity is needed
to form lubricant film at EHL contact. Water, however, has too low viscosity to form
water film. Therefore, phase separation of a water-based lubricant at the contact zone
is necessary for lubrication with water-based lubricants.
Figure 7.15 shows IR spectra obtained by in situ observation of EHL film lubri-
cating with O/W emulsion [17]. There are two main peaks of broad O–H peak from
water and sharp C–H peak from the base oil at 3400 and 2960 cm−1 , respectively. It
is clear that the absorbance of O–H decreases from the outside to the inside of the
contact zone and no water is observed at the center of EHL contact. This suggests that
phase separation occurs at the entrance of the contact zone, resulting in the formation
of oil film at EHL contact. A similar oil separation is observed with w/o emulsion
[18] and water soluble lubricants [19].

7.7.2 Semi-Solid Lubricant

Greases which are a mixture of thickener and base oil have been utilized as a semi-
solid lubricant in a wide range of bearings and gears. Thickeners form fibrous micelles
that can hold the base oil. The behavior of the thickener micelles around the contact
zone, which affects the tribological property of grease, is not clear. In order to make
clear the role of thickener, the lubricant film formed on a specimen is analyzed after
lubrication test by micro-FTIR [20]. The concentration of thickener can be estimated
230 S. Mori

Fig. 7.16 In situ IR spectra and distribution of thickener concentration of Li-grease (a) and urea-
grease (b)

from the absorbance ratio of N–H peak at 3300 cm−1 for urea-type thickener and C=O
peak at 1580 cm−1 for soap-type thickener to the absorbance of C–H from the base
oil, respectively. Figure 7.16 shows the increase in the concentration of the urea-type
thickener at the contact zone during lubrication test [21]. The concentration behavior
depends on the molecular structures of thickeners, i.e., soap-type and urea-type.
Concentration of urea-type thickener increases more easily than that of soap-type
thickener as shown in Fig. 7.16a, b, respectively.
In situ observation of thickener concentration during EHL test reveals that the con-
centration increases gradually with lubrication time under the lubricating condition,
but the concentration decreases steeply by shearing (Slide Roll Ratio: SRR  70%)
under the sliding condition (Fig. 7.17) [22]. This result suggests that a boundary film
of thickener is formed again at the sliding contact even when the film is removed with
friction. The thickness of the boundary film is determined by the balance between
formation and removal of the film. Increase in concentration of amide as an additive
is also observed during lubrication of a gel-like lubricant with dialkyl amides [23,
24].
7 In Situ Observation of Lubricating Films by Micro-FTIR 231

Fig. 7.17 Formation and removal of thickener film and effect of shearing on film formation of urea
grease

7.7.3 Paper Friction Material (Wet Clutch)

One of the important factors to control the friction property of paper friction plates
for wet clutches is the lubricant component between the paper material and the steel
plates. The lubricant component cannot be analyzed by transmission or reflection
IR spectroscopy, because the IR beam does not reflect on paper material. ATR was
applied to analyze the additive concentration in PAO or paraffinic mineral oil (P)
between the paper material and the diamond of the ATR unit as shown in Fig. 7.18a
[25]. It is interesting that the concentrations of polybutenyl saccimide (Fig. 7.18b)
and N-oleoylsulcosine (Fig. 7.18c) increase at 10 N of load under the effects of high
and low sliding speeds, respectively. The results of additive concentrations affect the
friction behavior of wet clutch with the additives. Water film between a hydrogel and
a diamond of ATR unit can be observed by ATR method [26].

7.7.4 Wire Drawing

In metal working processes, such as cutting, drawing, ironing, and rolling, unique
lubricants are demanded for high-energy efficiency, good finishing of metal surface,
and higher reliability of the process. Because lubricity is controlled by the component
and structure of the lubricant film between the metal and the working tool, in situ
232 S. Mori

Fig. 7.18 Effect of contact conditions on concentration of additives at the interface of paper material
for wet clutch. Additive; PBSI: polybutenyl succimide, OS: oleoylsarcosine, Base oil; P: paraffinic
mineral oil, PAO: polyalphaolefin

Fig. 7.19 Photo (a) and a schematic cross section (b) of diamond die in a holder

observation with IR has been applied to investigate the lubricant film between the
metal and the working tool. Figure 7.19 shows a wire drawing die made of diamond
[27]. The holder has a side window for an IR beam which reflects on a metal wire
surface. The IR spectrum of the lubricant film can be taken during drawing with
the die. Reduction of brass wire was carried out with oleic acid in PAO as a base
oil. The oil film thickness, estimated from the absorbance of C–H vibration, linearly
increases with the distance of the reduction region of the die. It is found that metal
oleate, which is formed tribochemically from the oleic acid and brass during drawing,
accumulated at the entrance of the bearing region of the die [28].
7 In Situ Observation of Lubricating Films by Micro-FTIR 233

7.8 Future Work

There are many analytical tools to investigate lubricating films which affect tribo-
logical properties. Among them, IR spectroscopy is powerful for in situ analysis of
lubricating films, especially for tests in air. However, IR spectroscopy has several dis-
advantages, such as limited sensitivity, spatial resolution, and chemical information.
IR spectroscopy is more suitable to analyze organic compounds, such as lubricant
components, rather than inorganic compounds, such as tribofilm formed from addi-
tives. Surface Enhanced Infrared Reflection Absorption Spectroscopy (SEIRAS) has
been applied to chemical analyses of nano-film for electrochemistry and catalysis
[29], and can be applied to the chemical analysis of lubricant films.
Because Raman spectroscopy can analyze many inorganic compounds, such as
MoS2 , at high spatial resolution of 1 µm, it has been applied for in situ observation of
lubricant films [30–32]. The advantage of Raman spectroscopy is the use of a glass
plate as a window instead of IR transparent materials such as diamond and calcium
fluoride for tribology test equipment. A combination of IR and Raman spectroscopies
can be applied as powerful tools to investigate a wide range of lubricating films
through in situ observation of lubricating films during EHL and boundary lubrication
conditions.
A new optical technique for in situ observation is SFG spectroscopy, which is
so sensitive enough to analyze the chemical structure of monolayer films with high
spatial and time resolutions [33]. SFG measurement of polyethylene reveals sur-
face segregation of organic additives in polymers, which affects surface mechanical
properties such as friction [34]. SFG was applied to investigate lubricating interface
of n-dodecane, n-C12 H26 , with stearic acid under shearing condition. A strong peak
of CH2 -ss (symmetric stretching vibration) is observed at 2840 cm−1 , only with
stearic acid containing oil under dynamic conditions (Fig. 7.20). When deuterated
n-dodecane (n-C12 D26 ) was used as a base oil, the peak at 2840 cm−1 disappears.
SFG spectra reveal that the interfacial structure of stearic acid changes negligibly, but
n-dodecane is aligned on the adsorbed film of stearic acid under shearing condition
[35]. To understand tribological phenomena fundamentally, a suitable combined sys-
tem of SFG spectrometer and a proper tribotester will provide fruitful information on
the chemical structures of lubricant films under dynamic conditions through in situ
observation of tribo-interface at high spatial and time resolutions.
234 S. Mori

Fig. 7.20 Molecular

orientation of base oil,
n-dodecane at the interface
between oleic acid adsorbed
layer and lubricant observed
by SFG spectroscopy

References

1. H.A. Spikes, In situ methods for tribology research. Tribol. Lett. 14(1), 1 (2003)
2. G.J. Johnston, R. Wayte, H.A. Spikes, The measurement and study of very thin lubricant films
in concentrated contacts. STLE Tribol. Trans. 34(2), 187 (1991)
3. J. Lauer, K. Vincent, Fourier emission infrared microspectrophotometer for surface analysis.
I. Application to lubrication problems. Infrared Phys. 19(3–4), 395 (1979)
4. P.M. Cann, H.A. Spikes, In lubro studies of lubricants in EHD contacts using FTIR absorption
spectroscopy. Tribol. Trans. 34(2), 248 (1991)
5. P.M. Cann, H.A. Spikes, In-contact IR spectroscopy of hydrocarbon lubricants. Tribol. Lett.
19(4), 289 (2005)
6. G. Socrates, Infrared and Raman Characteristic Group Frequencies, Table and Charts, 3rd
edn. (Wiley, New York, 2004)
7. K. Takiwatari, H. Nanao, S. Mori, Effect of high pressure on molecular interaction between
oleic acid and base oils at elastohydrodynamic lubrication contact. Lubr. Sci. 22, 89 (2010)
8. K. Takiwatari, H. Nanao, E. Suzuki, S. Mori, Stabilization of hydrogen bonding in polypropy-
lene glycol at EHL contact region. Lubr. Sci. 22(9), 367 (2010)
9. Y. Hoshi, N. Shimotomai, M. Sato, S. Mori, Change of concentration of additives under EHL
condition. J. Japanese Soc. Tribologists 44(9), 736 (1999)
10. K. Takiwatari, H. Nanao, I. Minami, S. Mori, The interaction between oleic acid and base oils
at elastohydrodynamic lubrication. J. Japanese Soc. Tribologists 54(1), 48 (2009)
11. R.W.M. Wardle, R.C. Coy, P.M. Cann, H.A. Spikes, An ‘in lubro’ study of viscosity index
improvers in end contacts. Lubr. Sci. 31(1), 45 (1990)
12. K. Takiwatari, H. Nanao, Y. Hoshi, S. Mori, Molecular interaction originating from polar
functional group in lubricants and its relationship with their traction property under elasthy-
drodynamic lubrication. Lubr. Sci. 27, 265 (2015)
7 In Situ Observation of Lubricating Films by Micro-FTIR 235

13. S. Mori, H. Iwata, Relationship between tribological performance of liquid crystals and their
molecular structure. Tribol. Int. 29(1), 35 (1996)
14. R. Lu, S. Mori, H. Tani, N. Tagawa, S. Koganezawa, Tribol. Int. 113, 36 (2017)
15. F. Mangolini, A. Rossi, N.D. Spencer, Tribol. Lett. 45, 207 (2012)
16. F.M. Piras, A. Rossi, N.D. Spencer, Growth of tribological films: In situ characterization based
on attenuated total reflection infrared spectroscopy. Langmuir 18, 6606 (2002)
17. N. Shimotomai, S. Mori, In situ observation of the lubrication film I O/W emulsion by mocro
FT-IR. J. Japanese Soc. Tribologists 56(1), 47 (2011)
18. Y. Shitara, S. Yasutomi, S. Mori, Direct observation of W/O emulsion in concentrated contact
by FT-IR microspectroscopy. J. Japanese Soc. Tribologists 55(10), 736 (2010)
19. Y. Hoshi, I. Minami, S. Mori, Change in concentration of water-glycol hydraulic fluid near the
EHL contact region–Observation by micro FT-IR-. J. Japanese Soc. Tribologists 49(11), 878
(2004)
20. P.M. Cann, Grease lubrication of rolling element bearings—Role of the grease thickener. Lubr.
Sci. 19(3), 183 (2007)
21. Y. Hoshi, K. Takiwatari, H. Nanao, H. Yashiro, S. Mori, In situ observation of EHL films of
greases a micro infrared spectroscopy. J. Japanese Soc. Tribologists 60(2), 153 (2015)
22. Y. Hoshi, K. Takiwatari, H. Nanao, S. Mori, In situ observation of transient responses in grease
lubrication by a micro infrared spectroscopy. J. Japanese Soc. Tribologists 61(11), 784 (2016)
23. K. Takahashi, Y. Shitara, S. Mori, Direct observation of thermo-reversible gel-lubricants in
EHL by FT-IR micro-spectroscopy. Tribol. Online 3(2), 131 (2008)
24. K. Takiwatari, Y. Hoshi, H. Nanao, T. Hojo, S. Mori, In situ observation of lubricant film of
semi-solid lubricants at EHL contact using micro-FTIR. Tribol. Online 11(2), 346 (2016)
25. T. Ichihashi, M. Kudo, S. Mori, Relation between the friction characteristics of wet clutches
and the concentration of additives obtained by in-situ observation of oil film. J. Japanese Soc.
Tribologists 58(8), 581 (2013)
26. Y. Hoshi, K. Takiwatari, H. Nanao, H. Yashiro, M. Wada, H. Furukawa, S. Mori, In situ
observation of hydrogel frictional interface by a micro FTIR-ATR spectroscopy. J. Japanese
Soc. Tribologists 60(1), 68 (2015)
27. H. Nanao, Y. Hoshi, T. Shizuku, K. Takiwatari, S. Mori, Direct observation of lubricant compo-
nents between wire and diamond die for wire drawing with a micro-FTIR. Tribol. Lett. 60(1),
12 (2015)
28. Y. Hoshi, H. Nanao, K. Takiwatari, S. Mori, T. Shizuku, In situ observation of lubricant film
in a diamond die for wire drawing by micro-FTIR. Tribol. Online 11(2), 88 (2016)
29. M. Osawa, In-situ surface-enhanced infrared spectroscopy of the electrode/solution interface,
in Electrochemical Science and Engineering, ed. by R.C. Alkire, D.M. Kolb, J. Lipkowski,
P.N. Ross, p. 269 (2006)
30. C.U. Amanda Cheong, P.C. Stair, In situ measurements of lubricant temperature and pressure
at a sliding contact. J. Phys. Chem. 111, 11314 (2007)
31. S. Zhang, Y. Liu, J. Luo, In situ observation of the molecular ordering in the lubricating contact
area. J. Appl. Phys. 165, 014302 (2014)
32. H. Okubo, S. Sasaki, In situ Raman observation of structural transformation of diamond-like
carbon films lubricated with MoDTC solution: Mechanism of wear acceleration of DLC films
lubricated with MoDTC solution. Tribol. Int. 113, 399 (2017)
33. G.A. Somorjai, K.C. Chou, M. Yang, Sum frequency generation vibrational spectroscopy char-
acterization of surface monolayers: Catalytic reaction intermediates and polymer surfaces. J.
Surf. Sci. Nanotech. 2, 106 (2004)
34. D.H. Gracias, D. Zhang, Y.R. Chen, G.A. Somorjai, Surface chemistry-mechanical property
relationship of low density polyethylene: IR+visible sum frequency generation spectroscopy
and atomic force microscopy study. Tribol. Lett. 4, 231 (1998)
35. S. Watanabe, M. Nakano, K. Miyake, S. Sasaki, Analysis of the interfacial molecular behavior
of n-dodecane containing stearic acid under lubricating conditions by sum frequency generation
spectroscopy. Langmuir 32, 13649 (2016)
Chapter 8
Micro-scale Real-Time Wear Dynamics
Investigated by Synchroton Radiation

M. Belin, Jean-Michel Martin, J. Schou, I. L. Rasmussen, R. Feidenhans’l,

T. Straasø and N. J. Mikkelsen

Abstract In situ wear measurements on a hard coating of TiAlN and CrN layers
deposited on vitreous carbon have been carried out with synchroton radiation. The
results show that wear dynamics can be successfully monitored on a lateral microm-
eter scale and with a submicrometer depth resolution. The wear process is highly
irregular and the local wear rate may vary strongly from one position to another in
the same wear track. Most of the ridges and grooves are generated within the first
500 nm and exist over several micrometers.

Friction and wear play an overwhelming role in our everyday life from motion of
human bodies, car driving, tool handling and manufacturing [1]. Nevertheless, since
the contact between two macroscopic bodies is of extremely complex nature, the
basic mechanisms at macroscopic level are generally described on empirical basis

M. Belin (B) · J.-M. Martin

Laboratoire de Tribologie et Dynamique des Systèmes, CNRS UMR 5512,
Ecole Centrale de Lyon, 69134 Ecully, France
e-mail: michel.belin@ec-lyon.fr
J.-M. Martin
e-mail: jean-michel.martin@ec-lyon.fr
J. Schou · I. L. Rasmussen · T. Straasø
DTU Fotonik, Technical University of Denmark, Risø Campus, 4000 Roskilde, Denmark
e-mail: josc@fotonik.dtu.dk
I. L. Rasmussen
e-mail: inge.lise.rasmussen@regionh.dk
T. Straasø
e-mail: straasoe@gmail.com
R. Feidenhans’l · T. Straasø
Niels Bohr Institute, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen,
Denmark
e-mail: robert@nbi.ku.dk
N. J. Mikkelsen
CemeCon AS, Lokesvej 5, DK–8230 Åbyhøj, Denmark
e-mail: njm@lindberg.com

© Springer Nature Switzerland AG 2018 237

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_8
238 M. Belin et al.

[2–4]. On the nanoscale, the atomic force microscope (AFM) has greatly advanced
the investigation of wear. AFM is ideally suited to gain insight in nanoscale wear
as the tip can act as a counterpart and simultaneously yield detailed information
of adhesion of atoms to the tip and single asperity contact in vacuum experiments
[5–7]. In situ fast scans over micrometer areas of the size up to 180 µm2 have been
demonstrated as well [3]. In contrast, macroscopic studies of the surface topography
during wear have been carried out primarily with ex situ images from scanning
electron microscopy (SEM) and energy dispersive X-ray (EDX) measurements [8,
9]. These techniques hardly allow us to scan a wear track continuously during the
material removal or modification by the counterpart, and profilometry at the mm-
scale has been conducted [10]. However, the gap between nanoscale AFM-based wear
and real-time wear dynamics on a micrometer scale is practically unexplored. In the
present work, for the first time, we have measured the wear thanks to a focused X-ray
beam of high brilliance, that enables us instantaneously to observe the diversity of
wear features through different layers of nitrides with micrometer lateral resolution.
The extraordinary combination of synchrotron radiation and controlled wear gives
us a non-invasive depth resolution of ~100 nm, which is unique compared with
macroscopic wear experiments.
Removal of material during abrasive wear by a repeated sliding motion of a coun-
terpart on a solid generates a track with grooves and ridges on the micrometer level.
These features are not stationary, but change dynamically under the wear process
[10]. In addition, debris from one microstructure can be transferred to other pits or
grooves [11, 12]. Systematic macroscopic investigations of wear are complicated
because of the statistical fluctuations and because of the comparatively small size
(~µm) of the features. Our investigations tell us that once a ridge or a groove of width
(>10 µm) larger than the grain size in the counterpart are created, their chance of
surviving through several micrometers through layers of different nitrides is large.
The interfaces do not significantly disturb the propagation of any of the features.
However, a significant fraction of features also disappears within the wear of 1 µm
depth.
We have designed a coating with a marker layer of CrN which gives us a signifi-
cant change in X-ray transmission during thickness loss, and has a significant peak
in the XANES (X-ray absorption near-edge structure) spectrum, which disappears
when the chromium is removed from the sample. The wear of the samples was car-
ried out in atmospheric air by a nomad instrument, a portable linear reciprocating
tribometer constructed for in situ experiments [13], which keeps a constant normal
force on the counterpart of 3 N during reciprocating motion. The counterpart itself
is a hemispherical molded resin pad with embedded diamond grains (Fig. 8.1). The
grains are of size of 2–4 µm (from Diprotex diamond powder) dispersed in a two-
component resin (Mecaprex KM-EM and KMR). This mixture produces a wear rate
of the hard coating on a substrate within a reasonable period of time. A typical wear
time for removing all nitride layers is 6–10 h corresponding to 60,000–80,000 cycles
including intermissions for measuring. The counterpart was moving back and forth
over the 1.5 mm long track with an average velocity of 15 mm/s and a repetition rate
of 5 Hz.
8 Micro-scale Real-Time Wear Dynamics Investigated by Synchroton … 239

Fig. 8.1 Experimental setup

for sample with a 3-layer
coating. The thickness of the
coatings is exaggerated for
clarity (TiAlN, blue, and
CrN, grey). The counterpart
with the wear pad (white)
was mounted on vitreous
carbon as well. The
absorption measurements
including the XANES
spectra were performed,
when the counterpart was
kept outside the beam line as
shown in Fig. 8.1

The measurements were carried out at beam line ID 24 at the European Syn-
chrotron Radiation Facility (ESRF) [14] with a beam of a lateral spatial extension of
5 × 8 µm2 FWHM (Full Width at Half Maximum) to pursue a detailed record of wear
features at a half-cylinder-like track of length 1.5 mm and width 0.8 mm. The worn
sample is kept at exactly the same position, while the counterpart with was moving
back and forth in a reciprocating motion (Fig. 8.1). The counterpart was moved to
the end of the track every time X-ray transmission and XANES spectra should be
recorded.
The substrates were made in vitreous carbon disks (thickness 1.0 ± 0.1 mm, disk-
shape of diameter 18 ± 0.1 mm) from SIGRADUR G. Subsequently, a hard coating
of nitride layers of thickness from 0.5 to 1.5 µm were deposited on the carbon discs
with PVD at CemeCon Scandinavia [15, 16]. The hard coating consisted of a top-
layer of t 1  1.51 ± 0.07 µm TiAlN (titanium aluminum nitride) on a layer t 2 
0.57 ± 0.03 µm of CrN (chromium nitride), which in turn covers a bottom layer of
t 3  1.27 ± 0.03 µm TiAlN (Fig. 8.1). The precise thickness of the layers indicated
above was determined from a cross section in a Scanning Electron Microscope.
The wear depth was monitored by the transmission in the region from 5.996
to 6.037 keV (see the indication in Fig. 8.2) in 80 track positions, of which those
from 30 to 50 were in the bottom of the track. The instantaneous coating thickness
was determined from the transmitted intensity I from the standard expression of
absorption:

I  I0 exp(−µ1 t1 ) exp(−µ2 t2 ) exp(−µ1 t3 ) exp(−µ4 t4 ) (8.1)

where I0 is the initial intensity, µ1 the absorption coefficient of TiAlN, µ2 the coeffi-
cient of CrN and µ4 the coefficient of vitreous carbon, available in [17]. The dominant
part of the absorption takes place in the relatively thick carbon, such that only 0.218
of the intensity is transmitted through the carbon disc. The transmission through
the coated sample rises from 0.142 at the very beginning until all three coatings are
240 M. Belin et al.

Fig. 8.2 XANES (X-ray absorption near edge spectrum) around the Cr K-edge. The red spectrum
with the peak for the edge at 6.007 keV is taken for a position outside the wear track, while the
blue one in the middle of the track is recorded for each 10,000 cycles. The hatched area indicates
the region in which the absorption measurements were taken

removed during the wear, which is seen when the transmission reaches the value of
0.218. The monitoring of the wear thus relies on a sufficiently distinct transmission
change (in our case 0.076), which is limited by the fact that the carbon disc has to
be so thick that it can sustain the wear from the counterpart without breaking or
bending.
An independent control of the wear depth was acquired with XANES (X-ray
absorption near edge spectra) around the chromium K-edge (Fig. 8.2) between 5.955
and 5.119 keV. The K-edge of chromium in the nitride binding (CrN) is located about
10 eV higher than the edge of metal chromium—as also reported in the literature
[18]). As seen from Fig. 8.2, once the first interface with the CrN layer is reached,
the spectrum starts to decrease, and the second interface is reached when the absorp-
tion has vanished. The final depth profile of the tracks was controlled by ex situ
profilometry with a Dektak instrument.
The cross section of the wear track with increasing depth during wear is shown in
Fig. 8.3 as a function of the number of wear cycles. The wear track is about 800 µm
wide and monitored with 80 positions across the track ranging from the unworn part
on one side to the unworn part on the other side with about 25 positions in the track.
Since the distance between each position is less than 20 µm, it means that we identify
solely the features on a microscale level. On the other hand, these are the most severe
ones for wear control, while asperities of the order of few µm usually are of minor
importance.
8 Micro-scale Real-Time Wear Dynamics Investigated by Synchroton … 241

Fig. 8.3 Instantaneous wear cross section of the track. The lines indicate the instantaneous depth
after the number of wear cycles for 25 positions across the track. The depth is calculated from
(8.1) starting from the surface at 1003.4 µm. The TiAlN-carbon interface is at 1000 µm. “A” is an
example of a ridge, “B” a groove and “C” a ridge which changes to a groove during the wear. The
distance between each position is 13.5 µm

Within the track there are twice as many ridges as grooves that persist after wear
of more than 2 µm. Most of them are more than 0.5 µm higher or deeper than the
surrounding level. For each of these major features there is approximately one short-
living feature that persists for less than 1 µm depth. In all cases, the interfaces do
not play a significant role, but there is a tendency to extend wear in the second layer.
The stochastic nature of the wear process is demonstrated in Fig. 8.4, which
shows the instantaneous depth as a function of cycles for three different positions
(36, 37 and 42) in the middle of the track (Fig. 8.3). The behavior in each individual
position is highly irregular and far from any linear dependence. Initially nothing
happens, but after 10,000 cycles the wear starts with a more and less regular depth
rate with pronounced deviations from a regular behavior. In some cases the thickness
increases locally—probably because of transport of debris [11, 12], which is carried
into the position by the counterpart. Generally, in the upper layer of TiAlN there is no
substantial difference between the maximum value (0.074 ± 0.01 nm/cycle) and the
minimum value (0.065 ± 0.02 nm/cycle) of the wear depth rate, defined as the first
time and last time the interface between the upper layer and the CrN layer is reached.
This value agrees actually very well with the control value (0.063 ± 0.01 nm/cycle)
obtained from the K-edge of the XANES spectrum. The irregular behavior of the wear
is even more pronounced in the CrN layer, and each position behaves differently. For
the blue squares (position 42) the CrN is worn out in a few thousands cycles, while for
green triangles (position 37) the wear rate is more and less constant. However, for the
242 M. Belin et al.

Fig. 8.4 Instantaneous depth for three positions in the track. Red position (36), green position (37)
and blue position (42). The wear starts on the surface at 1003.4 µm and with the carbon-TiAlN
interface at 1000 µm. The surface and the two other interfaces are shown in the figure. The depth in
the vitreous carbon is indicated in an equivalent TiAlN absorption depth, since the exact composition
of bottom of the grooves, of debris of TiAlN and carbon, is not known

red circles (position 36) the wear seems to stop around the interface for almost 15,000
cycles, which presumably indicates slow wear and continuous supply of debris. Once
the wear has reached the lower TiAlN layer, the wear process continues more or less
regularly with similar slope. At the interface between the TiAlN layer and the carbon
substrate there is a large amount of material removed within 1000 cycles except for
the green position that exhibits large jumps presumably resulting from mass transfer
out of and into the position. The depth below the initial interface between the bottom
layer of TiAlN and vitreous carbon is indicated in equivalent TiAlN depth, since
the groove contains an unknown mixture of carbon and debris of TiAlN. The latter
component is clearly observed by SEM as well.
We have shown that wear dynamics can be monitored on a lateral microm-
eter scale using an X-ray beam in the synchrotron radiation. The wear pro-
cess is highly irregular and the local wear rate varies strongly from one posi-
tion to another in the wear track. The ridges, grooves as well as the trans-
port of debris from the track were for the first time followed on a microme-
ter level, confirming the stochastic nature of this type of wear. Most of the fea-
tures are generated within the first 500 nm and exist over several micrometers.
However, also a significant fraction of grooves and ridges disappear by wear
over less than one micrometer. The results from transmission were supported by
results from XANES measurements of the K-edge of Cr from the-CrN marker
8 Micro-scale Real-Time Wear Dynamics Investigated by Synchroton … 243

layer. The XANES spectra also possess a great potential for identifying oxi-
dation during wear, since the position of the K-edge depends on the chemical
environment.

Acknowledgements The authors thank Cornelius Strohm and other staff members at beam line
ID 24 at ESRF for competent assistance. This work has been supported by the Danish Strategic
Research Council with the NABIIT grant 2106-05-0035.

References

1. B.N.J. Persson, U. Tartaglino, O. Albohr, E. Tosatti, Nat. Mater. 3, 882–885 (2004)

2. K. Holmberg, A. Matthews, Coatings Tribology. Properties, Mechanisms, Techniques and
Applications in Surface Engineering (Elsevier, Amsterdam, The Netherlands, 2009)
3. B. Gotsmann, M.A. Lantz, Phys. Rev. Lett. 101, 125501 (2008)
4. A. Erdemir, C. Donnet, J. Phys. D: Appl. Phys. 39(18), (2006)
5. H. Bhaskaran et al., Nat. Nanotechnol. 5, 181 (2010)
6. A. Filippov, V.L. Popov, M. Urbakh, Phys. Rev. Lett. 106, 025502 (2011)
7. N.N. Gosvami et al., Phys. Rev. Lett. 107, 144303 (2011)
8. Q. Luo, PEh Hovsepian, Thin Solid Films 497, 203–209 (2006)
9. X. Li, C. Li, Y. Zhang, H. Tang, G. Li, C. Mo, Appl. Surf. Sci. 256, 4272 (2010)
10. Q. Luo, Tribol. Lett. 37, 529 (2010)
11. N. Hiraoka, E. Yamane, Tribol. Lett. 41, 479 (2011)
12. M. Duarte, I. Vragovic, J.M. Molina, R. Prieto, J. Narciso, E. Louis, Phys. Rev. Lett. 102,
045501 (2009)
13. L. Joly-Pottuz, J.-M. Martin, M. Belin, F. Dassenoy, G. Montagnac, B. Reynard, Appl. Phys.
Lett. 91, 153107 (2007)
14. S. Pascarelli, O. Mathon, M. Munoz, T. Mairs, J. Susini, J. Synchroton Rad. 13, 351–358 (2006)
15. I.L. Rasmussen, N. Pryds, H.C. Pedersen, J. Schou, N.J. Mikkelsen, J. Phys. D Appl. Phys. 41,
135307 (2008)
16. I.L. Rasmussen, M. Guibert, J.-M. Martin, M. Belin, N.J. Mikkelsen, H.C. Pedersen, J. Schou,
Tribol. Lett. 37, 15 (2010)
17. www.cxro.lbl.gov
18. S. Kimura, S. Emura, H. Ofuchi, Y.K. Zhou, S. Hasegawa, H. Asahiet, J. Crystal Growth
301–302, 651 (2007)
 Part III
In-Situ Microscopy and Topography
Measurements
Chapter 9
Electrochemical Friction Force
Microscopy

Florian Hausen

Abstract Tribology is a very interdisciplinary area of research where physicists,

engineers, materials scientists and chemists all contribute to obtain a better under-
standing and control of friction and wear of surfaces in sliding contacts. As friction
and wear is strongly affected by the chemical nature of the interacting surfaces
electrochemistry plays an important role in tribology. Within this chapter the use
of established electrochemical methods to control surface chemistry and how this
affects tribological properties will be discussed.

9.1 Introduction

Tribology is a very interdisciplinary area of research where physicists, engineers,

materials scientists and chemists all contribute to obtain a better understanding and
control of friction and wear of surfaces in sliding contacts. As friction and wear is
strongly affected by the chemical nature of the interacting surfaces electrochemistry
plays an important role in tribology. Within this chapter the use of established elec-
trochemical methods to control surface chemistry and how this affects tribological
properties will be discussed.
Let us begin with the following question: When is the study of friction forces in an
electrochemical environment of relevance? One might divide this question into two
parts: On one hand the solid-liquid interface which has been shown to significantly
influence friction and sliding and on the other hand the controlled charging of the

F. Hausen (B)
Forschungszentrum Jülich, Institute of Energy and Climate Research, IEK-9,
52425 Jülich, Germany
e-mail: f.hausen@fz-juelich.de
F. Hausen
Institute of Physical Chemistry, RWTH Aachen University, Landoltweg 2,
52074 Aachen, Germany
F. Hausen
Jülich-Aachen Research Alliance, Section JARA-Energy, Jülich, Germany

© Springer Nature Switzerland AG 2018 247

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_9
248 F. Hausen

interface. Regarding the first part, most sliding and hence, frictional processes occur
at solid-liquid interfaces, especially when liquid lubricated contacts are involved.
Such lubrication was fundamental for the industrialization as it secured long-life of
machinery parts by reducing friction and wear. It is still of utmost importance in
the world we are living in as demonstrated by increasing sales of lubricants. How-
ever, when thinking about smaller scales such as micrometer contacts which are
relevant for microelectromechanical systems (MEMS), different strategies must be
considered as many surfaces exhibit a thin layer of adsorbed water under ambient
conditions. Although this water layer is typically extremely thin, it can significantly
alter the friction and wear properties of the materials present in the sliding contact.
Coming to the question of charges at the interface: When two non-identical sur-
faces are sheared against each other, charge separation might occur. This leads to
building up of opposite charges on the surfaces of the two materials depending on
their electron affinity. The same effect is known from rubbing hair with a balloon.
The separation of charges is still present when removing the balloon from the hair
causing an attractive force and spiky hair as long as there is no conductive medium
in between enabling discharge. It can be easily seen that this effect diminishes as
soon as electrically conductive materials or liquids are present. However, most of
the conventional lubricants in the market today are mineral or synthetic oils and thus
electrically insulating. Buildup of charges in machinery parts can cause severe dam-
ages when the potential becomes high enough that a spontaneous and unforeseeable
discharge occurs. In such a lightning event extreme temperatures and currents can
be reached, destroying the material. Thus, understanding and controlling the effect
of surface charges are of significance for tribological applications. Making use of
established electrochemical methods enables the experimentalist to easily realize
various surface morphologies and structures in a controlled environment. Further, it
has been demonstrated in the history of electrochemistry that electrochemical exper-
iments possess comparable control of surface properties with respect to cleanliness
as in ultra-high vacuum conditions.
The aim of this chapter is to provide the reader with an overview of current
research in electrochemical friction studies. Various effects on friction that have
been studied electrochemically are reviewed and recent research as well as open
questions are discussed. The overall discussion is focused on experiments performed
by means of atomic force microscopy (AFM). Additionally, electrochemical pin-
on-disk experiments are included where appropriate. Questions that are addressed
include: Does an applied potential to a surface change the overall friction of the tribo-
pair directly? Can electrochemical methods be used to control friction and switch
between certain friction regimes? Is the substrate or is the electrolyte more important
for controlling friction and what are the underlying mechanisms in each case?
The chapter is organized as follows: First basic electrochemical principles that are
crucial for understanding the later described tribological experiments are introduced.
Second, a general discussion about electrochemical cell design and the adoption of
electrochemical techniques into standard tribological tests is outlined. Subsequently,
the main part of the chapter is separated into two parts: Following a discussion of tri-
bological experiments performed in aqueous electrolytes, focusing on ion adsorption
9 Electrochemical Friction Force Microscopy 249

and oxidation, ionic liquid electrolytes and their special behavior in electrochemi-
cally controlled friction studies are presented.
Despite the fact that corrosion is undoubtedly an electrochemical process and
similar techniques are used to study the effect, the subject of simultaneous occurrence
of wear and corrosion, known as tribocorrosion [1], will not be addressed in this
chapter.

9.2 Electrochemical Foundations

When an interface between a solid and a liquid is formed the chemical potentials
μi of both phases are typically not equal. The system intends to achieve chemical
equilibrium and as a result an electrochemical double layer (EDL) is formed at the
interface, which consists of charges of opposite sign. In a classical view the interface
is characterized by electrostatic interactions between the ions in the electrolyte and
the electrode. Hence, the ions are attracted towards the electrode surface as close as
possible. According to Helmholtz this picture leads to the formation of two parallel
rigid layers of opposing charges, as depicted in Fig. 9.1. However, the formation
of a hydration shell can hinder ions in the electrolyte from reaching the electrode
surface and force them to remain at a distinct distance from the electrode. The plane
through the center of these molecules is called the outer Helmholtz plane (OHP) and
is defined by the radius of the non-specifically adsorbed hydrated ions. In the case
of a specific adsorption the hydration shell is partially removed and the center of the
ions lies in the inner Helmholtz plane (IHP) defined by the radius of the ions. This
differentiation becomes important in order to understand the influence of different
anions on friction forces.
More developed models of the electrochemical double layer take the thermal
motion of the ions into account. Gouy and Chapman concluded that the double layer
is completely diffuse with no immobilized counter ions on the electrolyte side, i.e.
no IHP or OHP is formed. According to Stern the electrostatic interactions in the
first layer are that strong that this layer is still rigid. Hence, this model describes a
combination of the Helmholtz and Gouy-Chapman model. All of the presented mod-
els of the electrochemical double layer have been developed in order to describe the
behavior of ions in diluted aqueous electrolytes. Therefore, these models only con-
sider interactions between ions and the electrode but neglect the interactions between
ions themselves. Theoretical work regarding classical models has been reviewed by
Henderson and Boda [3].
In addition to the electrochemical double layer, ions, solvent dipoles or molecules
can be adsorbed at the electrode’s surface by van-der-Waals interactions. Depending
on the potential of the electrode such adsorptions can be strengthened, weakened or
completely abolished due to Coulomb interactions. The point of zero charge (PZC)
of an electrode is defined as the potential where no excess charge is accumulated
at the surface or in the double layer. Hence, the exact value of the PZC depends
strongly on the nature of each electrochemical system [4]. Generally, it is found
250 F. Hausen

that the characteristics of the surface have a larger influence than the electrolyte. As
the surface charge is sensitive to the pH of the electrolyte, the PZC can be probed
by approaching two surfaces towards each other in solutions of varying pH-values.
It was shown experimentally that the forces between the two surfaces approaching
each other change from repulsive to attractive at the PZC [5]. The ability to control
friction by adjusting surface charges via the pH is illustrated in Fig. 9.5b.
Despite the successful description of EDL for dilute aqueous electrolytes, no
classical model of the electrochemical double layer is applicable in the case of ionic
liquids [6]. In this special class of electrolytes either a distinct layering of alternating
anions and cations is observed or a checkerboard arrangement of anions and cations is
formed at charged and solid interfaces, exhibiting an exponentially decaying profile
into the bulk liquid. The concept of layering as a generic feature of ionic liquids at
solid and charged interfaces, as shown in Fig. 9.2, was firstly introduced by Mezger
et al. by using X-ray reflectivity measurements [7]. Since its initial observation it
has been regularly observed by Atomic Force Microscopy [8–10] and Surface Force
Apparatus [11, 12]. However, the exact nature of the EDL in ionic liquids is still
controversial [13].
The potential difference E across the EDL, that is between the bulk potentials
of the solution and the electrode is called galvani potential. It cannot be measured

Fig. 9.1 Classical view of the electrochemical double layer (EDL). For details see text. Taken
from [2]
9 Electrochemical Friction Force Microscopy 251

Fig. 9.2 a Scheme of the double layer arrangement of ionic liquids, from [7] and b as typically
observed by AFM, from [9]

directly and a reference electrode with defined potential must be added to the system.
As the electrode with a half-cell potential of 0.00 V at standard conditions (activity of
protons are 1 mol/L) the standard hydrogen electrode (SHE) was chosen. For a wide
variety of electrodes the standard potentials E0 have been measured against SHE and
are tabulated [14]. However, the SHE is not practicable in all experiments and thus a
variety of other reference electrodes are commonly used. In aqueous systems reliable
reference electrodes are the Saturated Calomel Electrode (SCE) exhibiting a standard
potential of +0.24 V and the widely used Ag/AgCl—Electrode in 3 M KCl with E0
= 0.21 V. In electrolytes containing Cu ions, a Cu wire is often used as reference
electrode, resulting in a Cu/Cu2 + reference electrode (E0 = 0.34 V). In non-aqueous
systems potentials are frequently quoted with respect to the ferrocene redox couple
Fc+ /Fc with a nominal shift of +400 mV versus NHE. It is important to note that
reference potentials depend critically on the experimental conditions [15]. A platinum
wire is often introduced as a quasi-reference electrode (QRE) when extreme clean
conditions are required. It is frequently used when the work involves ionic liquids
where aqueous reference electrodes would cause impurities. Such a QRE should
be calibrated against a reference electrode with known potential. However, often
only qualitative experiments are performed and the calibration procedure is omitted
so that the exact reference potential remains unclear. A discussion about reference
electrodes to use with ionic liquids has been published recently by Bonnaud et al.
[16].
Cyclic voltammetry (CV) is one of the most prominent techniques to follow
electrochemical processes [17]. In a CV experiment the potential applied to the
electrode surface, typically being the working electrode (WE), is linearly cycled
between an upper and a lower potential value and then reversed with a certain scan
rate. In order to measure the potential at the WE accurately, the second electrode is
split into the reference electrode (RE) and the counter electrode (CE). This layout
is known as three—electrode—arrangement and has the advantage that no current
252 F. Hausen

Fig. 9.3 Cyclic Voltammograms of a Au(100) and b Au(111) single crystal electrodes in 0.05 M
H2 SO4 . Scan rate: 50 mV s−1 . The difference in the signature of the two crystallographic orientations
is remarkable. For discussions of the corresponding peaks see text

flows through the reference electrode. In a two electrode setup the reference electrode
is simultaneously acting as CE and RE. All electrodes are connected to a potentiostat
which adjusts the measured potential to a predefined desired potential by sending
a current through the counter electrode. In a cyclic voltammogram this current is
plotted as a function of the applied potential as demonstrated in Fig. 9.3 and is
characteristic for each electrochemical system. Each electrochemical process like
the formation of adlayers, oxidation steps, or redox reactions causes a distinct signal
in the CV. It is worth mentioning that the exchange of electrons between electrode
and electrolyte and thus a faradaic current is not necessary for the appearance of
peaks in cyclic voltammetry. Capacitive currents originating from a re-orientation of
the EDL, e.g. due to the lifting of a reconstruction or phase changes of adsorbates,
also generate signals in the respective CV.
As previously mentioned, the peaks in a CV can be related to the processes occur-
ing at the electrode, e.g. the peaks at 1.35 and 1.39 V as well as at 1.40 and 1.57 V
for Au(100) and Au(111) respectively are due to an electrochemical oxidation of the
gold surfaces. Only the first atomic layers are electrochemically oxidized enabling
the possibility to reduce the surface subsequently. Such a reduction corresponds to the
peak at 1.16 V and is very similar for both surfaces, as expected from the amorphous
character of the gold oxide. Individual aspects of CVs for different electrochemical
systems will be discussed where appropriate. A thorough analysis of cyclic voltam-
mograms and the respective underlying reactions in various electrolytes for gold
surfaces can be found in [18]. For further insights into electrochemistry in closer
detail the interested reader is referred to well-written textbooks on this topic [19,
20].
9 Electrochemical Friction Force Microscopy 253

9.2.1 Electrochemical Cell Design

Nowadays, most AFM manufacturers offer liquid cells and electrochemical standard
cells as an optional add-on to their commercial systems. Such cells are often designed
to match the instrumental requirements very well but do not allow for more sophisti-
cated electrochemical conditions like exchange of electrolyte during experimentation
or special electrode geometries, which might be required. This has led to a variety
of home-build electrochemical cells for electrochemical AFM (EC-AFM) measure-
ments. Generally, an electrochemical cell is a container in which electrodes can be
fixed and immersed in the electrolyte. The geometry of the electrodes should ensure
a homogeneous electric field between WE and CE and specified distance between
WE and RE. Thus, the perfect electrochemical cell must be adapted for individual
needs and instrumentation. As a general guideline the counter electrode should be
placed parallel or in form of a ring surrounding the working electrode. Furthermore,
it is advantageous if the CE is larger in size compared to WE and from a material
already used in the setup as contamination might occur by dissolving ions from the
counter electrode. As the counter electrode must only supply currents, the actual
material it is composed of is not critical. Much more care with respect to impurities
must be taken with the reference electrode as it might contain unwanted ions. The
electrode of choice must ensure a stable potential throughout the experiment which
in some cases requires putting the reference electrode in a separate compartment
of the electrochemical cell. Even very small amounts of impurities can contribute
significantly to the overall shape of a cyclic voltammogram. A good example for
this effect is the appearance of Fe2 + /Fe3 + redox-peaks at around 0.5 V in CVs of
Au(111) single crystals caused by the release of iron from the cantilever holder of
the AFM into the solution as recognizable in Fig. 9.7b and c. As cleanliness is cru-
cial in electrochemical experiments any contamination must be avoided. Thus, the
electrochemical cell itself should also be easily cleanable.
A good summary of design principles and how to incorporate them in electrochem-
ical cells for scanning probe microscopy is given by Valtiner et al. who introduced
a very versatile still widely applicable electrochemical cell [21]. In principle, the
same considerations hold for implementing an electrochemical cell in a pin-on-disk
configuration or for implementing EC capabilities into the Surface Force Apparatus.
Pin-on-disk tribometers with included electrochemical capabilities have been real-
ized by Argibay et al. [22] and Ismail et al. [23]. Also examples of electrochemically
modified Surface Force Apparatus for friction studies are realized [24].
A good starting point for friction force microscopy experiments using a com-
mercial AFM under controlled potentials would be to use the manufacturer’s elec-
trochemical cell and adapt it subsequently to individual needs. Labuda et al. went
one step further and introduced an atomic force microscope which was especially
designed for friction force microscopy under potential control with extremely high
resolution down to the atomic scale as depicted in Fig. 9.4 [25].
Even though it might be convenient using a built-in potentiostat provided by the
AFM manufacturer in order to control the WE potential, it is not desired. As elec-
254 F. Hausen

Fig. 9.4 High resolution lateral force map of the desorption of a copper adlayer on Au(111) by
switching the applied electrode potential from 150 to 350 mV at 2.0 nN normal load (Electrolyte:
0.1 M HClO4 + 1.5 mM Cu(ClO4 )2 + 10−5 M HCl, Potentials are given with respect to the Ag/AgCl
reference electrode). The respective structure and difference in lattice dimensions is clearly visible.
From [25]

trochemical control of the sample is in principal separated from the operation of

the AFM, all commercially available potentiostats can be used. The electrochemical
signals can be recorded separately from the AFM environment but being displayed
by using auxiliary input channels as supported by most AFMs. The latter allows the
experimentalist to combine all signals spatially and temporally to precisely reveal the
potential/current—structure/friction relationships for a given system. It is of impor-
tance to realize that the electrochemical signals are typically resulting from the entire
sample (or even include artifacts from the cell design) rather than a local variation as
probed by the AFM tip. Depending on the exact conditions the surface area exposed
to the applied pressure of the tip might exhibit variances in electrochemical behavior
compared to the overall sample. As is the case for most friction force microscopy
experiments, contact mode cantilevers with rather small force constants should be
employed. While it is common to use coated cantilevers in order to increase the reflec-
tivity and gain a more intense signal on the photodiode, particularly if the cantilever
is immersed in liquids, such metallic adlayers (usually aluminum or gold) might be
attacked by the electrolyte resulting in dissolution or delamination of the coating.
In both cases the properties of the cantilever will likely vary over the course of an
experiment, making the interpretation of the results complex. Hence, special care
must be taken when using such cantilevers or only uncoated cantilevers should be
used in EC-FFM experiments. Secondly, it is important to think about the electrical
conductivity of the cantilever. Electrostatic interactions may contribute significantly
to the recorded forces. Such electrostatic interaction influences the overall normal
load which affects the measured friction and can vary as a function of pH for iden-
tical surfaces. In order to diminish electrostatic interaction the ionic strength of the
electrolyte should be rather high (>50 mmol) and the cantilever should be made of
highly doped material enabling dissipation of charges.
9 Electrochemical Friction Force Microscopy 255

Fig. 9.5 a COF between two Pt sheets as a function of applied potential in diluted sulfuric acid.
From [29] b COF between Al2 O3 and Fe2 O3 as a function of pH. The solid line represents the
calculated COF while the points are measured values. For discussion see text. From [30]

9.3 Historical Development

The interest in the role of an applied potential to an electrode on sliding friction,

nowadays consolidated in the term “triboelectrochemistry” [1, 22, 26] dates back to
the pioneering work of Thomas Edison in 1875 [27]. In his patent for an improved
telegraph he described that a potential-induced decomposition of a surface oxide
resulted in a change of friction and thus, the mechanical forces of moving surfaces.
One of the first scientifically motivated studies was conducted in the mid-20th century
when the coefficient of friction (COF) measured between two platinum surfaces
immersed in 0.1 N (or 0.05 M) sulfuric acid was observed to depend significantly
on the applied potential [28]. As depicted in Fig. 9.5a the COF is maximized at
potentials where no ions are adsorbed on the surface (at about 0.35 V vs. NHE). For
both, larger cathodic or anodic potentials, i.e. with adsorption of layers of hydrogen
or oxygen, respectively, the coefficient of friction was observed to decrease.
As early as in 1970 R.B. Waterhouse stated in his review entitled “Tribology and
Electrochemistry” that chemisorbed adlayers on electrodes are expected to influence
the COF more strongly than physisorbed molecules [29]. This statement will be
reviewed in the discussion of anion adsorption onto electrodes. Furthermore, Water-
house concluded, based on the result of a potential-independent COF in the presence
of adsorbed impurities that the nature of the adsorbate is of major importance and
not the applied potential itself.
Kelsall et al. studied the influence of potentials between an Al2 O3 pin on a plate of
various metal oxides as a function of pH [30]. The authors developed a mathematical
equation to relate the electrostatic force between two surfaces to the coefficient
of friction. In Fig. 9.5b the calculated coefficient of friction is shown next to the
measured data for the combination of Al2 O3 on Fe2 O3 as a function of pH. The PZC of
Al2 O3 is at pH  9 and the PZC of Fe2 O3 is at pH  6.3. Hence, for pH values between
the PZCs, a strong attractive force is acting resulting in an “additional” normal load
256 F. Hausen

and thus higher friction. The work clearly demonstrates that electrostatic interactions
are an important aspect in tribology. Xie et al. reviewed the influences on lubrication
under charged conditions by separating them according to different lubrication states
[31]. Controlling surface charges by pH is one way to influence friction, however,
the use of classical electrochemical methods allow for a direct control of surface
charges precisely without changing the composition of the electrolyte.

9.4 Friction Force Microscopy Under Electrochemical

Conditions

9.4.1 Aqueous Electrolytes

Soon after the invention of the atomic force microscope in 1986 by Binnig, Quate
and Gerber [32] the potential of such apparatus for studying friction at the nanoscale
was first demonstrated by Mate et al. [33]. In their ground-breaking work the friction
between a tungsten tip and a graphite surface was measured under ambient conditions,
allowing the authors to visualize the atomic structure of the graphite lattice the friction
signal of the AFM. The same surface was examined for electrochemical friction force
microscopy in 0.1 M NaClO4 solution as pioneered by Binggeli et al. [34, 35]. In these
early electrochemical friction force microscopy (EC-FFM) experiments the friction
at monoatomic high step edges was analyzed and separated into a geometrical and a
chemical contribution. The authors found that the overall friction is predominate by
the chemical component and attributed this to a ploughing effect of adsorbed water
layers under the assumption that the resistance of such near-surface water layers to
ploughing is potential dependent. A similar argument of increased electro-viscosity
has been adopted by Valtiner et al. who found increased friction for mica—gold
contact in 10−3 M nitric acid upon charging the gold surface to 400 mV [24]. The
authors could exclude changes in the separation between the sliding surfaces by
following the absolute distance between mica and gold in situ by means of surface
force apparatus.
Weiland et al. extended the previously described experiments on HOPG in 0.1 M
NaClO4 and analyzed friction at atomic steps under applied potentials of 1.0 and
−1.0 V [36]. While the topography remains unaffected by a change of the potential the
friction force at the step edge is significantly altered. This result was later attributed
to the intercalation of perchlorate ions into the steps of HOPG as a function of
potential [37]. In agreement with this conclusion no change of friction was observed
on the basal plane. Thus, the applied potential itself does not change the sliding
between the AFM tip and the substrate in any form as it was already concluded
by Waterhouse [29]. Weiland et al. also found that friction increased independent
whether the tip slips up or down the step [36]. Similar observations have been made
for monoatomic high steps on Au(111) in sulfuric acid [38] and for steps up to 5
layers in height on graphite, MoS2 and NaCl under ambient conditions [39, 40].
9 Electrochemical Friction Force Microscopy 257

The effect is attributed to the Schwoebel-Ehrlich barrier which is known to hinder

diffusion across step edges [41, 42]. This impressively illustrates that friction is
very sensitive to surface effects. However, if the step height becomes considerably
larger than atomic dimensions, topographic influences dominate as suggested by
Sundarajan et al. [43]. HOPG is among the best examined system in electrochemical
friction studies beside gold surfaces. The reason for the popularity of gold lies in its
well-known electrochemical behavior which offers numerous possibilities to alter
the surface properties in situ [44–46].

9.4.1.1 Deposition of Ionic Adsorbates on Surfaces

One advantage of electrochemistry is the possibility to adsorb (sub)monolayers of

cationic or anionic adsorbates onto electrodes in a very controlled way by the applied
potential [44, 47]. For studying fundamental friction processes, this effect can be used
to deposit and alter surface films in a distinct way.
As discussed in the electrochemical foundation section anions can adsorb onto an
electrode at various potentials in the inner (IHP) or outer Helmholtz plane (OHP).
Even at a negative electrode anions can adsorb due to attractive van der Waals inter-
actions. The effect of anions being adsorbed in the IHP or OHP on friction has first
been described by Kautek et al. who studied the adsorption of halogenides on silver
by means of friction force microscopy [48]. Only specifically adsorbing anions in
the IHP, such as bromide and iodide, showed an influence on friction while fluoride
as weakly bound species in the OHP cause no significant change in friction. The
authors assumed an easier displacement of the weakly bound anion by the AFM
tip compared with the greater required force to displace specifically adsorbed ions,
which hinder the movement of the tip, enhancing dissipation.
Another example of a strongly binding anion in the inner Helmholtz plane is the
adsorption of sulfate onto gold electrodes which is a well-studied system in elec-
trochemistry [45, 49]. In order to yield a more comprehensive view of the situation
−
the sulfate ion SO2− 4 and the isoelectric perchlorate ion ClO4 is compared. Both
ions are used in standard electrochemical electrolytes. The adsorption of sulfate onto
Au(111) can be monitored by cyclic voltammetry as depicted in Fig. 9.6. The black
curve shows the same CV as in Fig. 9.3a while the red curve is a magnification of
the potential region between 0.0 and 0.9 V versus Ag/AgCl. At 0.3 V a first peak in
the anodic scan occurs that can be attributed to the lifting of the reconstruction of the
Au(111) surface. This peak is followed by a very broad peak between 0.4 and 0.7 V
corresponding to the adsorption of sulfate onto the electrode in a random fashion.
At 0.85 V a small spike is detected in the CV
√ which is assigned to a phase transition
√ 
of the adsorbed sulfate forming a regular 3x 7 R19.1◦ -adlayer [50] which can
be probed by EC-STM [51] and EC-AFM [52]. The cathodic scan demonstrates the
reversibility of the transitions.
With the adsorption of sulfate an increase of friction was observed at rather high
normal loads of 65 nN [38]. By means of high resolution EC-FFM the same system
258 F. Hausen

Fig. 9.6 Cyclic

voltammograms of Au(111)
in 0.05 M H2 SO4 . Potentials
are given with respect to
Ag/AgCl. The red curve
represents a magnification of
the CV between 0.0 and
0.9 V in order to visualize
the signatures of sulfate
adsorption in CV. For
discussion see text after [52]

Fig. 9.7 a Dependency of friction on applied potential and normal load for Au(111) in 0.05 M
H2 SO4 , potentials are given versus Ag/AgCl. After [52] The arrow indicates the shift of the onset
of friction towards more cathodic potentials with increasing normal load. b Frictogram of Au(111)
in 0.1 M HClO4 at a normal load of 6.5 nN. Friction remains unchanged in a potential regime
between 0.3 and 0.8 V. After [53] c Frictogram of Au(100) in 0.05 M H2 SO4 at a normal load of
nN. Sulfate is less strongly adsorbed compared to the Au(111) surface and no clear effect on friction
is observed

has further been investigated in greater detail [52] and it was revealed that the detec-
tion limit of friction is a complex interplay between applied potential and normal
load as illustrate in Fig. 9.7a. While for low normal loads of 1.0 nN no modulation
of friction resulting from adsorption of sulfate on the gold surface was detected in
the potential regime between 0.2 and 0.6 V, the onset of friction for higher normal
loads is shifted to lower potentials. The effect was attributed to forcing the sulfate
√ √ 
adlayer into the regular 3x 7 R19.1◦ -adlayer under the pressure of the tip. This
conclusion was supported by atomically resolved lateral force maps of the sulfate
super lattice at rather negative potentials at which this structure cannot be resolved
by EC-STM methods [52]. As it can also be seen from Fig. 9.7a the increase of
friction due to the electrochemical oxidation of the gold surface is not influenced by
the normal load and occurs for all loads at about 0.8 V.
How can it be proven that the observed potential dependent characteristics on
friction are indeed a signature of the adsorption state of the anion? The perchlorate
ion is isoelectrical to the sulfate ion but undergoes only a weak adsorption on Au(111).
Labuda et al. present frictograms for Au(111) in 0.1 M HClO4 as shown in Fig. 9.7b
9 Electrochemical Friction Force Microscopy 259

[53]. Throughout this chapter the term frictogram will be used for simultaneously
recorded CVs and friction data if presented in an overlaid diagram such as shown in
Fig. 9.7b and c. The authors cannot find any detectable influence on friction by the
weakly adsorbed perchlorate ion.
Another way of verifying the importance of specific binding of ions to influence
friction is to utilize a Au(100) surface on which the sulfate anion binds less strongly
because of geometrical constraints. When examined on this surface, no influence on
friction was found for sulfate adsorption on Au(100) as revealed by frictograms of
the system as depicted in Fig. 9.7c.
For the case of cationic adsorbates the deposition of copper on gold surfaces is
among the best studied system and act as a model system as it is electrochemically
well understood. Figure 9.8b shows a cyclic voltammogram of the deposition of
copper on Au(111) and the corresponding friction values for various normal loads.
We begin the discussion of the CV at a potential of 0.4 V versus Cu/Cu2+ where a
bare gold surface is present as also indicated by the high-resolution friction force
map shown in Fig. 9.8a. When lowering the potential from 0.4 V a first broad peak
is visible which is assigned to the formation of a submonolayer Cu on Au(111) [54].
It was found that sulfate is co-adsorbed onto this submonolayer [54]. In the friction
force map a change of the structure of the glide plain is visible with a clearly larger
lattice constant (potential regime 0.17–0.03 V). Shortly before 0.0 V versus Cu/Cu2+
a full monolayer of Cu is deposited on the gold electrode.
The frictograms depicted in Fig. 9.8b reveal various interesting characteristics of
the Cu/Au system when sliding against a silicon AFM tip. Comparing the frictograms
at various normal loads, the transitions between distinct friction regimes can be
correlated with the peaks in the cyclic voltammogram. This correlation highlights
that the applied potential to the electrode does not change the overall friction of
the tribo-pair. Modulations of friction are resulted by electrochemically induced
changes of the surface structure or composition. More specifically, friction depends
critically on the adsorbed species (bare gold—submonolayer Cu with coadsorbed
sulfate—monolayer Cu). Friction is found to be low on the submonolayer coverage
until the monolayer is formed shortly before the applied potential reaches 0.0 V. The
formation of the monolayer is accompanied by an increase in friction. The higher
level of friction is maintained when reversing the potential sweep until the monolayer
is dissolved at about 0.07 V. Interestingly, during the transition from submonolayer
coverage to bare gold a very strong increase in friction is observed before the same
friction level as before the electrochemical cycle is reached on bare Au. A clear
explanation for this behavior is still lacking but similar observations of an increase
in friction during electrochemically induced reformation of a bare gold surface can
be found in [38] as well as for the reduction of gold-oxide on Au(100) as shown in
Fig. 9.7c.
When comparing the friction versus normal load behavior of the submonolayer
and the monolayer of Cu on Au, a clear deviation is observed as illustrated in Fig. 9.8c
and d. While friction remains low and almost independent of normal load for the
submonolayer a linear increase of friction with normal load is observed for the full
monolayer coverage. This result indicates that the coadsorbed sulfate plays a vital
260 F. Hausen

Fig. 9.8 a High resolution friction force map of the deposition of Cu on Au(111). The numbers
on the left indicate the applied potential. b Corresponding frictograms recorded at various normal
loads as indicated. Open symbols correspond to the cathodic shift of the potential, closed symbols
correspond to the anodic shift of the potentials. a and b are taken from [55]. c The friction versus
normal load plots of the submonolayer coverage and the monolayer coverage of copper on gold
reveal distinct differences and indicate a strong lubricating effect of the coadsorbed sulfate layer on
the submonolayer of copper on gold

role, acting as an effective lubricant for low loads. Nielinger et al. reported friction
results for Cu deposition on Au(111) for considerably higher normal loads of 30
and 120 nN [56]. A strong increase in friction for all examined copper coverages on
Au(111) was observed. However, the distinction between submonolayer and mono-
layer copper coverage was only possible at 120 nN. This observation was rationalized
under the hypothesis that variations in the measured coefficient of friction as a func-
tion of normal load was likely the result of the AFM tip penetrating the lubricating
sulfate adlayer.

9.4.1.2 Oxidized Surfaces

When studying tribological interactions oxidized surfaces, either formed naturally or

intentionally, are often of particular interest because of their typically high hardness
and wear resistance. Oxides are also easily accessible under electrochemical condi-
tions. Friction studies using electrochemical oxidation has been reported for various
9 Electrochemical Friction Force Microscopy 261

Fig. 9.9 AFM friction

image of an iron surface in
0.02 M NaSO4 (pH  10)
after three potential cycles
between −1.2 and 0.4 V
versus SCE. As clearly
visible a very heterogeneous
surface composition with
respect to the COF arose
after electrochemical
treatment. From [30]

metals like Fe [26], Ta [57], Cu [22] or Au [53]. When iron is immersed in 0.1 M
NaOH electrolyte it can be oxidized in a two-step process to FeOOH which can be
easily followed by CV [26]. While the pure Fe surface exhibits relatively high COF
of 0.15 Zhu et al. find a strong decrease to μ < 0.1 accompanied with a smoothening
of the surface [26]. As shown in Fig. 9.9 the authors find patches of low and high
friction values next to each other across the surface after three voltammetric cycles.
When forming a Fe-octanate rather than the peroxide the authors found only a less
pronounced decrease of the COF to about 0.05 but could follow the growth of a
surface film modifying friction.
This work impressively shows that friction is affected in several directions under
electrochemical control. In addition to the change of the chemical nature of the
surface film, that is being the bare metal, an oxide film or an adsorbant, also the
roughness of the surface can influence the overall friction. The spatial distribution of
a film generated as a result of the chemical reactions mediated by the shearing action
of the two surfaces, a tribofilm, must also be taken into consideration. Labuda et al.
studied the electrochemical oxidation of gold and observed roughening of the surface
upon oxidation as well as a strong increase in friction as compared to the bare gold
surface [53]. As the electrogenerated gold oxide is only in the range of a few atomic
layers the authors could demonstrate an impressive reversibility between low and
high friction states corresponding to bare or oxidized gold surface. High resolution
friction force maps are shown in Fig. 9.10 and reveal the lattice parameters for the
Au(111) surface and a roughened structure of the gold oxide.
The authors attributed the higher friction observed on the gold oxide to its
increased roughness. This experiment illustrates that the structure of the surface
is as important as its chemical nature, which is further underlined by friction experi-
ments by Hausen et al. on Au(100) in sulfuric acid electrolyte [58]. The structure of
the Au(100) surface was electrochemically controlled between a so-called pseudo-
hexagonal phase at low potentials (<0.26 V vs. Ag/AgCl) which corresponds to the
262 F. Hausen

Fig. 9.10 Lateral force map revealing the atomic structure of bare Au(111) (upper part) and after
electrochemical oxidation. While the Au(111) shows a regular lattice, the oxidized gold film exhibits
irregular stick slip pattern, revealing its amorphous character. From [53]

reconstructed form of the electrode and the normally expected quadratic 1 × 1 struc-
ture at potentials positive of 0.26 V versus Ag/AgCl. Upon the transition, which
can be followed by CV, the chemistry of the surface remained unchanged while
the atomic arrangement was alternating between a more open 1 × 1 structure and a
densely packed hexagonal form.
In summary, it has been identified that friction can be controlled electrochemi-
cally by intercalation and adsorption of ions to electrode surfaces while the adsorption
state determines the effect on friction. The chemical nature of surface adlayers con-
tributes to the overall friction and leads to spatially different patterns when adsorbed
heterogeneously. At the same time, such inhomogeneities can introduce roughness
to the sliding surfaces typically correlating with enhanced friction. Roughness plays
a crucial role in controlling friction electrochemically on different length scales from
subtle effects of reconstructed surfaces to significant changes of the roughness when
electrochemically roughening a surface by oxidation. It is striking that friction was
always observed to increase except in the case of electrochemically induced oxida-
tion of iron. The difference in that one case is that the surface was smoothened upon
oxidation forming a more “perfect” surface afterwards, while in all other cases the
surface was roughened. Given that in real applications the expectation of a perfect
surface is rather low indicates that electrochemical methods might open paths to not
only modify but also to reduce friction during the application by targeted surface
finishing.
All the processes mentioned so far clearly demonstrate that friction is extremely
sensitive to subtle changes of the sliding contact which sets the requirements for
a controlled environment. However, for most realistic applications it is typically
unwanted to alter the glide plane dramatically, or it is not feasible to change the
chemistry or structure of the glide plane due to boundary conditions given by the
surrounding. This brings us to the idea of not changing the surfaces in contact through
the electrochemical potential but rather finding a way of controlling the properties
of the electrolyte, or more specifically in tribology the lubricant, by electrochemical
means.
9 Electrochemical Friction Force Microscopy 263

9.4.2 Ionic Liquids as Electrolytes

Ionic liquids (ILs), despite having been known for more than hundred years [59],
are recently gaining a lot of interest in tribology [60–63], in energy storage applica-
tions [64], as solvents in synthesis [65], in catalysis or for cellulose [66], as well as
in electrochemistry [67]. ILs show remarkable properties such as non-flammability,
high resistivity against oxidation and reduction, as well as negligible vapor pressure
and high thermal stability. ILs solely contain ions and hence, they are conductive and
offer a broad electrochemical window. When working with ionic liquids the quan-
tity of available compounds and the endless possibilities in tailoring task-specific
new ionic liquids make a proper selection of the IL ideal for the system challeng-
ing. In the last 15 years the interest of using ILs in tribological applications has
increased due to their special properties and strong film formation capabilities. The
anti-wear and lubrication properties have been attributed to the charged induced
adsorption of anions and cations onto the sliding surfaces, building up a tribofilm.
Most often hydrophobic anions such as Bis(trifluoromethylsulfonyl)imide (Tf2 N) or
Tris(pentafluoroethyl)trifluorophosphate (FAP) are used for EC-FFM experiments in
ILs. Historically the first air-and humidity stable ILs, so-called second generation of
ILs, consist of Tetrafluorobarate (BF4 ) and Phosphorhexafluorate (PF6 ) anions and
imidazolium-based cations [68], and are therefore widely distributed. A drawback of
these ions for tribological applications is their tendency of (partial) hydration when in
contact with water, forming corrosive HF. Table 9.1 illustrates the chemical structure
of common anions and cations for ionic liquids.
Figure 9.11 shows friction force maps of Au(111) in the ionic liquid 1-Butyl-
1-methylpyrrolidiniumtris(pentafluoroethyl)trifluorophosphat ([Py1,4 ]FAP) at −2 V
and 0.0 V versus Pt-QRE. Monoatomic high steps are present at the electrode sur-
face as well as some islands, most probably generated by lifting the herringbone
reconstruction of Au(111), are visible in Fig. 9.11a. The color represents the overall
friction with brighter colors corresponding to higher friction. Upon changing the
electrode potential by 2.0–0.0 V the overall friction becomes larger and the image
gets blurred (Fig. 9.11b). However, the monoatomic high steps on the gold surface
are still visible, clearly indicating that the observed change in friction as represented
by the change in color is caused by changes of the ionic liquid electrolyte. It is
worth mentioning that such a potential step in aqueous electrolytes would cause the
formation of gases due to the limited electrochemical window of water.
As discussed previously ionic liquids form alternating layers of anions and cations
on charged surfaces [69]. Without controlling surface charges Smith et al. showed
that friction between two atomically smooth mica surfaces depended strongly on the
number of layers between the interacting surfaces [70]. As a consequence, various
friction regimes exist for a given load. The ability of ionic liquids to form mechan-
ically stable layers adjacent to charged surfaces and there conductivity, resulting
from the fact that they solely contain ions, makes it straightforward to control the
surface layers by an applied electrode potential. It has been shown by Hayes et al.
that indeed the interfacial structure of ionic liquids is sensitive to the potential of
264 F. Hausen

Table 9.1 Common cations and anions for ionic liquids

Cations

Immidazolium – based [R1R2IM]

Pyrrolidinium – based [PyR1R2]

Anions

Tetrafluoroborate (BF4)

Hexafluorophosphate (PF6)

Bis(trifluoromethylsulfonyl)imide (NTf2; TFSI)

Tris(pentafluoroethyl)trifluorophosphate (FAP)

Fig. 9.11 Friction force maps of Au(111) in [Py1,4 ]FAP at various applied potentials. Monoatomic
high steps are clearly visible at −2 V versus Pt-QRE. The color represents the friction with brighter
colors higher friction. For 0 V versus Pt-QRE the overall friction is clearly increased compared to
−2 V as indicated by the color. However, the surface steps are still visible indicating that the change
in friction is not resulted by a modification of the substrate but rather originated from changes in
the ionic liquid interfacial layer
9 Electrochemical Friction Force Microscopy 265

the electrode [9]. Numerous theoretical works have verified an oscillating nature of
anion and cation layers at charged surfaces, either electrochemically controlled or
naturally charged [69, 71–73].
The first tribological study of ionic liquids by means of EC-FFM was presented
by Sweeney et al. in 2012 where the authors analyzed the friction between a gold
surface and silicon AFM-tips in [Py1,4 ]FAP [74]. A clear relation between the applied
potential to the gold surface and friction was observed. As shown in Fig. 9.12a low
friction was measured for low potentials (i.e. < 1.5 V vs. Pt-QRE) whereas friction
was considerably higher for positive applied potentials. As no other substances were
present in the electrolyte, the authors attributed the effect to changes in the interfacial
structure of the ionic liquid and the intrinsically different lubrication properties of
the respective anions or cations. It is worth mentioning that the cyclic voltammogram
of [Py1,4 ]FAP exhibits various peaks in the cathodic as well as in the anodic scan.
As it can be clearly seen in Fig. 9.12a friction does not change linearly as a function
of applied potential but corresponds to the peaks observed in the electrochemical
signal. Such peaks in the CV are likely caused by re-orientations of ions of the ionic
liquid at the surface of the electrode. Sum frequency generation (SFG) measurements
have shown that the tilt angle between the ionic liquid cation and the surface normal
direction depends critically on the applied potential at the electrode [75, 76]. A
detailed study of the influence of the tilt angle on friction has recently been provided
by Watanabe et al. in a pin-on-plate configuration using various imidazolium-based
ionic liquids, exhibiting different tilt angles due to attractive electrostatic forces [77].
The authors found a strong correlation between tilt angle and friction, where larger
tilt angles resulted in lower coefficients of friction. However, various other factors
of the individual interfacial properties of the tribopair, such as the roughness of the
shear plane or molecular densities of the adsorbed layers and thus variations in ion
mobility, were also discussed and may contributed to the observed friction. This
difficulty is further underlined by experiments conducted by Li and coworkers [78,
79].
In a first set of experiments the authors investigated the influence of
the cation alkyl chain length by studying friction as function of normal
load on Au(111) in 1-ethyl-3-methyl-imidazolium ([EMIM]), 1-butyl-3-methyl-
imidazolium ([BMIM]) and 1-hexyl-3-methyl-imidazolium ([HMIM]) in combina-
tion with the tris(pentafluoroethyl)trifluorophosphate (FAP) anion. The results are
depicted in Fig. 9.12b. In all cases, the highest friction was found at positive applied
potentials, corresponding to an anion enriched interface. The authors attributed this
to a relatively weak interaction of the FAP anion with the gold surface and disordered
coil morphology. Such disorder results in a rougher shear plane of the sliding contact
and thus, high friction. For the stronger adsorbed imidazolium-based cations, the
trend is less clear, which was explained by the influence of the cation chain length
with respect to the tilt angle and the formation of more resistant shear planes. In
Fig. 9.12b also the opposite behavior, that is lower friction at positive potentials, is
shown if the anion is exchanged with iodide. Although the change in shape of the
anion (iodine is more spherical than FAP) seems to be a reasonable explanation for
the observed difference in friction, a clear understanding that this is indeed the phys-
266 F. Hausen

Fig. 9.12 a Frictogram of an Au(111) surface in [Py1,4 ]FAP. Variations of friction are related to
the electrochemical behavior of the IL. The inset shows a lateral force map and the corresponding
line section as indicated at an applied electrode potential of −1.3 V versus Pt-QRE, revealing the
lattice parameters of the underlying gold surface. From [74] b Lateral force as a function of normal
load for various ionic liquid compositions as indicated in the graphs and electrode potentials on
Au(111) surfaces. While for all FAP anion containing ILs the highest friction values are obtained
for positive applied electrode potentials (a–c), this effect is reversed when iodide is used as anion
(d). From [79]

ical mechanism has not yet established. Results from the same research group were
published on the lubricating properties for a sliding an AFM silicon tip on a HOPG
surface immersed in [HMIM]FAP [78]. In stark contrast to the results of [HMIM]FAP
on Au(111), as discussed previously, a superlubricating regime, or a COF of <0.01, is
found for positive applied potentials, where a FAP anion enriched interface would be
expected. Negative applied potentials correspond to higher friction. In this case, the
authors assigned their observation of reduced friction at positive potentials to reori-
entations of ions in the shear plane resulting in a smoother interface in the case of
adsorbed anions. The huge variation in the COF for the same ionic liquid on different
surfaces demonstrate that next to the overall electrochemical control of the layered
structure of the ionic liquid at the electrode interface, direct substrate—ionic liquid
interactions must be taken into account in order to reach a comprehensive view of
ionic liquid lubrication. However, as the electrode potentials are quoted with respect
to Pt-QRE, a direct comparison is limited.
As discussed previously, electrochemically controlled friction in ionic liquids has
not yet been fully understood and it has been identified that numerous effects play a
role. Further experiments are required before general guidelines can be expressed. It
seems undoubtedly that it is very important to understand the various contributions
in ionic liquid nanotribology such as the influence of chain length, cation and anion
type, interaction with the surface, tilt angle, as well as shape of the ions. As it becomes
very ambiguous to investigate the various contributions in experiments separately,
theoretical work are an essential part in contributing to further understanding of the
9 Electrochemical Friction Force Microscopy 267

lubricating properties of ionic liquids. Canova et al. showed that subtle variations of IL
composition lead to different lubrication mechanisms. Comparing the shape of NTf2
and BF4 anions for 1-butyl-3-methyl-imidazolium ([BMIM]) cations, the authors
found larger fluctuations in the case of NTf2 and consequently irregular shearing
dynamics [80]. Fajardo et al. extended the theoretical work towards electrochemical
friction studies and predicted a shift of the shear plane from the solid-liquid interface
into the ionic liquid layers for larger surface charge densities, leading to overall lower
friction [81].
When considering ionic liquid lubrication it is of utmost importance to control
the environmental conditions very precisely. It has been shown that even very small
traces of water changes the interfacial properties of ionic liquids significantly as
well as their frictional behavior [82, 83]. The determination of the water content by
Karl-Fischer titration can be misleading as the amount of water in the near-surface
region and thus, in the region responsible for the tribological behavior, can be strongly
deviating from the measured bulk concentrations. As many experiments studying the
lubrication properties of ionic liquids are performed under ambient conditions, the
water content can vary dramatically between different experiments, even for the same
ionic liquid. Also, due to the very deviating hydrophilicity of ionic liquids composed
of different ions direct comparisons between experiments are challenging.
Thinking about applications, ionic liquid lubrication becomes interesting only in
niche markets because of the currently very high costs of ionic liquids. The interest
is based on the strong surface films formed by ILs which are not repelled from the
contact even under very high loads as demonstrated by atomic force microscopy,
where a last strongly bound layer withstands high pressures of a sharp AFM tip [9].
Secondly, it seems possible to use only atomically thin layers of ionic liquids for
lubrication, making them interesting as additives in conventional lubricants. Further
experimental research is necessary to elucidate the individual contributions to the
overall friction force by help of theoretical studies.

9.5 Summary

Combined electrochemical methods with tribological instruments are a powerful

method to elucidate underlying principles of friction in situ, and support efforts to
gain a more physical and scientific understanding of the origins of friction. Elec-
trochemical friction force microscopy has been shown to significantly contribute to
the understanding of tribological processes and lubrication at solid-liquid interfaces.
Within a single experiment, various factors influencing friction can be investigated
and isolated, including surface morphology, chemical identity as well as interfacial
properties. Furthermore, these properties can be probed in a controlled and reversible
manner. The control of surface charges by electrochemical means on metal surfaces
has been discussed in this chapter. However, electrochemical effects on friction are
not limited to metallic surfaces. While they have been excluded from the current
268 F. Hausen

discussion, the impacts of electrochemistry on friction on other surfaces, such as

ceramic materials, are of great interest [84].
In summary, the applied potential to an electrode contributes considerably to the
nanoscopically and macroscopically observed friction. Various aspects, including
roughness, chemistry, adsorption state, electrolyte composition all influence friction
and make the interpretation of the data challenging. However, given that friction is
extremely sensitive to even subtle changes of the sliding surface, tribological exam-
ination of materials under electrochemical control cannot only be used in order to
understand frictional processes better, but also to study such electrochemical phe-
nomena by following the frictional response upon changing the electrochemical
conditions. With the merging interest in precise control of interfaces, particularly at
very small scales EC-FFM will shed light onto various aspects of nanotechnology,
e.g. analyzing processes in energy storage systems or in the design of novel materials.

References

1. S.Ã. Mischler, Triboelectrochemical techniques and interpretation methods in tribocorrosion:

a comparative evaluation. Tribol. Int. 41, 573 (2008)
2. D.M. Kolb, An atomistic view of electrochemistry. Surf. Sci. 500, 722 (2002)
3. D. Henderson, D. Boda, Insights from theory and simulation on the electrical double layer.
Phys. Chem. Chem. Phys. 11, 3822 (2009)
4. D.M. Kolb, Reconstruction phenomena at metal-electrolyte interfaces. Prog. Surf. Sci. 51, 109
(1996)
5. A.C. Hillier, S. Kim, A.J. Bard, Measurement of double-layer forces at the electrode/electrolyte
interface using the atomic force microscope: potential and anion dependent interactions. J. Phys.
Chem. 100, 18808 (1996)
6. A.A. Kornyshev, Double-layer in ionic liquids: paradigm change? J. Phys. Chem. B 111, 5545
(2007)
7. M. Mezger et al., Molecular layering of fluorinated ionic liquids at a charged sapphire (0001)
surface. Science 322, 424 (2008)
8. R. Hayes, G.G. Warr, R. Atkin, At the interface: solvation and designing ionic liquids. Phys.
Chem. Chem. Phys. 12, 1709 (2010)
9. R. Hayes et al., Double layer structure of ionic liquids at the Au(111) electrode interface: an
atomic force microscopy investigation. J. Phys. Chem. C 115, 6855 (2011)
10. X. Zhang et al., Probing double layer structures of Au (111)-BMIPF6 ionic liquid interfaces
from potential-dependent AFM force curves. Chem. Commun. 48, 582 (2012)
11. S. Perkin, Ionic liquids in confined geometries. Phys. Chem. Chem. Phys. 14, 5052 (2012)
12. R.M. Espinosa-Marzal, A. Arcifa, A. Rossi, N.D. Spencer, Microslips to ‘Avalanches’ in con-
fined, molecular layers of ionic liquids. J. Phys. Chem. Lett. 5, 179 (2014)
13. D. Henderson, J. Wu, Electrochemical properties of the double layer of an ionic liquid using a
dimer model electrolyte and density functional theory. J. Phys. Chem. B 116, 2520 (2012)
14. CRC Handbook of Chemistry & Physics. (CRC Press; Taylor & Francis group, 2017)
15. V.V. Pavlishchuk, A.W. Addison, Conversion constants for redox potentials measured versus
different reference electrodes in acetonitrile solutions at 25 °C. Inorg. Chim. Acta 298, 97
(2000)
16. C. Bonnaud, I. Billard, N. Papaiconomou, E. Chainet, J.C. Lepre, Rationale for the implemen-
tation of reference electrodes in ionic liquids. Phys. Chem. Chem. Phys. 18, 8148 (2016)
17. R.G. Compton, C.E. Banks, Understanding Voltammetry. (World Scientific, 2007)
9 Electrochemical Friction Force Microscopy 269

18. B.E. Conway, Electrochemical oxide film formation at noble-metals as a surface-chemical

process. Prog. Surf. Sci. 49, 331 (1995)
19. C.H. Hamann, W. Vielstich, Elektrochemie, 4. Auflage. (Wiley-VCH, 2005)
20. A.J. Bard, L.R. Faulkner, Electrochemical Methods—Fundamentals and Applications. (Wiley)
21. M. Valtiner, G.N. Ankah, A. Bashir, F.U. Renner, Atomic force microscope imaging and force
measurements at electrified and actively corroding interfaces: challenges and novel cell design.
Rev. Sci. Instrum. 82, 23703 (2011)
22. N. Argibay, W.G. Sawyer, Frictional voltammetry with copper. Tribol. Lett. 46, 337 (2012)
23. M.N.F. Ismail, T.J. Harvey, J.A. Wharton, R.J.K. Wood, A. Humphreys, Surface potential
effects on friction and abrasion of sliding contacts lubricated by aqueous solutions. Wear 267,
1978 (2009)
24. M. Valtiner, X. Banquy, K. Kristiansen, G.W. Greene, J.N. Israelachvili, The electrochemical
surface forces apparatus: the effect of surface roughness, electrostatic surface potentials, and
anodic oxide growth on interaction forces, and friction between dissimilar surfaces in aqueous
solutions. Langmuir 28, 13080 (2012)
25. A. Labuda et al., High-resolution friction force microscopy under electrochemical control. Rev.
Sci. Instrum. 81, 83701 (2010)
26. Y. Zhu, G.H. Kelsall, H.A. Spikes, Triboelectrochemistry on a nanometre scale. Tribol. Lett.
2, 287 (1996)
27. T. Edison, US Patent 158787 (1875)
28. F.P. Bowden, D. Tabor, The Friction and Lubrication of Solids. (Clarendon Press, 2008)
29. R.B. Waterhouse, Tribology and electrochemistry. Tribology 3, 158 (1970)
30. G.H. Kelsall, Y. Zhu, H.A. Spikes, Electrochemical effects on friction between metal oxide
surfaces in aqueous solutions. J. Chem. Soc., Faraday Trans. 89, 267 (1993)
31. G. Xie, D. Guo, J. Luo, Lubrication under charged conditions. Tribology Int. 84, 22 (2015)
32. G. Binnig, C.F. Quate, C. Gerber, Atomic force microscope. Phys. Rev. Lett. 56, 930 (1986)
33. C.M. Mate, G.M. McClelland, R. Erlandsson, S. Chiang, Atomic-scale friction of a tungsten
tip on a graphite surface. Phys. Rev. Lett. 59, 1942 (1987)
34. M. Binggeli, R. Christoph, H.-E. Hintermann, J. Colchero, O. Marti, Friction force measure-
ments on potential controlled graphite in an electrolytic environment. Nanotechnology 4, 59
(1993)
35. M. Binggeli, R. Christoph, H.-E. Hintermann, Observation of controlled, electrochemically
induced friction force modulations in the nano-Newton range. Tribol. Lett. 1, 13 (1995)
36. E. Weilandt, A. Menck, O. Marti, Friction studies at steps with friction force microscopy. Surf.
Interface Anal. 23, 428 (1995)
37. B. Schnyder, D. Alliata, R. Kötz, H. Siegenthaler, Electrochemical intercalation of perchlorate
ions in HOPG: an SFM/LFM and XPS study. Appl. Surf. Sci. 173, 221 (2001)
38. F. Hausen, M. Nielinger, S. Ernst, H. Baltruschat, Nanotribology at single crystal electrodes:
influence of ionic adsorbates on friction forces studied with AFM. Electrochim. Acta 53, 6058
(2008)
39. H. Hölscher, D. Ebeling, U. Schwarz, Friction at atomic-scale surface steps: experiment and
theory. Phys. Rev. Lett. 101 (2008)
40. P. Egberts, et al., Environmental dependence of atomic-scale friction at graphite surface steps.
Phys. Rev. B—Condens. Matter Mater. Phys. 88, 1 (2013)
41. R.L. Schwoebel, E.J. Shipsey, Step motion on crystal surfaces. J. Appl. Phys. 37, 3682 (1966)
42. G. Ehrlich, Atomic displacements in one- and two-dimensional diffusion. J. Chem. Phys. 44,
1050 (1966)
43. S. Sundararajan, B. Bhushan, Topography-induced contributions to friction forces measured
using an atomic force/friction force microscope. J. Appl. Phys. 88, 4825 (2000)
44. E. Herrero, L.J. Buller, H.D. Abruna, Underpotential deposition at single crystal surfaces of
Au, Pt, Ag and other materials. Chem. Rev. 1001, 1897 (2001)
45. J. Lipkowski, Z. Shi, A. Chen, B. Pettinger, C. Bilger, Ionic adsorption at the Au(111) electrode.
Electrochim. Acta 43, 2875 (1998)
270 F. Hausen

46. H. Angerstein-Kozlowska, B.E. Conway, A. Hamelin, L. Stoicoviciu, Elementary steps of

electrochemical oxidation of single-crystal planes of Au Part II. A chemical and structural
basis of oxidation of the (111) plane. J. Electroanal. Chem. Interfacial Electrochem. 228, 429
(1987)
47. O.M. Magnussen, Ordered anion adlayers on metal electrode surfaces. Chem. Rev. 102, 679
(2002)
48. W. Kautek, S. Dieluweit, M. Sahre, Combined scanning force microscopy and electrochemical
quartz microbalance in-situ investigation of specific adsorption and phase change processes at
the silver/halogenide interface. J. Phys. Chem. B 101, 2709 (1997)
49. Z. Shi, J. Lipkowski, Investigations of So42- Adsorption at the Au(111) electrode in the presence
of underpotentially deposited copper adatoms. J. Electroanal. Chem. 364, 289 (1994)
50. F.C. Simeone, D.M. Kolb, S. Venkatachalam, T. Jacob, Die Au(111)-Elektrolyt-Grenzschicht:
eine Tunnelspektroskopie- und DFT-Untersuchung. Angew. Chemie 119, 9061 (2007)
51. D.M. Kolb, Elektrochemische oberflächenphysik. Angew. Chem. 113, 1198 (2001)
52. F. Hausen, N.N. Gosvami, R. Bennewitz, Anion adsorption and atomic friction on Au(111).
Electrochim. Acta 56, 10694 (2011)
53. A. Labuda et al., Switching atomic friction by electrochemical oxidation. Langmuir 27, 2561
(2011)
54. Z. Shi, S. Wu, J. Lipkowski, Coadsorption of metal atoms and anions: Cu upd in the presence
of SO42–, Cl– and Br–. Electrochim. Acta Surf. Struct. Electrochem. React. 40, 9 (1995)
55. Bennewitz, R., Hausen, F., N.N. Gosvami, Nanotribology of clean and modi fi ed gold surfaces.
J. Mater. Res. 28, 1279 (2013)
56. M. Nielinger, H. Baltruschat, Nanotribology under Electrochemical Conditions: influence of
a copper (sub)monolayer deposited on single crystal electrodes on friction forces studied with
atomic force microscopy. PhysChemChemPhys 9, 3965 (2007)
57. D. Huitink, F. Gao, H. Liang, Tribo-electrochemical surface modification of tantalum using
in situ AFM techniques. Scanning 32, 336 (2010)
58. F. Hausen, J.A. Zimmet, R. Bennewitz, Surface structures and frictional properties of Au(100)
in an electrochemical environment. Surf. Sci. 607, 20 (2013)
59. P. Walden, Ueber die Molekulargrösse und elektrische Leitfähigkeit einiger geschmolzenen
Salze. Bull. Acad. Sci. St. Petersbg. 8, 405 (1914)
60. M.-D. Bermudez et al., Ionic liquids as advanced lubricant fluids. Molecules 14, 2888 (2009)
61. A. Somers, P. Howlett, D. MacFarlane, M. Forsyth, A review of ionic liquid lubricants. Lubri-
cants 1, 3 (2013)
62. F. Zhou, Y. Liang, W. Liu, Ionic liquid lubricants: designed chemistry for engineering appli-
cations. Chem. Soc. Rev. 38, 2590 (2009)
63. I. Minami, Ionic liquids in tribology. Molecules 14, 2286 (2009)
64. M. Watanabe et al., Application of ionic liquids to energy storage and conversion materials and
devices. Chem. Rev. 117(10), 7190 (2017)
65. J.P. Hallett, T. Welton, Room-temperature ionic liquids: solvents for synthesis and catalysis. 2.
Chem. Rev. 111, 3508 (2011)
66. A. Pinkert, K.N. Marsh, S.S. Pang, M.P. Staiger, Ionic liquids and their interaction with cellu-
lose. Chem. Rev. 109, 6712 (2009)
67. P. Hapiot, C. Lagrost, Electrochemical reactivity in room-temperature ionic liquids. Chem.
Rev. 108, 2238 (2008)
68. J.S. Wilkes, M.J. Zaworotko, Air and water stable 1-Ethyl-3-Methylimidazolium Based Ionic
Liquids. J. Chem. Soc.—Chem. Commun. 965 (1992)
69. M.V. Fedorov, A.A. Kornyshev, Ionic liquids at electrified interfaces. Chem. Rev. 114, 2978
(2014)
70. A.M. Smith, K.R.J. Lovelock, N.N. Gosvami, T. Welton, S. Perkin, Quantized friction across
ionic liquid thin films. Phys. Chem. Chem. Phys. 15, 15317 (2013)
71. M.V. Fedorov, A.A. Kornyshev, Ionic liquid near a charged wall: structure and capacitance of
electrical double layer 112, 11868 (2008)
9 Electrochemical Friction Force Microscopy 271

72. C. Merlet, B. Rotenberg, P.A. Madden, M. Salanne, Computer simulations of ionic liquids at
electrochemical interfaces. Phys. Chem. Chem. Phys. 15, 15781 (2013)
73. M.Z. Bazant, B.D. Storey, A.A. Kornyshev, Double layer in ionic liquids: overscreening versus
crowding. Phys. Rev. Lett. 106, 046102 (2011)
74. J. Sweeney et al., Control of nanoscale friction on gold in an ionic liquid by a potential-
dependent ionic lubricant layer. Phys. Rev. Lett. 109, 155502 (2012)
75. S. Baldelli, Surface structure at the ionic liquid-electrified metal interface. Acc. Chem. Res.
41, 421–431 (2008)
76. S. Baldelli, Interfacial structure of room-temperature ionic liquids at the solid-liquid interface
as probed by sum frequency generation spectroscopy. J. Phys. Chem. Lett. 4, 244 (2013)
77. S. Watanabe, M. Nakano, K. Miyake, R. Tsuboi, S. Sasaki, Effect of molecular orientation angle
of imidazolium ring on frictional properties of imidazolium-based ionic liquid. Langmuir 30,
8078 (2014)
78. H. Li, R.J. Wood, M.W. Rutland, R. Atkin, An ionic liquid lubricant enables superlubricity to
be ‘switched on’ in situ using an electrical potential. Chem. Commun. 50, 4368 (2014)
79. H. Li, M.W. Rutland, R. Atkin, Ionic liquid lubrication: influence of ion structure, surface
potential and sliding velocity. Phys. Chem. Chem. Phys. 15, 14616 (2013)
80. F. Federici Canova, H. Matsubara, M. Mizukami, K. Kurihara, A.L. Shluger, Shear dynamics
of nanoconfined ionic liquids. Phys. Chem. Chem. Phys. (2014). https://doi.org/10.1039/c4cp
00005f
81. O.Y. Fajardo, F. Bresme, A.A. Kornyshev, M. Urbakh, Electrotunable lubricity with ionic liquid
nanoscale films. Sci. Rep. 5, 7698 (2015)
82. A.M. Smith, M.A. Parkes, S. Perkin, Molecular friction mechanisms across nano fi lms of a
bilayer- forming ionic liquid. J. Phys. Chem. Lett. 5, 4032 (2014)
83. R.M. Espinosa-Marzal, A. Arcifa, A. Rossi, N.D. Spencer, Ionic liquids confined in hydrophilic
nanocontacts: structure and lubricity in the presence of water. J. Phys. Chem. C 118, 6491 (2014)
84. A. Kailer et al., Influence of electric potentials on the tribological behaviour of silicon carbide.
Wear 271, 1922 (2011)
Chapter 10
In Situ Friction Tests in a Transmission
Electron Microscope

Fabrice Dassenoy

Abstract Post mortem characterization techniques are essential for the

understanding of the lubrication mechanisms of complex tribological systems. The
information obtained using these techniques can be used to propose hypothesis of
mechanisms that have then to be definitively validated by probing the interfacial
material in real time during the friction test. To go further in the understanding of
the action modes of some tribological systems, it is important to set up an analytical
methodology using in situ experimental techniques in order to (i) probe directly the
behavior of the interfacial material in the contact zone and (ii) dissociate, for a bet-
ter understanding, the different components of the tribological stress (pressure and
shear). There are several techniques that combine mechanical stress and in situ anal-
ysis (Raman tribology, Infra-Red tribology, EXAFS under pressure …). However,
the most interesting techniques are certainly those that permit a visualization of the
contact area in real time area during the friction test, down to a nanometer scale. The
objective of this chapter is to present through some examples the potential of in situ
friction tests done inside a transmission electron microscope.

10.1 Introduction

The dream of any tribologist is certainly to observe in real time and down to the small-
est detail the behavior of the interfacial material in the tribological contact during
mechanical stress. When the observation takes place at the atomic scale, this dream
becomes a grail. The only technique for this is the Transmission Electron Microscopy
(TEM). Recently, the development of new sample holders for TEM including either
a nanoindenter or an atomic force microscope (AFM) made possible compression
and friction tests in a TEM, thus allowing the simulation of the tribological contact.

F. Dassenoy (B)
Laboratoire de Tribologie et Dynamique des Systèmes, CNRS, UMR 5512,
Ecole Centrale de Lyon, 69134 Ecully, France
e-mail: fabrice.dassenoy@ec-lyon.fr

© Springer Nature Switzerland AG 2018 273

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_10
274 F. Dassenoy

These techniques have a very high potential because they allow combining friction
and in situ observation/analysis, thus permitting to go further in the understanding
of the reaction mechanisms operating in a tribological contact.
In situ TEM techniques were used in many works to determine nanoscale mechan-
ical properties concurrently with real-time microstructure characterization of various
types of samples [1] such as thin film [2–4], carbon nanotube [5], nanowire [6, 7]
and various types of nanoparticles [8–13]. Minor et al. [2, 3] used an in situ TEM
nanoindenter holder to investigate dislocation nucleation and motion in silicon and
aluminium thin films. Deneen et al. [8] used the same device to study the mechanical
behaviour of individual silicon nanospheres. The experiments were carried out on
some 200 nm diameter particles. Both elastic and plastic deformation as well as frac-
ture was observed. In situ TEM Nano indentation experiments were also performed
on polycrystalline CdS hollow spheres ranging in diameter from 200 to 450 nm [7]
and on some clusters of silicon particles around 50 nm in size [10]. In the first work,
it was shown that the particles can achieve both a high compression to failure and
withstand very high shear stresses with respect to their ideal strength, while in the
second one a rotation of the nanoparticles was observed, followed by cracking at
the interface between two nanoparticles. More recently Carlton et al. [11] investi-
gated the deformation of silver nanoparticles under compression using diffraction
contrast and phase contrast TEM. The authors observed the disappearance of dis-
location in single crystal silver nanoparticles after nanoindentation during real time
experimental observations. The use of in situ TEM techniques for solving complex
tribological problems has emerged recently. The objective of this chapter is to illus-
trate through few examples of in situ compression and sliding tests performed on
single nanoparticles the potential of this technique for tribological applications.

10.2 Example 1: In Situ Compression Test on a Single

MoS2 Nanoparticle

In Lahouij et al. [13], performed compression tests of individual IF-MoS2 nanopar-

ticles in a High Resolution Transmission Electron Microscope using an in situ TEM
nanoindenter. Thanks to this technique, in situ observations of the behavior of sin-
gle nanoparticles under pressure were performed for the first time. In addition, the
mechanical response of the particle under compression was followed and force-time
and force-displacement curves were recorded. The compression experiments were
carried out using a Nanofactory Instruments HN200 single tilt Nanoindenter Micro-
Force probing holder (Fig. 10.1) mounted on the specimen holder of a High resolution
JEOL 2010 FEG microscope operating at 200 kV accelerating voltage. The nanoin-
dentation sensor used for the experiments was equipped with a diamond tip. The force
range of the sensor was between 0 and 3 mN and its stiffness constant was 3500 N/m
with a force resolution of 100 nN. Thanks to this system, in situ force measurement
recording was possible. This TEM nanoindenter was used for (i) the characterization
10 In Situ Friction Tests in a Transmission Electron Microscope 275

Signal amplifying
electronics specimen
sensor
positioning

e- beam direction

30 mm

Fig. 10.1 Schematic of the TEM nanoindenter sample holder produced by Nanofactory company
[13]

of the mechanical properties of the fullerenes, (ii) their nanomanipulation at the sub-
nanometer scale, (iii) the observation of their elastic and plastic deformation during
the test. The nanoindenter was also used to carry out sliding experiments by moving
sideways the substrate on which the particles were deposited by nanomanipulation
experiment. Bright Field imaging and the observation of the stress/strain-induced
structural changes in the particle were possible in addition of TEM nanoindentation
experiments. The TEM-Nanoindentor permits to image the nanocompression process
with real time imaging, to record movies, as well as to acquire the force-displacement
and force-time data. For compression and sliding experiments, substrates on which
the nanoparticles have to be previously deposited must be nanomachined to be thin
enough for electron transparency in one dimension, long enough in second dimen-
sion and moderately short in the third one, in order to avoid crashing onto the sensor.
A short wedge sample made in silicon was used. FIB-SEM was used to prepare the
Si-wedge on which the nanoparticles were deposited. Figure 10.2 shows the Si wedge
after machining. The presence of 10 µm long “terraces” used for the nanoparticles
deposition can be observed. A thin gold wire was used to fix the Si wedge to the
positioning part of the TEM sample holder.
Figure 10.3 shows a series of images captured from a video recording during a
nanocompression test carried out on a single IF-MoS2 particle of about 100 nm diam-
eter. The particle does not present any empty core and presents a highly crystalline
and faceted structure. Compression tests were carried out with a truncated diamond
tip of 1 µm width. Nanoscale alignment of the particles and tip along the microscope
electron beam axis was performed by monitoring the relative focus between the tip
and the particles at high TEM magnification. This operation is not straightforward
and can take a long time. Compression was performed by forward actuation of the
support wedge along the z-axis using the fine piezoelectric tube. The exact 3D shape
of the entire particle is difficult to know here because the TEM image is a pro-
276 F. Dassenoy

Fig. 10.2 Si wedge after

FIB machining. The
micromachining enables the
creation of some 10 µm
steps used for the particles
deposition

jection in two-dimension. Each stage consisted of 100 nm displacement controlled

compression of the tip at a loading rate of approximately 13 nm s−1 .
Figure 10.4a, b show the corresponding force/time and displacement/time curves
for this test.
Figure 10.3a shows the nanoparticle on the Si substrate just before compression. It
corresponds to t  0 s on the force/time curve. From this stage, the nanocompression
test starts and the Si substrate moves towards to the tip. At t  2 s, the tip is brought
into real contact with the fullerene surface (Fig. 10.3b). The particle is further loaded
and the evolution of the force can be followed on the force/time curve (Fig. 10.4a).
On Fig. 10.3c, the particle starts to change of shape in the TEM image. This change
is more and more pronounced with the load increase (Fig. 10.3d–g). The facetted
shape of the particle changes progressively to an ellipsoidal shape. To be sure that
the thickness of the particle has increased in the direction of the electron beam to
form the oblate, the intensity of the electron signal in the center of the particle was
measured. The signal is proportional to the grey level because the images have been
recorded practically in the same illumination conditions. The results indicate that
the particle is oblate after compression and accommodates the contact pressure and
progressively rotates and lies parallel to the two contacting surfaces (Fig. 10.3g).
When the maximum force is reached (10 µN at t  7.5 s), then the particle is ejected
from the contact zone and slides outside the compression zone certainly due to low
adhesion to the substrate and the fact that the two surfaces are not strictly parallel
(Fig. 10.3h, i). The sudden sliding of the oblate particle causes an exfoliation of its
outer sheets where the curvature of the lamellar is maximum. These isolated sheets
stick to the diamond tip (Fig. 10.3i—zoomed part). Afterwards, when the particle is
10 In Situ Friction Tests in a Transmission Electron Microscope 277

Diamond
(a) tip (b) (c)

Si substrate

Loading direction
t= 0s t= 2s
t= 3s

(d) (e) (f)

t= 4s t= 6s

20nm
(g) (h) (i)
Exfoliation of
externe sheet

Unloading
t= 7s direction
t=13s

Fig. 10.3 Image captures obtained from a video recorded during a compression experiment carried
out with a single MoS2 particle [13]

unloaded, it remains deformed (Fig. 10.3i). Unfortunately, due to intrinsic motion of

the sample, the resolution in the images does not permit to observe the basal planes
of MoS2 in the structure of the particle during solicitation. The question of either the
particle elastically rotates or plastically deforms cannot be answered properly here,
because of the lack of 3-D imaging in the TEM. However, the observation of the
force/time curve (Fig. 10.4a) does not show any sliding effect, suggesting that the
particle is partially plastically deformed.
It is possible to estimate the contact area between the tip and the sample during
loading by assuming that (i) the contact area of the faceted MoS2 nanoparticle is
a square and (ii) the contact between the tip and the faceted MoS2 nanoparticle is
a flat-on-flat configuration. The contact area between the fullerene and the tip was
measured in the selected video images. Thus, on Fig. 10.3g, the estimated contact
278 F. Dassenoy

Fig. 10.4 Image captures (a)

obtained from a video
recorded during the f
compression experiment
shown in Fig. 10.3 carried
out with a single MoS2
d
particle [13] c
i
b g

(b)

area is ~10,000 nm2 . If we consider that the maximum force applied on the particle
is 10 µN, this gives an estimated average contact pressure of 1 GPa.
In this work, Lahouij et al. [13] investigated in real time the deformation and
degradation behavior of spherical and well crystallized single inorganic fullerene
nanoparticles of MoS2 under compression using a HRTEM equipped with a nanoin-
dentation holder. For the first time, the exfoliation of the outer sheets of a fullerene
nested structure was imaged. However, it was found that well crystallized and round
shape nanoparticles were difficult to exfoliate. Under 1–1.5 GPa uniaxial pressure,
it was found that the shape of single IF-MoS2 nanoparticles was preserved and only
external layers were exfoliated. A uniaxial pressure higher than 1.5 GPa was found to
be necessary to crush the particles. The resistance of the IF-MoS2 to exfoliation was
attributed to their good mechanical properties due to their well crystalized structure
10 In Situ Friction Tests in a Transmission Electron Microscope 279

and the round shape of the particle. However, the behavior of nanoparticles during
mechanical stress strongly depends on the size, the morphology and the crystallinity
of the particle.

10.3 Example 2: In Situ Compression Test on a Single WS2

Nanoparticle

Figure 10.5 shows a compression test carried out on a single WS2 nanoparticle. The
particle present a hollow core compared to the MoS2 nanoparticle studied earlier.
Test was done with a displacement of 45 nm leading to an applied maximum force of
3.5 µN and a maximum contact pressure estimated at 0.45 GPa. The load/time and
displacement/time curves are shown in Fig. 10.6. The behavior under pressure of this
WS2 nanoparticle is completely different to that of the MoS2 nanoparticle. As soon
as the particle is loaded, the sheets are deformed. The particle seems to lengthen (t
 8 s). The central cavity loses volume and is no longer visible (t  11 s). Despite
the loss of image resolution due to vibration, we can see that the particle starts to be
strongly damaged (t  15 s) giving rise to a large number of WS2 sheets (from t 
15 s to t  45 s). At the end of the test, the particle is completely damaged.

10.4 Example 3: In Situ Sliding Test on a Single MoS2

Nanoparticle

In [15], Lahouij et al. gave evidence that the lubricating properties of poorly crys-
tallized MoS2 nanoparticles (with many defects) were better than those of perfectly
crystallized spherical ones. On the basis of the post-mortem analyses of the worn sur-
faces, the authors attributed the good lubricating properties of the poorly crystallized
IF-MoS2 to their ability to exfoliate immediately and to form rapidly a homoge-
neous tribofilm on the surface made of MoS2 layers aligned in the sliding direction.
At the opposite, it was observed that the tribofilm obtained with well-crystallized
IF-MoS2 nanoparticles was heterogeneous and composed of a mixture of MoS2 lay-
ers, iron oxide nanoparticles and intact particles embedded in the tribofilm. It was
also observed that the shape of most of the well crystallized particles was preserved
after friction. The authors ascribed this finding to the perfect crystallinity of the
particles together with the large number of closed layers that confer them with a
higher mechanical resistance. It was also suggested that these perfect particles could
be considered to behave as genuine nano-ball bearings, at least temporarily, until
they gradually deform and start to exfoliate their outer layers. However, only a direct
observation of the sliding contact allows accessing the real behavior of the fullerenes
during a mechanical stress and to answer the question of the influence of the struc-
ture/morphology of the particles on their lubrication mechanisms (exfoliation, sliding
280 F. Dassenoy

Diamond tip

t=0s t=8s t=11s

Si substrate

t=14s t=15s t=18s

t=22s
t=32s t=45s

Fig. 10.5 Image captures obtained from a video recorded during a compression experiment carried
out with a single WS2 particle [14]

and rolling). In [16], the authors focused on the influence of the crystal structure of two
types of IF-MoS2 nanoparticles (namely perfectly crystallized IF-MoS2 and poorly
crystallized IF-MoS2 ) on their behaviors during sliding tests performed inside the
TEM. Figure 10.6 shows a series of image captures obtained from a video recorded
during a sliding experiment carried out with a single perfectly crystallized IF-MoS2
nanoparticle. Figure 10.6a corresponds to the particle compressed between the tip
and the Si substrate before starting the sliding test at t  t 0 . Figure 10.6h shows the
particle at the end of the sliding test at t  t final . A white point was arbitrary placed
on the particle in order to easily follow its movement during the sliding test. The
white arrow marks the starting point on the diamond tip. When the sliding test starts
(by moving the Si substrate parallel to the tip—the white arrow on the Si substrate
indicates the direction of the movement of the Si substrate), it can be observed from
the image captures that the IF-MoS2 particle rolls in the contact. The distance trav-
10 In Situ Friction Tests in a Transmission Electron Microscope 281

Fig. 10.6 Image captures obtained from a video recorded during a sliding experiment carried out
with a single perfectly crystallized IF-MoS2 nanoparticle. a–h shows the rolling of the particles
and g–h shows the exfoliation of an outer layer. A white point was arbitrary placed on the particle
in order to easily follow the movement of the particle during the test. The white arrow marks the
starting point on the diamond tip. The distance travelled by the particle from t  t 0 to t  t final is
estimated to be 160 nm [16]

elled by the particle from Fig. 10.6a–h was estimated to be of ~160 nm. This means
that the particle has travelled a little more than half of its circumference calculated
to be of ~250 nm. It can be also observed from these images that the structure and
the shape of the particle are preserved during sliding and that only an outer layer is
delaminated as shown on the zoomed parts of Fig. 10.6g, h).
In the same way, the behavior of a single poorly crystallized IF-MoS2 particle
was observed during a sliding test. Figures 10.7 and 10.8 show respectively two
sequences recorded during a sliding test performed from a single isolated IF-MoS2
of about 30 nm of diameter. The two sequences of the sliding test were performed
with an average normal force of, respectively, 1 and 4 µN and a constant velocity of
1 nm/s. Figure 10.7a shows the single particle on the Si substrate just before starting
the sliding test. The surface of the diamond tip is covered (contaminated) by some
MoS2 layers deposited during previous tests. Figure 10.7b–f show the behavior of
the particle during the sliding test. Figure 10.7b shows the particle when the normal
force of 1 µN is applied to the particle (without shear stress) corresponding to an
estimated contact pressure of 110 MPa. No important structural change of the particle
is observed, except for a small compression. When the shear stress is applied, a huge
compression with ensuing exfoliation and material transfer is observed. This leads
to a delivery of exfoliated MoS2 layers in the contact as shown in Fig. 10.7c–e. At
the end of this first sequence, the normal force is unloaded, and it can be seen in
Fig. 10.7f that both the surface of the wedge and the tip are now covered by some
MoS2 layers aligned along the sliding direction. The second sequence consists in
the observation of the deformation of the exfoliated MoS2 layers during the sliding
282 F. Dassenoy

Fig. 10.7 Image captures obtained from a video recorded during a sliding experiment carried out
with a single poorly crystallized IF-MoS2 nanoparticle demonstrating exfoliation. The black arrow
on the Si substrate indicates the direction of the movement of the Si substrate [16]

process (Fig. 10.8). Figure 10.8a corresponds to the final state of the first sequence
(Fig. 10.7f) and is also the initial state of the second sliding sequence. The substrate
covered by exfoliated MoS2 layers was placed in contact with the MoS2 layers
deposited on the diamond tip. A normal load of 4 µN was applied. Figure 10.8b
shows the MoS2 layers compressed between the wedge and the tip. When the wedge
starts to move sideways, a transfer of material from the wedge to the tip can be
observed (Fig. 10.8c). The displacement of the wedge induces the shearing of MoS2
layers (Fig. 10.8d, e). At the end of the test, we can see that a few sheared MoS2
layers were detached from the tip and adhere now on the wedge (Fig. 10.8f). This
sequence shows the easy shearing properties of the MoS2 film under the combined
effect of pressure and shear stress.

10.5 Example 4: Friction Coefficient Measured on a Single

WS2 Nanoparticle

Through an in situ series of experiments using a special TEM-AFM holder that per-
mitted the simulation of a tribological contact, Lahouij et al. in [13, 16] described
the lubrication mechanism of Inorganic Fullerene Like nanoparticles made of metal
disulfide. The experiment provided valuable qualitative results. Rolling of MoS2
nanoparticles was observed at estimated contact pressure up to 100 MPa, while exfoli-
ation at approximately 1 GPa. It was observed that the nanoparticles can behave differ-
ently under seemingly similar experimental conditions: rolling or sliding/exfoliating.
In their experiments, the sample holder used had increased sensibility for measuring
10 In Situ Friction Tests in a Transmission Electron Microscope 283

Fig. 10.8 Image captures obtained from a video recorded during a sliding experiment carried out
with a single poorly crystallized IF-MoS2 nanoparticle demonstrating the easy shearing between the
exfoliated MoS2 layers. The white arrow on the Si substrate indicates the direction of the movement
of the Si substrate [16]

normal forces, in the direction of the main axis of the device. It was able to move
in normal and lateral direction, but it was not able to measure lateral force, hence
the friction coefficient. Recently, Jenei et al. [17] went further in the investigation of
the lubrication mechanisms of metal disulfide nanoparticle. An in situ friction test
was carried out on a single WS2 nanoparticle in a transmission electron microscope
(TEM) using a nano indentation device able to measure normal and lateral forces
independently. For the first time, quantitative results were reported: friction forces
were recorded during the test, and thus friction coefficient values were calculated.
The exfoliation of the particle was observed during the experiment, and was linked
to the friction coefficient modification. The in situ tests were conducted in an FEI
Titan environmental transmission electron microscope (ETEM) operated on 300 kV
accelerating voltage, equipped with a Cs corrector, with the help a Pi-95 Picoindenter
from Hysitron, Inc. The sensor (transducer) of the picoindenter is a MEMS based
device, which has a diamond tip attached to it. The transducer is capable of control-
ling the tip with a 0.02 nm displacement resolution and it is capable of measuring
forces with a 3 nN resolution. The use of a 2D transducer makes it possible to mea-
sure normal and lateral forces at the same time, independently from each other. The
tip of the transducer used in this experiment is a so called “at punch”, it has a flat
surface section with a length of approximately 160 nm (see Fig. 10.9).
A tribological contact was mimicked with a single WS2 nanoparticle being held
between the silicon wedge and the diamond tip of the transducer. A “nanofriction
test” was designed in the following way: the diamond tip was aligned to the selected
WS2 nanoparticle, a constant load was applied on the nanoparticle after which the tip
284 F. Dassenoy

Fig. 10.9 a Schematic representation of the silicon wedge, the plateau is depicted in light blue. b
Schematic representation of a nanoparticle deposited on the plateau, the diamond tip slides along
the length of the wedge. c A TEM image of the WS2 nanoparticle on the silicon wedge before the
experiment [17]

Fig. 10.10 a The control function for the in situ friction test and b the measured normal force and
lateral displacement during the experiment [17]

was moved laterally to the right and then to the left (right and left strokes). This was
considered as a cycle or a segment. At the end of the segment the load was increased.
Between each stroke there was a short pause. In total 4 segments were recorded. The
test was performed in the so called “load control” mode which means that during
the test the instrument made sure that the predefined load was being applied to the
nanoparticle. During the test the normal and lateral forces were recorded, as well as
the normal and lateral displacement of the tip. A TEM image in Fig. 10.9c shows the
WS2 nanoparticle on the silicon wedge with the aligned diamond tip. The normal
load and lateral displacement during such a test was predefined. Figure 10.10a shows
the control function for the experiment. The length of one stroke is 50 nm, exceptions
are the first and last strokes where the displacements are only 25 nm. The normal
load in the first segment was 4.5 µN, then it is set to increase with a 0.5 µN step
in each segment, reaching a maximum of 6.5 µN at the end of the experiment. The
length of the friction test was set to be a little longer than three minutes. The actual
normal force and lateral displacement values measured during the experiment can
be seen in Fig. 10.10b.
10 In Situ Friction Tests in a Transmission Electron Microscope 285

Fig. 10.11 The normal and lateral forces and the friction coefficient values recorded during the
experiment [17]

Figure 10.11 shows the normal and lateral force and the calculated friction coef-
ficient (COF) during the whole experiment. The COF is calculated as the ratio of the
lateral and normal force. During the first couple of strokes, the friction force is rela-
tively high (compared to the rest of the experiment), after which there is a significant
decrease of the COF. The friction coefficient reduces from a 0.55 level to 0.15. At
the beginning of the last stroke, where the normal load is increased to 6.5 µN, the
nanoparticle slips out from between the diamond tip and the silicon wedge. The COF
increases instantly. It is interesting to see that as the stroke progresses, the friction
coefficient follows a decreasing trend, although here the diamond tip is sliding on
the silicon wedge. Figure 10.12 shows some image captures from the video of the
“nano” friction test. In Fig. 10.12a the exfoliation during the right stroke in segment
1 can be observed. In Fig. 10.12b the adhesion of the previously exfoliated layer is
visible; in (c): exfoliation during the right stroke in segment 1; in (d) exfoliation at
the left stroke in segment 1; in (e) exfoliation at the end of the left stroke in segment
1, this layer is not completely detached from the nanoparticle; in (f) exfoliation in
the left stroke of segment 2; in (g) an exfoliation can be observed at the right cap of
the particle just before is slips out from the contact; in (h) several exfoliated layers
can be seen between the tip and the silicon wedge after the particle got out from the
contact.
This experiment permitted to follow in real time the evolution of the friction
coefficient during the tribotest done in situ inside a TEM and to establish a correlation
with the results obtained at the micro scale.
286 F. Dassenoy

Fig. 10.12 Frames captured from the video of the “nano” friction test performed on a WS2 nanopar-
ticle [17]

10.6 Conclusion

Friction experiments carried out in a Transmission Electron Microscope (TEM) con-

stitute a way to go further in the understanding of the reaction mechanisms operating
in a tribological contact. They enable a direct visualization of the behavior of the
interfacial material during the tribological test, thus making possible the follow in
real time of any structural and morphological change in the stressed material. With
the emergence of new microscopes that allow carrying out experiments in a con-
trolled environment, it will be soon possible to perform similar tests in presence of
gas, thus making possible the in situ monitoring of chemical reactions between the
stressed materials and this gas. Complementary techniques such as Electron Energy
Loss Spectroscopy (EELS) will be in this respect very useful.
10 In Situ Friction Tests in a Transmission Electron Microscope 287

References

1. J.M. Howe, H. Mori, Z.L. Wang, In situ high-resolution transmission electron microscopy in
the study of nanomaterials and properties. MRS Bull. 33(2), 115–121 (2008)
2. A.M. Minor, J.W. Morris Jr., E.A. Stach, Quantitative in situ nanoindentation in an electron
microscope. App. Phys. Lett. 79, 1625–1627 (2001)
3. E.A. Stach, T. Freeman, A.M. Minor, D.K. Owen, J. Cumings, M.A. Wall, T. Chraska, R. Hull,
J.W. Morris Jr., A. Zettl, U. Dahmen, Development of a nanoindenter for in situ transmission
electron microscopy. Microsc. Microanal. 7, 507–517 (2001)
4. N. Li, J. Wang, J.Y. Huang, A. Misra, X. Zhang, In situ TEM observations of room temperature
dislocation climb at interfaces in nanolayered Al/Nb composites. Scr. Mater. 63, 363–366
(2010)
5. Z.L. Wang, P. Poncharal, W.A. de Heer, Measuring physical and mechanical properties of
individual carbon nanotubes by in situ TEM. J. Phys. Chem. 61, 1025–1030 (2000)
6. X. Han, K. Zheng, Y.F. Zhang, X. Zhang, Z. Zhang, Z.L. Wang, Low-temperature in situ
large-strain plasticity of silicon nanowires. Adv. Mater. 19, 2112–2118 (2007)
7. A. Asthana, K. Momeni, A. Prasad, Y.K. Yap, R.S. Yassar, In situ observation of size scale
effects on the mechanical properties of ZnO nanowires. Nanotechnology 22, 265712 (2011)
8. J. Deneen, W.M. Mook, A.M. Minor, W.W. Gerberich, C.B. Carter, In situ deformation of
silicon nanospheres. J. Mater. Sci. 41, 4477–4483 (2006)
9. Z.W. Shan, G. Adesso, A. Cabot, M.P. Sherburne, S.A. Syed Asif, O.L. Warren, D.C. Chrzan,
A.M. Minor, A.P. Alivisatos, Ultrahigh stress and strain in hierarchically structured hollow
nanoparticles. Nat. Mat. 7, 947–952 (2008)
10. A.J. Lockwood, B.J. Inkson, In situ TEM nanoindentation and deformation of Si nanoparticle
clusters. J. Phys. D Appl. Phys. 42, 035410 (2009)
11. C.E. Carlton, P.J. Ferreira, In situ TEM nanoindentation of nanoparticles. Micron 43,
1134–1139 (2012)
12. I. Lahouij, F. Dassenoy, L. De Knoop, J.M. Martin, B. Vacher, In situ TEM observation of the
behavior of an individual fullerene-like MoS2 nanoparticle in a dynamic contact. Tribol. Lett.
42, 133–140 (2011)
13. I. Lahouij, F. Dassenoy, B. Vacher, J.M. Martin, Real time imaging of compression and shear
of single fullerene-like MoS2 nanoparticle. Tribol. Lett. 45, 131–141 (2012)
14. I. Lahouij, Ph.D. thesis, Ecole Centrale de Lyon, France (2013)
15. I. Lahouij, B. Vacher, J.M. Martin, F. Dassenoy, IF-MoS2 based lubricants: influence of size,
shape and crystal structure. Wear 296, 558–567 (2012)
16. I. Lahouij, B. Vacher, F. Dassenoy, Direct observation by in situ TEM of the behavior of
IF-MoS2 nanoparticles during sliding tests. Lubr. Sci. 36(3), 163–173 (2014)
17. I. Jenei, F. Dassenoy, Friction coefficient measured on a single WS2 nanoparticle: an in situ
transmission electron microscope experiment. Tribol. Lett. 65, 86 (2017)
Chapter 11
In Situ Digital Holography for 3D
Topography Analysis of Tribological
Experiments

Martin Dienwiebel and Pantcho Stoyanov

Abstract This chapter reviews topography measurement techniques that allow

performing the in situ quantification of roughness and the analysis of the surface mor-
phology during sliding experiments. The method of digital holographic microscopy
is then introduced in detail and examples of topography evolution are presented.

11.1 Introduction

In situ tribometry has significantly advanced the field of Tribology over the last 2
or 3 decades because temporal information of the state of the sliding contact is very
beneficial in obtaining in-depth understanding of the processes that lead to differences
in the friction and wear performance of a tribosystem. However the term is widely
used to describe very different methods and experimental situations. It is mostly used
when it is possible to look into the tribological by using a transparent counter surface.
This setup allows to measure the real area of contact between surfaces (see e.g. chap.
9 by Dassenoy), changes in chemistry (see chaps. 4, 6 and 7) or the thickness of a
lubricant layer [1]. When it is not possible or desirable to use a transparent sliding
surface it is possible to monitor changes in the wear track close to the tribological
contact as described in a recent review by Wahl and Sawyer [2].
The term in situ is also used when the tribometer is located inside another instru-
ment such as a TEM or when it is operating in a special environment such as an ultra-
high vacuum chamber, as beautifully described in chap. 5 by Tysoe. This chapter will

M. Dienwiebel (B) · P. Stoyanov

Institute for Applied Materials, Microtribology Center μTC, Karlsruhe Institute of Technology
KIT, Strasse am Forum 7, 76131 Karlsruhe, Germany
e-mail: martin.dienwiebel@kit.edu; martin.dienwiebel@iwm.fraunhofer.de
P. Stoyanov
e-mail: pantcho.stoyanov@gmail.com
M. Dienwiebel · P. Stoyanov
Fraunhofer-Institute for Mechanics of Materials, Wöhlerstrasse 11,
79108 Freiburg, Germany

© Springer Nature Switzerland AG 2018 289

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_11
290 M. Dienwiebel and P. Stoyanov

give an overview of in situ topography analysis during a tribological experiment and

will describe the method of digital holographic microscopy (DHM) with a focus on
tribology.

11.1.1 Optical Roughness Measurements

Optical roughness measurement techniques or optical profilers are well established

and are being used in tribology research as well as in industrial quality control. Optical
profilers can be divided into instruments that are interferometry based and confocal
instruments. All optical techniques have in common that the lateral resolution is
typically an order of magnitude lower that the resolution in the normal direction
and that it depends on the used objective lens of the microscope. The optics also
leads to differences in the maximum slope that can be observed with either confocal
microscopy, phase shift or white light interferometry (Fig. 11.1).

Fig. 11.1 Top: Schematic of a combined White Light Interferometer and Tribometer. Bottom:
Friction as function of sliding cycles and snapshots of the wear track topography. Adapted from
[3] with permission from W . Gregory Sawyer and Kathryn J. Wahl
11 In Situ Digital Holography for 3D Topography … 291

With the introduction of the new standard EN ISO 25178 the measurement of 3D
surface parameters using different instruments and their calibration are nowadays
available. In this standard, parameters such as Ra , Rz and Rt are substituted by their
corresponding 3D parameters S a , S z and S t . As will be shown in this chapter the
analysis of sliding surfaces should not be limited to the measurement of roughness
parameters. Especially the observation of changes in time at the same location of the
sliding track can be very valuable.
The implementation of an optical profiler into a tribological experiment is possible
when the wear track can be accessed optically, i.e. when the length of the wear track
is larger than the diameter of the counter sample and the objective lens. To overcome
ambiguities in height due to phase jumps either the objective lens or the surface has
to be scanned vertically and later the topography data is constructed from a large
number of single images. The acquisition time thus depends on the total roughness
of the surface and is typically in the order of several seconds.

11.1.2 Digital Holographic Microscopy (DHM)

A digital holographic microscope uses a coherent object wave that is brought into
superposition with a reference wave. Figure 11.2 shows an optical configuration and
is described in detail elsewhere [4, 5]. Instead of a photographic film or plate that is
used to capture a classic hologram in digital holography the hologram is recorded with
a CCD camera [6, 7] and therefore the acquisition is not limited by the exposure of the
film and thus depends on illumination and frame rate of the CCD. The reconstruction
of the object wave is obtained in DH by numerical computation with a computer using
diffraction theory. Recent progress in computational speed and in the development
of CCD cameras has made DHM a quasi “real-time” method. However, similar to
the case of White Light Interferometry the reconstructed topography will suffer from
phase jumps when the roughness of the sample is larger than the wavelength λ of the
coherent light source. In such a case numerical phase unwrapping algorithms can be
used. However, these algorithms can fail under certain circumstances such as steep
wear tracks that extent over the whole field of view. Another way to circumnavigate
phase ambiguities is to use two or more lasers [8]. Also Kühn et al. show that by using
a dual-wavelength approach the a new synthetic wavelength  can be expressed as
[9]:
λ1 λ 2
Λ (11.1)
λ2 − λ1

For two laser diodes with λ1  680 nm and λ2  760 nm the resulting syn-
thetic wavelength, where phase jumps will occur is   6.46 μm. Therefore dual-
wavelength DHM allows measuring surfaces with larger roughness without the
numerical phase unwrapping.
292 M. Dienwiebel and P. Stoyanov

Fig. 11.2 Digital

holographic microscope for
reflection imaging. Reprinted
with permission from [4]
OSA

11.2 Integration of DHM into a Tribometer

In order to create a sequence of topographic images in a sliding experiment the

topography acquisition has to be performed at the same position after each pass of
the counter surface (Fig. 11.3). The precision of the positioning thus has to be better
than the lateral resolution of the microscope. In the homebuilt tribometer that was
constructed to integrate a DHM, this is accomplished by using a high precision x-y
table that is steered by electromagnetic fields and that is used to move the lower sliding
surface [10]. The position is monitored and adjusted optically with a resolution of
20 nm. The table can achieve a speed of 0.5 m/s and has a total travel of 200 mm
in each direction. The position repeatability is better than 1 μm. The x-y table is
mounted on a granite block and bridge in order to reduce mechanical vibrations that
also deteriorate the DHM image acquisition. In order to further reduce vibrations the
granite block is mounted on air dampers.
In a different example, a DHM was used to provide a better understanding of
the cavitation phenomenon in tribological contacts [F]. In particular, the authors
used DHM to measure cavitation bubble formation and thickness. The schematic
representation of this measurement technique is shown in Fig. 11.4.
In the following sections of this chapter, several examples are given in order to
demonstrate where DHM can help to obtain a better understanding of tribological
phenomena by investigating the topography evolution of sliding surfaces.
11 In Situ Digital Holography for 3D Topography … 293

Fig. 11.3 The concept of integrating DHM, pin (square), and an atomic force microscopy (AFM)
(circle). On the right, the travel path for the three components is indicated as AFM (dashed line),
DHM (dotted line) and pin (solid line). Reprinted with permission from [11]

Fig. 11.4 Schematic representation of using DHM to improves the understanding of the cavitation
phenomenon in tribological contacts. Reprinted under creative commons licence from [12]

11.2.1 Analysis of Plowing Friction

Plowing generally describes a situation where a sharp asperity indents into and plas-
tically deforms a softer surface. The friction that is generated in the case of plowing
is described as the sum of a shear contribution F S and a plowing contribution F P .
According to Bowden Moore and Tabor the two terms can be written as [13, 14]:

FF  FS + FP  τ A R + p  A (11.2)

Here τ is the shear strength of the softer material, AR the real area of contact, p
is the flow pressure and A the projected area of the asperity in direction of sliding.
Thus the plowing friction depends mainly on the shape of the asperity and the normal
294 M. Dienwiebel and P. Stoyanov

Fig. 11.5 Extracted profile lines from individual DHM frames. The profile lines are stitched
together to compensate for the repositioning of the DHM microscope. The black solid line indicates
the edge of the sliding track. Reprinted with permission from [15]

force. It is assumed in (11.2) that the asperity creates a single scratch and indeed
plowing often dominates the initial phase of the running-in phase, whereas in a later
stage the running-in of metals is governed by the formation of a third body. In the
case that the plowing track subsequently widens and deepens, the model has to be
modified. Therefore, in order to model friction in the case of multiple reciprocating
sliding, the sliding track created by a spherical indenter was monitored as function
of sliding cycles and correlated to the change in friction. The plowing experiments
were conducted on a polished high purity copper sample. As indenter a ruby sphere
with a diameter of 1 mm was chosen, mainly because of the high precision of the
shape and the low roughness of the sphere. The surface was lubricated using base
oil (PAO 8) in order to transport wear particles out of the sliding track and to prevent
oxidation. The experiments were performed at room temperature and a humidity of
approx 45% RH.
During the experiment the sliding track widened very rapidly and thus the DHM
had to be repositioned perpendicular to the sliding track in order to keep one side
of the track in the field-of-view of the microscope. At each sliding cycle one profile
line across the sliding track was evaluated. All profile lines were added to a profile
versus sliding cycle map. To correct for the repositioning all profile lines had to be
stitched together after the experiment (Fig. 11.5).
11 In Situ Digital Holography for 3D Topography … 295

Fig. 11.6 Friction coefficient μ (solid line) and widening rate (dotted line) for 8 N (a) and 4 N (b)
normal load and for 20 mm/s (a) and 15 mm/s (b) speed. Reprinted from [15]

From the position of the track edge the radius r of indentation can be extracted
assuming that the widening of the track is symmetric. According to (11.2) the plowing
contribution to friction should be proportional to r. However, in our experiments
we find a better correlation between the friction coefficient and the widening rate
per sliding cycle ṙ . The reason for this apparent discrepancy is the fact that (11.2)
describes the situation of a single scratch or pass whereas the experiments here
are performed by sliding multiple times in the same track. In this case a plowing
contribution to the friction force is only present if the sphere sinks in between two
subsequent cycles, since the effective cross section that contributes to the friction is
A  K i − K i−1 , and the plowing cross section
⎡  ⎤
 2
1⎣ 2 R R ⎦
r cos−1
i i
Ki  − Ri rs2 − (11.3)
2 s rs 2

Here r s is the radius of the sphere and Ri is half of the width of the plowing track
[16]. If the increase of Ri with every cycle is small, then it can be shown that A scales
with the widening rate. Since the track widens very fast during the first cycles the
above correlation can only be seen in the case of a ruby sphere sliding against copper
from cycles 75–100 onwards depending on the loading conditions (Fig. 11.6).

11.2.1.1 The Influence of Running-in on the Topography of Brass

In a very similar fashion as presented in the first example, different running-in exper-
iments were conducted on CuZn5 brass plates. The plates were annealed at a tem-
perature of 650 °C for 45 min. The average grain size after the heat treatment was
30 ± 3 μm. Counter surfaces were produced from100Cr6 spheres with a diameter of
3 mm which were first hand-grinded and then hand-polished on one side in order to
obtain flat-on-flat geometry. The tribocontact was lubricated by spraying poly-alpha
296 M. Dienwiebel and P. Stoyanov

Fig. 11.7 Average friction coefficient μ as a function of speed and nominal contact pressure.
(Reprinted with permission from [17], Elsevier)

olefin (PAO-8 with a viscosity of 45.8 mm2 /s at 40 °C). The oil temperature through-
out the sliding experiments was kept at 35 °C. For the experiments we use a linear
reciprocating path with a length of 120 mm. After each cycle the topography was
recorded at the same location by DHM. In order to influence the running-in behavior
the speed was varied from 10 to 20 mm/s and the nominal contact pressure from 1
to 4 MPa respectively.
Figure 11.7 shows the average friction coefficient recorded over approximately
5000 sliding cycles. Only few experiments located in a narrow range from 2.2 to
2.9 MPa (dashed lines) led to a significant friction reduction. The experiments that
were conducted in this pressure range show a markedly decrease in friction (from
0.1 to below 0.05) with considerably smaller scatter of the friction force from one
to another cycle. Figure 11.8 shows the evolution of the COF and for comparison
the average roughness as a function of sliding cycles for one experiment that was
conducted in the low-friction corridor. The roughness was computed from DHM
topography maps. In this case the roughness follows roughly the trend in the friction.
Another example of an experiment that is in terms of normal pressure located outside
the low-friction corridor is presented in Fig. 11.9. Here the friction force and also the
roughness show strong fluctuations although and possess a similar behavior although
there is no strict correlation between the two signals.
11 In Situ Digital Holography for 3D Topography … 297

Fig. 11.8 Coefficient of friction (solid line) and average roughness S a (dotted line) for a constant
sliding speed of 20 mm/s and a nominal contact pressure of 2.7 MPa. The experiment was performed
under lubrication using a synthetic PAO-8 base oil. (Adapted from [17])

Fig. 11.9 Coefficient of friction (solid line) and average roughness S a (dotted line) for a constant
sliding speed of 20 mm/s and a nominal contact pressure of 3.6 MPa. The insets show topography
maps of the wear track. The experiment was performed under lubrication using a synthetic PAO-8
base oil. (Adapted from [17])
298 M. Dienwiebel and P. Stoyanov

Fig. 11.10 An approach of studying dynamic interfacial processes leading to the variations in
the friction and wear. This concept involves linking on line experiments to atomistic simulations
with realistic bond order potentials. The structural/chemical changes observed in the simulations
are compared to ex situ analysis using transmission electron microscopy, X-ray photoelectron
spectroscopy, Auger Electron Spectroscopy, and Raman spectroscopy (Adapted from [21])

11.3 Topography and Wear Evolution in Metal/Ceramic

Tribosystems

11.3.1 Tungsten Sliding Against Tungsten Carbide

In one study, we used the DHM in order to evaluate the roughness in dry sliding of a
metal versus a ceramic in order to capture the dynamic interfacial processes leading
to the variations in the friction and wear. The overall approach is shown in Fig. 11.10,
linking in situ DHM characterization to atomistic simulations and ex situ analysis
using transmission electron microscopy, X-ray photoelectron spectroscopy, Auger
Electron Spectroscopy, and Raman spectroscopy. For these experiments tungsten and
tungsten carbide were chosen in order to perform molecular dynamics simulations
with realistic potentials. The simulations are presented elsewhere [18–20]. In the
present example a 20× objective lens is used that is not operating in immersion,
therefore allowing a larger field of view and measurement of the complete wear
track.
Upon initial sliding the friction is approximately 0.1 and subsequently levels off
to 0.6 where it remains nearly constant, as shown in Fig. 11.11a. Similarly, the
scatter in the coefficient of friction increases after 40 cycles. The average roughness,
11 In Situ Digital Holography for 3D Topography … 299

measured by DHM, followed the trend of the friction coefficient closely. A similar
behavior is observed from the roughness values parallel and perpendicular to the
sliding direction. However, there are evident differences between the two directions,
as shown in Fig. 11.11b; while the roughness in the perpendicular direction increases
instantaneously (i.e. within 10 cycles) to nearly 80 nm, the roughness in the parallel
direction took more than four times as long to reach the same value. This can be
explained by the formation of grooves and scratches along the sliding direction and
has been previously referred to as an adaptation of the surface topographies between
the sliding interfaces [22]. These processes typically consist of plowing or cutting
events caused by the asperities of the harder material onto the softer one (see [14]).
The wear depth values, obtained using the DHM data, are used to calculate
the wear rate (i.e. depth/sliding distance) and analysed in terms of cycle number,
Fig. 11.11c. The wear depth is calculated only up to the 80th cycle due to the width
of the wear track being larger than the image size for the remaining cycles. The
wear rate of the last cycle is also included in the analysis; however it is calculated
using the depth obtained from the confocal microscope at the end of the test. The
values measured by DHM and ex situ suggest that the wear rate remains constant
after the first 50 sliding cycles. Interestingly, while the coefficient of friction and the
average roughness are increasing, the wear rate is decreasing in the same phase of
the experiment.
The in situ digital holography approach revealed a different behavior in lubricated
ceramic/metal tribocouples (i.e. WC/W). The DHM images and corresponding fric-
tion coefficients are shown in Fig. 11.12. As expected, the friction is approximately
0.1 throughout steady state. Similarly to the unlubricated WC/W sliding contacts, the
DHM images upon initial sliding show grooves parallel to the sliding direction indi-
cating adaptation of the two surface topographies, as shown with the DHM images
in Fig. 11.12. However, unlike the lubricated sliding case, the grooves here remained
visible for the remainder of the test and the roughness values parallel to the sliding
was low throughout the duration of the test. This correlated well with the low and
steady friction values in the lubricated conditions.
An added benefit of integrating DHM into a tribometer is the observations of
individual particles from consecutive images within the lubricant. An example of the
consecutive images within the wear track is shown in Fig. 11.13. The observations
clearly indicate that the particles are mobile and float within the hexadecane for
the WC/W tribocouple with hexadecane. More details on this analysis can be found
elsewhere [23].

11.3.2 Tungsten Sliding Against Diamond-like Carbon

In situ digital holography was also used more recently on diamond-like carbon coat-
ings in order to provide a better understanding of the interfacial phenomena in lubri-
cated and unlubricated sliding conditions. Figure 11.14 shows the average friction
coefficient as a function of the cycle number for the two conditions. Similarly to the
300 M. Dienwiebel and P. Stoyanov
 11 In Situ Digital Holography for 3D Topography … 301

Fig. 11.11 a Evolution of friction coefficients and average roughness. b Roughness values obtained
using a thin rectangle parallel to the sliding direction and a thin rectangle perpendicular to the sliding
direction (right panel). The roughness values obtained from the two directions versus the cycles
are shown in left panel. c Wear rate for dry sliding W against WC. Roughness and wear rate are
obtained from in situ holographic microscopy. Due to the large wear track created with the WC
tip, wear rates are only possible to be obtained up to the 80th cycle. The last data point of the wear
rate is obtained using ex situ confocal microscopy. It represent the average wear over ~520 cycles
(Reprinted with permission from [23])

Fig. 11.12 Top: Evolution of the coefficient of friction as function of the sliding speed of a tungsten
carbide sphere sliding against tungsten lubricated by PAO for sliding speeds ranging from 0.2 to 20
mm/s. Bottom: Topography of the wear track at different stages of the experiment (Reprinted from
[23])

lubricated tungsten carbide/tungsten tribocouple. the friction coefficient of a-C:H

sliding against tungsten with hexadecane is approximately 0.1 and remains nearly
constant throughout the test (Fig. 11.14a). As a comparison, the dry sliding of a-C:H
against tungsten resulted in slightly higher steady state friction values (Fig. 11.14b)
however increased to above 0.4 between 2000 and 3000 cycles (Fig. 11.14c). The
DHM images of the different conditions are shown within the figures and the observa-
tions followed the friction behavior closely; the lubricated test showed smooth wear
tracks with lower roughness parallel to the sliding direction, while the dry sliding
resulted in higher roughness parallel to the sliding. The behavior is also observed with
the holographic images within the figures, where the lubricated conditions showed
smoother wear tracks compared to the dry sliding case. Consistently with the increase
in friction between 2000 and 3000 cycles for the unlubricated condition, the holo-
graphic images showed an increase in surface roughness and increased presence of
302 M. Dienwiebel and P. Stoyanov

Fig. 11.13 DHM images of two consecutive cycles (i.e. cycles 90 and 91) for a sliding velocity of
5 mm/s. Observations of individual particles indicate that the particles are mobile and float within
the hexadecane (Reprinted with permission from [23])

debris particles (Fig. 11.14c) different experiments. The black line represents the
average value obtained from two experiments, and the blue lines show the mini-
mum and maximum values at each cycle. (b) Coefficient of friction vs. cycle as well
as holographic images of the wear track under dry sliding conditions for the first
1000 cycles and (c) long experiments (i.e. between 1000 and 4000 cycles). Exper-
iments with higher cycle numbers show that the coefficient of friction increases up
to 0.5 between 2500 and 3000 cycles, after which it fluctuates between 0.2 and 0.5.
From [19].

11.3.3 Aluminum Sliding Against Sapphire

In a more recent study, in situ DHM was combined with optical in situ observation
of the contact in order to study the interfacial processes in lubricated metallic (i.e.,
aluminum based) sliding conditions, as shown in Fig. 11.15 (see [24]). The DHM
analysis in this study showed that the roughness evolution followed the coefficient of
friction trend closely, with initially low values followed by higher roughness during
steady state. Consistently, the optical in situ observation of the contact revealed that
the transfer film behavior correlated well with the roughness of the worn surfaces
and the subsurface microstructure of the worn surfaces [24].
11 In Situ Digital Holography for 3D Topography … 303

Fig. 11.14 a Friction coefficient as a function of the number of sliding cycles of a a-C:H coated
sphere sliding against a tungsten plate lubricated by hexadecane at room temperature. The friction
values are averages of two consecutive experiments. The insets show snapshots of the topography
recorded by DHM. b Friction coefficient vs. cycles for an unlubricated a-C:H sphere against tungsten
with 1000 cycles (b) and 4000 cycles (c). Instabilities in friction appear after approximately 3000
cycles (Reprinted with permission from [19])

Fig. 11.15 Schematic representation of combining in situ and on-line methods for studying the
interfacial processes in sliding couples. (Reprinted with permission from [24])
304 M. Dienwiebel and P. Stoyanov

11.4 Conclusion

This chapter has addressed and outlined fundamental aspects of topography mea-
surement techniques that allow performing in situ quantification of roughness, wear
and surface morphology during sliding experiments. The main focus was on holo-
graphic microscopy integrated in custom-build tribometers. The use of DHM within
tribometers broadens our understanding of the sliding processes and mechanisms
that govern the friction and wear behavior. This approach was demonstrated with
different material couples including metallic contacts in the lubricated and unlubri-
cated conditions. In addition, there exists several opportunities of combining DHM
with other in situ techniques (e.g. optical observation of the contact) in order to
study the interfacial processes in lubricated metallic. As a future outlook, combining
in situ techniques with ex situ transmission electron microscopy or scanning elec-
tron microscopy will provide a better understanding on the interfacial phenomena
and velocity accommodation modes of tribosystems.

References

1. R. Gohar, A. Cameron, Nature 200, 458 (1963)

2. W.G. Sawyer, K.J. Wahl, MRS Bull. 33, 1145 (2008)
3. K.J. Wahl, W.G. Sawyer, MRS Bull. 33, 1159 (2008)
4. E. Cuche, P. Marquet, C. Depeursinge, Appl. Opt. 38, 6994 (1999)
5. P. de Groot, Adv. Opt. Photonics 7, 1 (2015)
6. M.K. Kim, SPIE Rev. 1, 18005 (2010)
7. M. Matrecano, M. Paturzo, P. Ferraro, Opt. Eng. 53, 112317 (2014)
8. J. Gass, A. Dakoff, M.K. Kim, Opt. Lett. 28, 1141 (2003)
9. J. Kühn, T. Colomb, F. Montfort, F. Charrière, Y. Emery, E. Cuche, P. Marquet, C. Depeursinge,
Opt. Express 15, 7231 (2007)
10. S. Korres, M. Dienwiebel, (n.d.)
11. S. Korres, M. Dienwiebel, Rev. Sci. Instrum. 81, 63904 (2010)
12. T. Tang, N. Morris, J. Coupland, L. Arevalo, Tribol. Lett. 58, 5 (2015)
13. F.P. Bowden, A.J.W. Moore, D. Tabor, J. Appl. Phys. 14, 80 (1943)
14. F.P. Bowden, D. Tabor, The Friction and Lubrication of Solids, Part I (Clarendon Press, Oxford,
1950)
15. S. Korres, T. Feser, M. Dienwiebel, Wear 303, 202 (2013)
16. Y.S. Kim, J.S. Ha, W. J. Kim, in Des. Process. Prop. Adv. Eng. Mater. Pts 1 2, ed. by S.G.
Kang, T. Kobayashi (Trans Tech Publications Ltd, Stafa-Zurich, 2004), pp. 597–600
17. T. Feser, P. Stoyanov, F. Mohr, M. Dienwiebel, Wear 303, 465 (2013)
18. P. Stoyanov, P.A. Romero, T.T. Järvi, L. Pastewka, M. Scherge, P. Stemmer, A. Fischer, M.
Dienwiebel, M. Moseler, Tribol. Lett. (2013)
19. P. Stoyanov, P.A. Romero, R. Merz, M. Kopnarski, M. Stricker, P. Stemmer, M. Dienwiebel,
M. Moseler, Acta Mater. 67, 395 (2014)
20. P. Stoyanov, R. Merz, P. Romero, F.C. Wählisch, O.T. Abad, R. Gralla, P. Stemmer, M. Kop-
narski, M. Moseler, R. Bennewitz, M. Dienwiebel, ACS Nano 9, 1478 (2014)
21. P. Stoyanov, P. Stemmer, T.T. Järvi, R. Merz, P.A. Romero, M. Scherge, M. Kopnarski, M.
Moseler, A. Fischer, M. Dienwiebel, A.C.S. Appl, Mater. Interfaces 5, 6123 (2013)
22. S. Hanke, I. Samerski, J. Schöfer, A. Fischer, Wear 267, 1319 (2009)
11 In Situ Digital Holography for 3D Topography … 305

23. P. Stoyanov, P. Stemmer, T.T. Järvi, R. Merz, P.A. Romero, M. Scherge, M. Kopnarski, M.
Moseler, A. Fischer, M. Dienwiebel, ACS Appl. Mater. Interfaces 5, (2013)
24. P. Stoyanov, J.M. Shockley, M. Dienwiebel, R.R. Chromik, MRS Commun. 6, 301 (2016)
 Part IV
Computer Simulations
Chapter 12
Understanding the Tribochemistry
of Lubricant Additives by Ab initio
Calculations: The Case of Phosphites

M. Clelia Righi

Abstract The search for novel lubricants to improve the energy efficiency of engines
and help mitigate the environmental effects of carbon dioxide emissions has gained
increasingly importance in recent years. Commercial lubricants contain a wide range
of compounds including those that undergo tribochemical reactions and form friction-
or wear-reducing films. A microscopic understanding of tribochemical reactions
is of paramount importance for designing new, environmental-friendly lubricants.
However, many aspects of tribochemistry remain elusive due to the difficulties in
experimentally probing the sliding buried interface. Simulations can play a decisive
role in this context, in particular ab initio molecular dynamics (AIMD), where both
the ionic and electronics degrees of freedom are fully taken in into account. This is
essential for an accurate description of reactions in situations of enhanced reactivity
imposed by the tribological conditions. This chapter offers an example of application
of ab initio methods in tribochemistry. A twofold analysis of the reaction mechanisms
and effects of tribochemical reactions involving organophosphorus additives at iron
interfaces allowed to understand the mechanisms of function of phosphorus-based
additives in boundary lubrication.

12.1 Introduction

Many tribological phenomena can ultimately be described as chemical reactions

occurring in the presence of mechanical forces, i.e., tribochemical reactions (TRs).
One important example is represented by boundary lubrication, where tribologically-
induced modifications of the surfaces interacting with lubricant-additives (or with
other molecules present in the environment surrounding the sliding media) can dra-
matically change the adhesion and friction of the materials in contact. Therefore,

M. Clelia Righi (B)

Department of Physics, Informatics and Mathematics,
University of Modena and Reggio Emilia,
Via Campi 213/a, Modena, Italy
e-mail: mariaclelia.righi@unimore.it

© Springer Nature Switzerland AG 2018 309

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8_12
310 M. Clelia Righi

it is highly important understand their activation mechanisms. Our knowledge of

the reaction kinetics and thermodynamics driving forces at the open surface is not
sufficient for this aim, since the tribological conditions alter the reaction kinetics.
Tribological conditions include frictional heating, molecular confinement, shear
stress, and high pressure. Usually, the reaction rate is observed to increase, a phe-
nomenon referred to as “mechanical catalysis” in the field of mechanochemistry,
where chemical and materials syntheses are achieved with the application of mechan-
ical forces. Many aspects of mechanochemistry as well as tribochemistry remain
elusive due to the difficulties in directly probing the interface: From the analysis of
the reaction products, typically by means of post mortem spectroscopic analysis, it
is possible to acquire information on the type of chemical reactions occurred, but
very little can be inferred on the activation mechanisms, reaction pathways and rates.
Simulations can play a decisive role here, in particular ab initio molecular dynam-
ics (AIMD), where both the ionic and electronics degrees of freedom are fully taken
in into account. This is essential for an accurate description of bond-breaking and
bond-forming reactions in situations of enhanced reactivity imposed by the tri-
bological conditions. However, the use of AIMD in tribology has been very scarce
and most of the existing simulations of tribochemistry are based on force fields, i.e.,
on a parameterization of the atomic interactions. While many force fields are avail-
able for selected chemical environments, their transferability is often poor, especially
to describe the non-ordinary conditions present at the tribological interface. In this
chapter we show that ab initio methods can succesfully be applied to understand the
functionality of lubricant additives.
The design of lubricants to decrease friction and wear in machine components is
an important way to increase the energy efficiency of mechanical systems while tak-
ing into account restrictive environmental requirements and technological advances
[1]. Lubricants are formulated products composed of base oil and a package of addi-
tives designed for specific performance needs. The additives can be classified as
chemically active, i.e., designed to chemically interact with the surface and form
protective layers, or chemically inert, i.e., with the function of improving the phys-
ical properties of the bulk materials. Most of lubricant additives have a non-polar
part, usually consisting of a hydrocarbon chain, the functionality of which is to sol-
ubilize the molecule into the base oil and a functional polar group that reacts with
the surface. Here we consider extreme-pressure (EP) and anti-wear (AW) additives
where the key elements of the functional group are sulfur and phosphorus. EP, AW
additives are typically used in synchronizer systems, [2] which reduce the speed
difference between the shaft and the idler gear [3]. The synchronizing system is a
key component of manually operated vehicle transmissions, as fluent transmission
operations have become increasingly important to drivers in the recent years. Gear
oils must perform in conditions and applications that can vary drastically, therefore
they are enriched by additives that protect seals, improve thermal, oxidative and vis-
cosity stability, provide micropitting resistance, bearing corrosion protection, foam
resistance, enhanced demulsibility and load-carrying capacity. The concentration of
EP and AW Additives that work in severe conditions is higher in gear oils, which are
enclosed in gear boxes, than in engine oils [4, 5].
12 Understanding the Tribochemistry of Lubricant Additives … 311

S- and P-based compounds are largely used as EP and AW additives for gear
oils [6–8]. EP additives typically adsorb onto the metal surface either by physical or
chemical attraction [9, 10]. Under severe tribological conditions, they react with the
surfaces forming surface films that prevent the welding of opposing asperities and
avoid scuffing that is destructive to sliding surfaces under high loads [11].
The action mechanism of the sulfur-based EP/AW additives was first proposed by
Davey et al. [12] and then refined by Forbes et al. [13]. The sulfur compounds adsorb
on an iron surface and under tribological conditions form inorganic iron sulfide film,
which can reduce frictional wear and prevent seizure. The Fe–S formation under
boundary lubrication conditions has been identified in different studies [14–16].
Recently, Li et al. considered three novel S-containing alkyl phenylboric esters [17,
18]. After the tribotest, the worn surface was investigated by X-ray absorption near
edge structure spectroscopy (XANES) and Fourier transform infrared (FT-IR) spec-
troscopy. The surface analysis confirmed the presence of iron sulfide.
The chemical structure of P-containing additives influences the nature and the
tribological performances of the tribofilm as highlighted in [19–22], where phos-
phite and phosphate have been compared. Typically, organic phosphites function as
friction-modifiers whereas phosphates as anti-wear additives. Post mortem analy-
sis of the boundary lubrication film formed on frictional surfaces showed that an
iron phosphate compound is often obtained from phosphates [15, 22–24]. First, the
organic phosphate molecule adsorbs on the iron substrate and then, iron phosphate is
formed due to thermal/mechanical decomposition, [26–32] but also due to hydrolysis
[23, 25]. The formation of iron phosphate as a boundary lubrication film that pre-
vents wear-out has been observed with steel ball-on-disc tribotest followed by X-ray
photoelectron spectroscopy (XPS) [33]. Philippon et al. have proposed a decompo-
sition mechanism of a model EP additive, trimethylphosphite (TMPi), on nascent
iron surface [34] that leads to the formation of iron phosphide film under tribological
condition [35–37].
Ab initio calculations have been applied to investigate the functionality of P-
containing additives. A twofold analysis have been performed: (i) the tribochemical
reactions that lead to the formation of a tribofilm starting from P-containing additive
molecules have been identified by combining chemisorption studies and ab initio
molecular dynamics simulations; (ii) the functionality of the P-containing tribofilms
has been elucidated by first principles calculations of intrinsic tribological properties
of iron interfaces. The main stages of this analysis are described in the following to
provide an example of a computational protocol that can be applied to the research
in lubricant additives.
312 M. Clelia Righi

12.2 Combining First Principles Calculations and XPS

Experiments to Study the Adsorption of Lubricant
Additives

Heavy molecular weight organophosphorous additives like triphenylphosphite are

used as in commercial lubricants. The functional group containing phosphorous
affects the frictional properties while the long hydrocarbon chains have the function
to improve the additive solubility in oil. Since it is difficult to evaporate such heavy
molecular weight additives in UHV, 10−9 hPa, smaller molecules, having similar
active gropus, are used as model compounds for experimental research in lubricant
additives. Among all types of phosphite additives, trimethylphosphite (TMPi), with
chemical formulae P(CH3 O)3 , is probably the simplest compound containing the
phosphite ion, thus it was chosen as a model to provide a basic understanding on
the functionality of organophosphorous additives. First principles calculations based
on density functional theory (DFT) were applied to elucidate the mechanism of P
release from TMPi decomposition at the iron surface. The numerical results were
discussed in relation to the in situ XPS analysis performed on metallic iron after
TMPi thermally-controlled adsorption [38].
Spin-polarized DFT calculations were performed within the generalized gradi-
ent approximation proposed by Perdew–Burke–Ernzerhof (PBE) to describe the
exchange correlation functional [39]. The electronic wave-functions were expanded
on a plane waves basis and the ionic species were described by pseudopotentials.
The Fe(110) surface was considered since it is the most stable among the densely
packed iron surfaces [40]. The surface is modeled by means of periodic supercells
containing an iron slab of (4 × 4) in-plane size, e.g. 16 atoms per layer, and a vacuum
region 20 Å thick. It has been verified that the choice of the (4 × 4) in-plane size is
sufficient to avoid lateral interaction of TMPi with its periodic replicas. To identify
the microscopic mechanisms that govern the release of P and its adsorption onto the
Fe(110) surface by TMPi dissociation, the adsorption of TMPi was first studied and
then the reaction paths for molecular dissociation were calculated by means of the
Nudged elastic band (NEB) method [41]. The NEB method allows us to obtain the
minimum energy path (MEP), which is the path with the greatest statistical weight,
followed by the system in a transition from an initial to a final state. In the adsorption
study, the initial and final states corresponded to the undissociated and dissociated
adsorption configurations identified as the most stable ones.
Temperature-programmed reaction spectroscopy (TPRS), X-ray photoelectron
spectroscopy (XPS) and low-energy electron diffraction (LEED) [42] experiments
have shown that TMPi decomposes into adsorbed phosphorous and gaseous CO
and H2 via the methoxy (CH3 O) intermediate on clean Fe(110), leaving behind a
phosphide film. In agreement with these observations, the calculated reaction ener-
gies indicate that the molecular dissociation via methoxy detachment is a highly
exothermic process [38]. Furthermore, the calculations revealed that the adsorption
of molecular fragments is more favorable when the fragments are isolated on the
surface, i.e., they do not share metal atoms with neighboring adsorbed fragments.
12 Understanding the Tribochemistry of Lubricant Additives … 313

Fig. 12.1 Schematic representation of the TMPi dissociation into atomic phosphorus and adsorbed
methoxy groups on the Fe(110) surface (a). Initial and final states for each methoxy detachment
process (b) [31]. Fe is colored in blue, P in yellow, O in red, C in grey, and H in white

This result indicates that TMPi dissociative adsorption is favored in the presence of
open Fe sites, as evidenced by the experiments that highlighted the key role played
by the nascent metallic surfaces on the tribochemistry of phosphites [43, 44].
A schematic representation of TMPi dissociation in atomic phosphorous and
methoxy groups on the Fe(110) surface is offered in Fig. 12.1a, while the initial
and final states of each reaction of methoxy detachment are reported in Fig. 12.1b in
a ball-and-stick representation. The TMPi molecule binds to the surface forming a
P–Fe bond. Via methoxy detachment, the number of P–Fe bonds increases until the
P atom is adsorbed into a long bridge site, where it is bonded to four Fe atoms. The
energy of the dissociated adsorption configurations decreases with the number of
detached methoxy groups, and the most stable adsorption configuration is obtained
when all the methoxy groups are detached and atomic P is released at the surface.
The passivation of the Fe surface with phosphorus through TMPi dissociation is thus
an energetically favorable process.
The reaction paths and corresponding energy barriers obtained by means of the
NEB method are reported in Fig. 12.2. The results indicate that an activation energy
of E A  0.75 eV is necessary for the first methoxy detachment, and produces an
energy gain of E R  −0.73 eV (Fig. 12.2a). The remaining P(CH3 O2 ) is less stable
than the entire molecule, thus the second methoxy detachment requires a lower
activation energy (E A  0.23) and produced an higher gain (E R  −1.20) than the
first methoxy dissociation and (Fig. 12.2b). Such energy gain is sufficient to promote
the decomposition of the residual P(CH3 O) fragment in elemental P and adsorbed
methoxy, which is still an exothermic process (E R  −0.78).
The formation of iron phosphide from the TMPi adsorption onto metallic iron
at different temperatures has been investigated thanks to in situ surface analyses.
Figure 12.3 shows the P2p XPS spectra of phosphorous obtained in situ after adsorp-
tion of TMPi onto an etched steel surface at various temperatures. The chemisorption
314 M. Clelia Righi

Fig. 12.2 Reaction paths, energies, E R , and barriers, E A , for subsequent methoxy detachment from
the the TMPi molecule [31]

of TMPi molecules is evidenced by the appearance of the P2p XPS peak at 133.2 eV
at 100 °C. The characteristic peak of iron phosphide appears clearly at a tempera-
ture of 300 °C with a distinct characteristic binding energy of 130 eV. These results
strongly suggest that the formation of iron phosphide is thermally-activated. In fact,
temperatures above 200 °C increase the formation of phosphide to a large extent,
providing sufficient energy to dissociate the TMPi molecule. Although iron phos-
phide formation is evidenced, undissociated TMPi molecules are also observed on
the surface, as shown by the characteristic binding energy at 133.5 eV.
The combined computational and experimental analysis above described revealed
that the full TMPi dissociation, leading to phosphorus release, can occur once the
activation energy is provided, e.g. by heating the sample. In the next section the
effects of mechanical stresses (load and shear) in the activation of the reaction will
be considered.
12 Understanding the Tribochemistry of Lubricant Additives … 315

Fig. 12.3 In situ X-ray

photoelectron spectroscopy
P2p lines recorded after
TMPi thermally controlled
adsorption onto etched steel,
e.g. metallic iron [31]

12.3 Effects of Mechanical Stresses on Additive

Decomposition at the Tribological Interface

The tribological conditions, which include high temperatures, molecular confine-

ment, load and shear stresses can highly enhance the rate of chemical reactions.
The first proposed mechanisms [45] attributed the increase of the reaction rates to
extremely high temperatures, “flashes temperatures,” that develop at few discrete
“hot spots,” where the contact between the two surfaces occurs [46]. Such tempera-
ture increase has been also proposed to produce local excitations, “magama plasma,”
that decay rapidly into a local heating of the surface [47]. However, recent works have
shown that tribochemical reactions can take place even though the temperature raise
during sliding is negligible or limited and highlighted the primary role of mechanical
stresses in the activation of chemical reactions [48–57]. AIMD simulations of tribo-
chemical reactions involving water molecules confined at diamond interfaces have
shown that confinement under pressure is the driving force for molecular dissociation
and reaction rates are functions of load. In particular, it was found that the higher
the load, the more effective the surface passivation by water dissociation is, which
in turn leads to a reduction of friction [58, 59].
Applying AIMD to study the tribochemistry of additives at iron interfaces repre-
sents a great computational challenge due to difficulty imposed by the presence of
a metallic system where the ionic and electronic degrees of freedom can exchange
energy due to the absence of a band gap. In addition, the magnetic character of iron
imposes the use of spin polarization. This challenge has been successfully faced [60]
by adopting the Born Oppenheimer scheme, where the ions are moved according to
the Hellmann–Feynman forces obtained from the total electronic energy, which is
316 M. Clelia Righi

minimized at fixed ionic position every MD step. In this way the ionic and electronic
degrees of freedom are decoupled and the dynamics of metallic systems, even includ-
ing magnetization, can be described in accurate way. Iron interfaces, composed by
two self-mated Fe(110) surfaces, were modeled by means of two slabs with (4 × 4)
in-plain size. The Brillouin zone sampling was realized by means of a (2 × 2 × 1)
Monkhorst Pack grid [61]. The bottom layer of the lower slab (the substrate) was
held rigid, while the upper slab (the counter-surface) was moved at constant velocity
of 200 m/s. A vertical force was applied to the atoms belonging to the top layer of
the counter-surface to model an applied pressure the value of which was varied to
evaluate the dependence of the reaction rates on load. In the initial configuration,
common to all the performed simulations, a TMPi molecule per cell (of ~100 Å2
area) was adsorbed in its most stable configuration on the iron substrate and the sys-
tem was relaxed under the effects of the applied pressure. After the relaxation, the
system was equilibrated at a constant temperature of 300 K and the countersurface
was laterally moved with constant velocity.
In the first simulation, the pressure was fixed to 0.5 GPa. During a simulated
time interval of 7 ps the system dynamics consisted in molecular vibrations, mainly
rotations of each methyl group around the axis along the CO bond, and we did not
observe any dissociative reaction. By increasing the pressure to 2 GPa, molecular
dissociation is, instead observed after 1 ps. As can be seen in Fig. 12.4a, b, as soon
as the iron counter-surface gets closer to the molecule, the oxygen atoms become
attracted by the iron countersurface, the PO bonds get stretched and finally dissoci-
atevily adsorb on the countersurface while elemental phosphorus remains adsorbed
on the substrate (Fig. 12.4c).
The reaction path just described is perfectly consistent with that predicted by the
NEB method and observed experimentally at the open surface (Sect. 12.2). Molecular
dissociation occurs, in fact, through methoxy detachment and the importance of clean
Fe sites for favoring methoxy adsorption appears evident in the dynamic simulations,
where the clean Fe counter-surface attracts methoxy groups. We also observed that
the released elemental phosphorus at the end of the MD simulation was located at
a long bridge site, which was indicated as the most favorable site for P adsorption
at the Fe(110) surface (Fig. 12.1a). It is interesting to notice that the molecular
dissociation observed during the dynamic simulation occurred as expected on the
basis of thermodinamical driving forces. However, the activation time for TMPi
dissociation observed in the dynamic simulation under tribological conditions turned
out to be much shorter than the activation time obtained by applying the Arrhenius law
at room temperature and considering a dissociation barrier of 0.75 eV, as reported in
Sect. 12.2. This result, along with the different trajectories that we obtained for the
two simulations at 0.5 and 2 GPa, indicate that the mechanical stresses play a crucial
role in accelerating the reaction rates.
In the second part of the dynamic simulation at 2 GPa pressure, it was observed
the tendency of phosphorus at low concentration on the substrate to interact with the
counter-surface. Once Fe–P bonds were established across the interface the surface
separation was reduced (Fig. 12.4d) and a direct contact between few metal atoms
occurred (Fig. 12.4e). The direct metal-metal contact was sufficient to promote the
12 Understanding the Tribochemistry of Lubricant Additives … 317

Fig. 12.4 Snapshots acquired during AIMD simulation of TMPi tribochemistry at iron inter-
faces under 2 GPa pressure and room temperature. The simulation time increases from panel (a) to
panel (f) [60]

cold sealing of the two iron surfaces and the molecular fragments remained embed-
ded in the iron matrix (Fig. 12.4f). This event in real life would be detrimental,
corresponding e.g. to engine seizure. An increased phosphorous coverage is crucial
to prevent direct metal-metal contact [62].
A closer inspection of the molecular fragments embedded in the iron matrix
revealed that the methoxy group got dissociated into H atoms on CO molecules.
This result, in agreement with the observation that the dissociation of a CO molecule
on the Fe(110) surface is an endothermic process, [63] reveals that O adsorption
does not compete with P adsorption from TMPi decomposition. This can explain the
higher efficiency of phosphites than phosphates observed in the experiments [62].

12.4 Functionality of Phosphorous in Reducing Adhesion

and Friction

Gas phase lubrication (GPL) experiments are ideal experiments to provide under-
standing on the functionality of extreme pressure additives because the chemical
compounds are tested without the presence of the base oil, thus mimicking boundary
lubrication conditions. In a gas phase lubrication experiment the friction coefficient
of a steel-on-steel sliding contact was first measured in ultra high vacuum (UHV)
and then in the presence of 1 hPa of TMPi. As can be seen in Fig. 12.5a, TMPi turned
318 M. Clelia Righi

out to be very effective in reducing the friction coefficient of steel. The wear tracks
formed on the surface of the steel flat at the end of the experiments in the presence
of TMPi were observed by optical microscopy (Fig. 12.5b). A tribofilm was clearly
evidenced. To check this assumption, in situ surface analyses were carried out inside
and outside the wear tracks formed on the flat immediately after the friction experi-
ments. This is made possible by transferring the steel flat from the tribometer to the
analysis chamber thanks to the UHV intermediate preparation chamber. The signif-
icant advantage of in situ analyses is that the signals are not changed and/or masked
by an oxidation in air as in the case of ex situ surface analyses. This permitted to
clearly identify iron phosphide inside the wear tracks (Fig. 12.5b). Therefore, a film
composed by iron and elemental phosphorus was generated during the tribological
test and it was responsible for the observed friction reduction.
First principles calculations were carried out to shed light into capability of phos-
phorus to reduce iron friction.
A computational protocol to calculate from first principles two intrinsic tribolog-
ical properties of solid interfaces, namely the work of separation W and the ideal
interfacial shear strength, τ , was proposed in [64] These properties, which corre-
spond to the energy required to separate two surfaces from contact and to the static
friction force per unit area, are ruled by the physical/chemical interactions between
the surfaces in contact. First principles calculation based on Density Functional
Theory (DFT) can accurately describe surface-surface interactions, offering the pos-
sibility to characterize in silico the adhesive and shear strength of materials and their
change due to surface chemical modification. The protocol was applied to identify
the effects of surface passivation in diamond [65], to compare the friction reduction
properties of phosphorus and sulfur [66], and to understand the lubricating proper-
ties of graphene [67]. The protocol has been recently implemented as a workflow
to perform high throughput calculations of the intrinsic tribological properties of
solid interfaces [68]. Here we show how it can be applied to quantify the effects of
phosphorus on the adhesion and resistance to sliding og iron.
It is well known that phosphorus segregates in iron and forms 2D chemisorbed
overlayers, [69] but there is presently no experimental information on their structure
in the literature. The most favorable adsorption site for P at the Fe(110) surface,
was indentified by means of first principles calculations by comparing the adatom
energy at high symmetry locations [20]. The long bridge site, having the highest
coordination, turned out to be the most stable adsorption location. Then, the effects of
coverage were evaluated by comparing the P adsorption energy in the configurations
represented in Fig. 12.6. It turned out that a coverage increase with respect to 0.25
ML (Fig. 12.6b) destabilizes the overlayer. In particular, the adsorption energy per
atom increases by 6% for a coverage increase to 0.5 ML (Fig. 12.6c) and by 21% at
full coverage (Fig. 12.6d) [70].
The interfaces were modeled by self-mating the optimized surfaces represented in
Fig. 12.6. A ball-stick representation of the optimized interface structures obtained
for the PES minima is presented in Fig. 12.7, the corresponding work of separation,
W , and the distance between interfacial Fe layers, zeq , are reported in Table 12.1.
12 Understanding the Tribochemistry of Lubricant Additives … 319

Fig. 12.5 Steady-state friction coefficient obtained under GPL at room temperature for steel/steel
friction pair in vacuum and in the presence of TMPi (a). High resolution XPS spectra recorded
in situ after the tribo-test in the presence of TMPi with and without the wear tracks (b)

Fig. 12.6 Top view representation of the Fe(110) surface with different P coverages θ [20]
320 M. Clelia Righi

Fig. 12.7 Lateral view representation of the iron interfaces obtained by self-mating the surfaces
with different P coverages represented in Fig. 12.6. The lateral and vertical relative position of the
self-mated surfaces correspond to the optimal ones [62]

Table 12.1 The work of separation, W , and the equilibrium distance, zeq , separating the innermost
Fe layers at the interface are reported for the PES absolute minimum
Interfacial Wsep(J/m2 ) Z eq (Å) τ x (GPa) τ y (GPa)
coverage
θ 0 4.8 2.0 10.1 10.1
θ p  0.25 1.71 3.2 4.0 4.0
θ p  0.5 0.6 4.3 3.3 2.1
θ p  1.0 0.04 7.3 0.1 0.1

The shear strengths, τ , are calculated for the MEPs along the [−110] and [001] sliding directions
[20]

The adhesion energy was calculated for different relative lateral positions, con-
structing in this way the PES for the sliding interface. In Fig. 12.8 the PESes of
the bare interface and that obtained including interstitial phosphorus at an interfa-
cial coverage of 0.25 are compared. A common energy scale is used, the blue color
indicates the PES minimum, which is taken as reference and the red color is for the
PES maximum. As can be seen in Fig. 12.8a, the minima of the PES for the clean
interface are located at LB positions, i.e., where a new layer would be positioned
in the bcc structure of iron bulk. Two clean iron surfaces in contact undergo a cold
sealing (Fig. 12.7a): the optimized interfacial separation corresponds to the interlayer
distance in Fe bulk, the work of separation W sep is equal to the energy required to
form two Fe(110) surfaces and the shear strength τ x,y  10 GPa is typical of bulk
iron.
The situation is completely altered if the mating surfaces are P-terminated. At
interfacial coverage θ P  0.25 (Fig. 12.7b) the distance between the interfacial Fe
layers increases by 60% and the adhesion decreases by one order of magnitude with
12 Understanding the Tribochemistry of Lubricant Additives … 321

Fig. 12.8 Two-dimensional representation of the PESes obtained for the clean (a) and the P-
containing iron interface (P atoms are in yellow) at relative concentration of 0.25 (b). A common
energy scale is used for the potential corrugation [62]

respect to the clean iron interface. A decreased adhesion gives rise to a smoother PES,
as can be seen in Fig. 12.8b, and lower frictional forces: the shear stress decreases
by 60%.
In Fig. 12.7c, d it can be seen that by increasing the concentration of interfacial
phosphorus to θ P  0. 5 and θ P  1, the distance between the surfaces increases
and chemical bonds are no longer present across the interface, the surfaces are hold
together by physical interactions. The disappearance of chemical bonds produces a
considerable decrease in the adhesion and shear strength (last two columns of Table
12.1). In particular, in the case of two fully passivated surfaces in contact the short-
range repulsion between interfacial P layers causes a decrease of adhesion and shear
strength by two orders magnitude with respect to clean iron. The magnitude of the
physical interactions are most likely underestimated by DFT-PBE. However, such
consistent changes give a clear indication of the dramatic effect of phosphorus in
reducing friction and adhesion at iron interfaces, in agreement with the tribolog-
ical experiments and with the well known effect of iron embrittlement caused by
phosphorous.
The above described results clearly indicate that atomic phosphorsus is able to
effectively reduce the adhesion and shear strength of iron, in excellent agreement
with the experimental observation of a friction reduction promoted by the formation
of an iron phosphide tribofilm. This clearly exemplify the power of first principles
calculation in the research on lubricant material, as chemical surface modifications,
accurately described by first principles methods, can impact macroscale properties
as the friction coefficient.

12.5 Conclusions

The power of ab initio methods in the research in lubricant materials has been high-
lighted through an example study on the functionality of organophosphorus additives.
322 M. Clelia Righi

GPL experiment and spectroscopic analysis revealed that the ability phosphite addi-
tives to reduce friction relies on the formation of an iron phosphide tribofilm. Ab
initio calculations were applied to provide insights into these experimental findings
and a twofold analysis was performed with the aim at identifying: (i) the microscopic
mechanisms that lead to iron-phosphide formation through phosphite decomposition,
(ii) the functionality of phosphorus in reducing friction of iron. The first issue was
addressed by studying the molecular decomposition at the surface and at the tribolog-
ical interface by means of ab initio molecular dynamics simulations. The atomistic
mechanisms for P release were clearly identified and the role of mechanical stresses
in promoting the process were highlighted. The second issue was faced by applying
a computational protocol that allows one to calculate intrinsic interfacial properties
as adhesion and shear strength, which provided results in excellent agreement with
the experimental observations.

References

1. H. Spikes, Tribol. Lett. 60(5), 1–26 (2015)

2. X. Wanli, Z. Wai, S. Bin, X. Ximeng, Wear 328–329, 475–479 (2015)
3. A. Bouffet, SAE Paper 01-2022 (2004)
4. J. Vinci, US20060276352, Lubrizol Corporation (2006)
5. J. Zakarian, M. Haire, US20060252657, Chevron Texaco Corporation (2006)
6. S. Matsui, H. Komatsubara, US20070287643, Nippon Oil Corporation (2007)
7. T. Craig, T. Thomas, WO2007111741, Lubrizol Corporation (2007)
8. O. Furlong, F. Gao, Tribol. Int., 2008, 40, 699
9. Z. Tang, S. Li, Curr. Opin. Solid State Mater. Sci., 18, 119 (2014)
10. H. Spikes, Lubr. Sci. 20, 103 (2008)
11. J. Van Rensselar, Power Transm. Eng. (2003)
12. W. Davey, et al., Wear 1, 291 (1958)
13. E.S. Forbes, A.J.D. Reid, ASLE Trans. 16, 50 (1973)
14. R.C. Coy, T.F.J. Quinn, ASLE Trans. 18, 163 (1975)
15. S. Mori, K. Hori, Y. Tamai, Junkatsu 27, 505 (1982)
16. J. Tannous, B.M.I. de Bouchet, T. Le Mogne, P. Charles, J.M. Martin, Tribol.-Mater. Surf.
Interfaces 1, 98 (2007)
17. Z. Li, Y. Li, Y. Zhang, T. Ren, Y. Zhao, RSC Adv. 4, 25118 (2014)
18. J. Li, B. Fan, T. Ren, Y. Zhao, Tribol. Int. 88, 1 (2015)
19. A. Riga, J. Cahoon, W.R. Pistillo, Tribol. Lett. 9, 219 (2000)
20. M.I. De Barros-Bouchet, M.C. Righi, D. Philippon, S. Mambingo-Doumbe, T. Le-Mogne, J.M.
Martin, A. Bouffet, RSC Adv. 5, 49270 (2015)
21. F. Gao, P.V. Kotvis, D. Stacchiola, W.T. Tysoe, Tribol. Lett. 18, 377–384 (2005)
22. F. Gao, O. Furlong, P.V. Kotvis, W.T. Tysoe, Langmuir 20, 7557–7568 (2004)
23. O. Beek, J.W. Givens, E.C. Williams, Proc. R. Soc. London Ser. A 177, 103 (1940)
24. D. Godfrey, ASLE Trans. 8, 1 (1965)
25. F.T. Barcroft, S.G. Daniel, Trans. ASME Ser. D 87 (1965)
26. Y. Yamamoto et al., Wear 50, 343 (1978)
27. D. Sung, A.J. Gellman, Tribol. Lett. 13, 1 (2002)
28. A.W. Holbert, J.D. Batteas, A. Wong-Foy, T. S. Rufael, C.M. Friend, Surf. Sci. 401, l437 (1998)
29. D. Ren, G. Zhou, A.J. Gellman, Surf. Sci. 61, 475 (2001)
30. D. Ren, A.J. Gellman, Tribol. Int. 34, 353 (2001)
12 Understanding the Tribochemistry of Lubricant Additives … 323

31. M.C. Righi, S. Loehlé, M.I. De Barros-Bouchet, D. Philippon, J.M. Martin, RSC Adv. 5,
101162 (2015)
32. A. Rossi, F.M. Piras, D. Kim, A.J. Gellman, N.D. Spencer, Tribol. Lett. 23, 197 (2006)
33. K. K. Mistry, A. Morina, A. Erdemir and A. Neville, Tribol. Trans., 2013, 56, 4
34. D. Philippon, et al., Tribol. Int. 44, 113 (2011)
35. F.T. Barcro, Wear 3, 440 (1960)
36. D. Philippon, M.I. De Barros Bouchet, Th. Le Mogne, E. Gresser, J.M. Martin, Tribol.-Mater.
Surf. Interfaces 1, 113 (2007)
37. D. Philippon, M.I. De Barros-Bouchet, Th. Le-Mogne, B. Vacher, O. Lerasle, J.-M. Martin,
Thin Solid Films 524, 191 (2012)
38. M.C. Righi, S. Loehlé, M.I. de Barros Bouchet, D. Philippon, J.M. Martin, RSC Adv. 5,
101162–101168 (2015)
39. J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996)
40. P. Blonski, A. Kiejna, Surf. Sci. 601, 123 (2007)
41. G. Henkelman, B.P. Uberuaga, H.J. Jonsson, J. Chem. Phys. 113, 9978 (2000)
42. A.W. Holbert, A. Batteas, A. Wong-Foy, T.S. Rufael, C.M. Friend, Surf. Sci. 401, L437–L443
(1998)
43. D. Philippon, M.-I. De Barros-Bouchet, T. Le Mogne, O. Lerasle, A. Bouffet, J.-M. Martin,
Tribol. Int. 44, 684 (2011)
44. D. Philippon, M.-I. De Barros-Bouchet, T. Le Mogne, E. Gresser, J.-M. Martin, Tribol.-Mater.
Surf. Interfaces 1, 113 (2007)
45. T.E. Fischer, Ann. Rev. Mater. Sci. 18, 303 (1988)
46. G. Heinicke, Tribochemistry (Carl Hanser Verlag, Munchen, 1984)
47. P.A. Thiessen, K. Meyer, G. Heinicke, Grundlagen der Tribochemie (Akademie Verlag, Berlin,
1967)
48. N.N. Gosvami, et al., Science 348, 102 (2015)
49. B. Gotsmann, M.A. Lantz, Phys. Rev. Lett. 101, 125501 (2008)
50. C.R. Hickenboth et al., Nature 446, 423 (2007)
51. O.J. Furlong et al., ACS Appl. Mater. Interfaces 3, 795 (2011)
52. O.J. Furlong et al., RSC Adv. 4, 24059 (2014)
53. A. Ghanbarzadeh et al., Tribol. Lett. 1, 61 (2016)
54. J.R. Felts et al., Nat. Commun. 6, 6467 (2015)
55. D.B. Jacobs, R.W. Carpick, Nat. Nanotechnol. 8, 108 (2013)
56. H.L. Adams, M.T. Garvey, S.U. Ramasamy, Z. Ye, A. Martini, W.T., Tysoe, J. Phys. Chem. C
119, 7115 (2015)
57. J. Zhang, H. Spikes, Tribol. Lett. 63, 24 (2016)
58. S. Kajita, M.C. Righi, Carbon 103, 193 (2016)
59. G. Zilibotti, S. Corni, M.C. Righi, Phys. Rev. Lett. 111, 146101 (2013)
60. S. Lohelé, M. C. Righi, Lubricants 6, 31 (2018)
61. H.J. Monkhorst, J.D. Pack, Phys. Rev. B 13, 5188 (1976)
62. M.I. De Barros-Bouchet et al., RSC Adv. 5, 49270 (2015)
63. D.E. Jiang, et al., Surf. Sci. 570, 167 (2004)
64. G. Zilibotti, M.C. Righi, Langmuir 27, 6862 (2011)
65. M.I. De Barros Bouchet, G. Zilibotti, C. Matta, M.C. Righi, J. Phys. Chem. C 116(12),
6966–6972 (2012)
66. M.C. Righi et al., RSC Adv. 6, 47753 (2016)
67. P. Restuccia, M.C. Righi, Carbon 106, 118–124 (2016)
68. P. Restuccia et al. Comp. Mat. Sci. 154, 517 (2018)
69. W. Arabczyk, T. Baumann, F. Storbeck, H.J. Mussig, A. Meisel, Interaction between phospho-
rus and oxygen on the Fe(111) surface. Surf. Sci. 189, 190–198 (1987)
70. G. Fatti, P. Restuccia, C. Calandra, M.C. Righi (to be published)
Index

A Brilliance, 238
Ab Initio Molecular Dynamics (AIMD), 310 Butylated triphenyl phosphorothionate
Abrasion, 31 (b-TPPT), 178
Additives, 111 1-butyl-3-methyl-imidazolium ([BMIM]), 265
Adhesion, 31
Ag/AgCl – Electrode, 251 C
AISI 52100, 113 CoCr29Mo6, 34
Alkyl cyanobiphenyl (ACB), 225 17CrNiMo7-6, 48
Alkyl cyanocyclohexylphenyl (ACC), 225 18CrNiMo7-6, 36
Aluminum, 302 CoCrMo, 40
Amorphous carbon (a-C), 83 Collision cascade, 6
Analytical Electron Microscopes (AEM), 16 CrN, 238
Anti-Wear (AW) additives, 310 Cross sections, 7
Antimony trioxide, 25 CuZn5, 295
Antiwear mechanism, 201 Cyclic Strain Gradient, 33
Atomic Force Microscopy (AFM), 248 Cyclic Voltammetry (CV), 251
ATR crystal, 136 Cyclic Voltammogram, 252
ATR-FTIR, 177
Attenuated Total internal Reflection Infrared D
spectroscopy (ATR-IR), 136 Dark-field (DF), 15
Attenuated Total Reflection (ATR) IR Debris, 302
spectroscopy, 227 Delamination, 34
Au(111), 254 Density Functional Theory (DFT), 312
Auger stectroscopy, 138 Dialkyldithiophosphate (DDP), 159
Austenitic materials, 34 Diamond Anvil Cell (DAC), 224
Diamond-like carbon coating, 20
B Digital Holographic Microscopy (DHM), 290,
Beer-Lambert’s Law, 65 291
Bis(trifluoromethylsulfonyl)imide (Tf2N), 263 Dimethyl disulphide (DMDS), 141, 146
Blocking dislocations, 34 Dimple grinding, 52
Borate, 111 DLC, 176
Boundary lubrication, 108
Brass, 295 E
Bremsstrahlung, 17 EDS, 124
Bright-field (BF), 15 Electrochemical AFM (EC-AFM), 253

© Springer Nature Switzerland AG 2018 325

M. Dienwiebel and M.-I. De Barros Bouchet (eds.), Advanced Analytical Methods
in Tribology, Microtechnology and MEMS,
https://doi.org/10.1007/978-3-319-99897-8
326 Index

Electrochemical Double Layer (EDL), 249 I

Electrochemical Friction Force Microscopy IF-MoS2, 274, 275
(EC-FFM), 247, 254 Inner Helmholtz Plane (IHP), 249
Electrochemically modified Surface Force In situ AES, 117
Apparatus, 253 In-situ AFM, 182
Electron Backscattering Diffraction (EBSD), 3, In situ IR analysis, 217
10 In-situ lift-out, 8
Electron Energy Loss Spectroscopy (EELS), 44 In situ TEM, 274
Electron microscopy, 3 In-situ XANES, 182
Emulsions, 228 In-situ XAS, 183
EN ISO 25178, 291 Interfacial Force Microscope (IFM), 192
Entrance Grid Bias (EGB), 71, 73 Inverse pole figure maps, 13
Environmentally Controlled Analytical Ion damage, 6
Tribometer (ECAT), 109 Ionic Liquids (ILs), 263
Ethyl-3-methyl-imidazolium (EMIM), 265 Ion milling, 52
Extended X-ray Absorption Fine Structure IR spectrum, 221
(EXAFS), 64 ISE imaging, 7
Extreme-Pressure (EP), 310
K
F Kikuchi bands, 10
Fe(110), 316
Fe–O–S ternary system, 174 L
FeOOH, 261 Lateral force map, 262
Fermi’s Golden Rule, 69 LI-ZDDP, 168
Ferritic steel, 41 LIGA, 20
Fluorescence Yield (FY), 68, 71, 189 Ligand exchange reaction, 163
Focused Ion Beam (FIB), 3, 5, 52, 121 Liquid Metal Ion Source (LMIS), 5
Fourier-Transform Infrared spectroscopy Low-Energy Electron Diffraction (LEED), 143,
(FTIR), 216 312
Frictogram, 259 Lowest Unoccupied Molecular Orbital
(LUMO), 67
G
Gas Phase Lubrication (GPL), 108, 130, 146, M
317 Magnetically-guided imaging NEXAFS
Glassy polyphosphate, 188 spectroscopy, 91
Glycol, 224 Main wear mechanisms, 32
Gold oxide, 261 Mechanical mixing, 32, 49
Grain contrast images, 6 Metallic alloys, 3
Grease, 229 Methane, 132
Mica, 256
H Microanalysis, 16
Hard and Soft Acids and Bases (HSAB), 165 Microelectromechanical systems (MEMS), 20
Heat-Assisted Magnetic Recording (HAMR), Micro-FTIR, 216
83 Molecular Dynamics (MD) simulations, 88
Hexadecane, 301 Molybdenum disulfide (MoS2), 25
1-hexanethiol(C6H13-SH), 113, 116 Multivariate Statistical Analysis (MSA), 18
1-hexanethiol 1-hexyl-3-methyl-imidazolium
(HMIM), 265 N
Hip joints, 48 N-oleoylsulcosine, 231
Hydrogen-free tetrahedral amorphous carbon N-alkanethiols, 112
(ta-C), 88 NaClO4, 256
Hydrogenated amorphous carbon (a-C:H), 75, Nanocompression, 275
177 Nanocrystalline metals, 23
Hydrolytic decomposition, 169 Nanoindentation, 192
Index 327

Nanoindenter, 274 Schwoebel-Ehrlich barrier, 257

NaOH, 261 Selected-Area Diffraction (SAD), 15
Near edge X-ray absorption fine structure Self-Assembled Monolayers (SAMs), 112
(NEXAFS), 63 SEM cross section, 36
NEXAFS spectroscopy, 63 Shake-up structures, 68
Nudged Elastic Band (NEB) method, 312 Shear-induced surface-to-bulk transport, 155
Silicon-and oxygen-containing hydrogenated
O amorphous carbon (a-C:H:Si:O), 77
Oiliness agent, 220 Silicon oxide-doped diamond-like carbon
Oleic acid, 221 (SiOx-DLC), 85
Outer Helmholtz Plane (OHP), 249 Slide-to-Roll Ratio (SRR), 187
Spacer Layer Imaging Method (SLIM), 190
P Spectral image analysis, 17
P(CH3O)2, 313 Standard Hydrogen Electrode (SHE), 251
P-based additives, 163 Sum Frequency Generation (SFG), 227, 265
Partial Electron Yield (PEY), 71 Surface Fatigue, 31
Perfluoropolyether (ZDOL), 87 Surface-Enhanced Raman Spectroscopy
Phosphate, 111 (SERS), 135
Phosphite, 111 Surface-Force Apparatus (SFA), 189
Photoabsorption edge, 66 Synchrotron, 63
Photoelectric absorption coefficient, 65 Synchrotron beamline, 70
Photoemission Electron Microscopy (PEEM), Synchrotron radiation, 242
71, 87
Piston ring, 43 T
Plasma source ion column, 5 TEM-AFM holder, 282
Plowing friction, 293 TEM-Nanoindentor, 275
Point of Zero Charge (PZC), 249 Temperature-Programmed Desorption (TPD),
Polarized IR spectroscopy, 225 147
poly a-olefin (PAO), 147, 294 Temperature-Programmed Reaction
Polybutenyl saccimide, 231 Spectroscopy (TPRS), 312
Polybutylene glycol, 222 Tenacity, 196
Polypropylene glycol (PPG), 222 Tetrafluorobarate (BF4), 263
Precursor gas, 7 Thermal decomposition, 164
Pressure-viscosity coefficient, 222 Thermal films, 164
Principal Components Analysis (PCA), 18 Thermal liquid cell, 183
Protective layer, 7 Thermo-oxidative decomposition, 166
PTWA Coating, 43 Thickener, 229
Ti-doped (Ti-C:H) DLC coating, 177
Q TiAl6V4, 48
Quasi-Reference Electrode (QRE), 251 TiAlN, 241
Time-of-Flight Secondary-Ion Mass
R Spectrometry (ToF-SIMS), 121
Ratio of sliding to rolling, 187 Topography, 296
“Redhead” equation, 149 Total Electron Yield (TEY), 71, 189
Reference Electrode (RE), 251 Transition Dipole Moment (TDM), 69
Reflection-Absorption Infrared Spectroscopy Transition Metal Dichalcogenides (TMD), 25
(RAIRS), 135 Transition zone, 32
Roughness, 291 Transmission Electron Microscopy (TEM), 5,
Running-in, 294 12, 29, 30, 121
Transmission Kikuchi Diffraction (TKD), 8, 10
S Tribochemical reactions, 31, 32, 111, 309
Sandia’s Automated eXpert Spectral Image Tribochemistry, 130
Analysis (AXSIA), 18 Tribocorrosion, 249
Saturated Calomel Electrode (SCE), 251 Tribofilm, 116
328 Index

Tribomaterial, 39 Wear particle preparation, 54

Tricresylphosphate (TCP), 204 Wear particles, 49
Trimethylphosphite (TMPi), 311 Wear-induced recrystallization, 20
Tris(pentafluoroethyl)trifluorophosphate Wire drawing, 231
(FAP), 263, 265 Working Electrode (WE), 251
Tungsten, 298 WS2 nanoparticle, 279
Tungsten disulfide (WS2), 25
X
U X13CrMnMoN18-14-3, 34
Ultra-High Vacuum (UHV), 108, 140, 160 X5CrNiMo17-13-2, 34
Ultra-mild sliding wear, 29, 30 X6CrNiMoN22-10-4-3, 34
Ultrananocrystalline diamond (UNCD), 75 X-ray Photoelectron Spectroscopy (XPS), 110,
Urea-type thickener, 230 312
X-ray Absorption Near Edge Structure
V (XANES), 64, 238
Viscosity Index Improver (VII), 222 X-ray Photoemission Electron Microscopy
(XPEEM), 190
W
Water-based lubricants, 229 Z
Wavy slip, 34 ZDDP tribofilm, 166
Wear particle analyses, 47 Zincdialkyldithiophosphate (ZDDP), 159