SANDAR MYO MYINT · My sincere thanks also go to Dr. Su Zhoucheng, Dr. Song Shaoning, Dr. Muhammad Ridha and Mr. Alvin from Applied Mechanics Group. Special thanks to my colleagues,

TRIBOLOGICAL STUDY AND FINITE ELEMENT

MODELLING OF SURFACE TEXTURES ON SU-8

AND PDMS POLYMER SURFACES

SANDAR MYO MYINT

NATIONAL UNIVERSITY OF SINGAPORE

2015

TRIBOLOGICAL STUDY AND FINITE ELEMENT

MODELLING OF SURFACE TEXTURES ON SU-8

AND PDMS POLYMER SURFACES

SANDAR MYO MYINT

(B.E, YTU, M.E, YTU, M. Sc, NUS)

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF MECHANICAL ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2015

Declaration

i

Declaration

I hereby declare that this thesis is my original work and it has been written by

me in its entirety. I have duly acknowledged all the sources of information

which have been used in the thesis

This thesis has also not been submitted for any degree in any university

previously.

________________________________

Sandar Myo Myint

21 August 2015

sdmm

Stamp

Acknowledgement

ii

Acknowledgement

First and foremost, I would like to express my deeply gratitude to my

supervisors, A/Prof. Vincent Tan Beng Chye and Dr. Sujeet Kumar Sinha for

giving me an opportunity to work with them and for their valuable guidance,

encouragement and support throughout the period of my PhD. Secondly, my

thanks to Dr. Bhatia, C. S., co-supervisor, for his help and support. This

dissertation could not be finished without their generous guidance, help and

support.

I would also like to express my thanks to Dr. Satyanarayana Nalam

and Dr. Myo Minn for helping me whenever and whatever I needed any sort

of help since I joined to the tribology group. I am also grateful to Dr Yaping

and Mr. Tay Nam Beng for their advices and suggestions in the fabrication

process. I would also like to thank to Mr. Thomas Tan Bah Chee, Mr. Abdul

Khalim Bin Abdul, Mr. Ng Hong Wei and Mr. Juraimi Bin Madon from

Material Science Lab and Mr. Suhaimi Bin Daud from Micro/Nano-System

Initiative Lab for their support and assistance in many ways. My sincere

thanks also go to Dr. Su Zhoucheng, Dr. Song Shaoning, Dr. Muhammad

Ridha and Mr. Alvin from Applied Mechanics Group. Special thanks to my

colleagues, my fellow labmates and my seniors, especially to Khin Sandar Tun

who helped me for giving her suggestions.

Finally, I would like to thank my parents U Myo Myint and Daw Kywe

Kywe and three elder sisters for their support, understanding and

Acknowledgement

iii

encouragement for taking care of me throughout my life. This research

investigation work would not be completed without their love, and kindness.

Publication

iv

Journal Article

Sandar, M. M., Minn, M., Yaping, R., Satyanarayana, N., Sinha, S. K.

and Bhatia, C. S. (2013) ‘Friction and wear durability studies on the 3D

fingerprint and honeycomb textured SU-8 surfaces’, Tribology International,

60, 187-197.

Conference Publication/ Presentation

Sandar Myo Myint, Myo Minn, N. Satyanarayana, Vincent Beng Chye

Tan, Sujeet K. Sinha and Charanjit S. Bhatia, (2014) ‘Effects of 3D

negative and positive fingerprint Si/SU-8 textures on the tribological

properties’, TS1914689, ASIATRIB-2014, 17th

– 20th

February 2014, Agra,

India.

Conference poster presentation

Sandar Myo Myint, Myo Minn, Sujeet K. Sinha and Charanjit S. Bhatia,

‘Tribological properties of the negative fingerprint patterning on SU-8

polymer’, ICYRAM12-A-0685, International Conference of Young

Researchers on Advanced Materials, 1st July 2012, Singapore.

Table of Contents

v

Table of Contents

Declaration i

Acknowledgement ii

Table of Contents v

Summary xv

List of Tables xvii

List of Figures xx

List of Abbreviation xxvii

List of Symbols xxix

Chapter 1 Introduction 1

1.1.Introduction to tribology 2

1.2.Effect of surface texture in tribology 5

1.3.Application of surface texture 7

1.4.Glossary and choice of surface textures 8

1.5.Choice of polymers 9

1.6.Research objectives 11

1.7.Research methodology 12

1.8.Thesis content layout 14

Chapter 2 Literature Review 16

2.1. General properties of polymers 17

2.2. Surface textures on polymers 19

2.2.1. Patterning of micro/nano structures in

tribology 21

2.2.2. Fabrication methods of surface texture 24

2.2.2.1. Polymer jet printer 25

Table of Contents

vi

2.3. Tribology of polymers 26

2.3.1. Tribology of bulk polymer 27

2.3.2. Tribology of polymer thin film 28

2.3.2.1. Polymer film coating 30

2.4. Importance mechanisms in tribology 31

2.4.1. Friction mechanism 31

2.4.2. Adhesion in friction 36

2.4.3. Contact area and contact pressure 37

2.4.4. Friction force versus real contact area 40

2.4.5. Wear mechanism 42

2.4.5.1. Adhesive wear 43

2.4.5.2. Abrasive wear 45

2.4.5.3. Fatigue wear/rolling wear 46

2.5. Effect of different normal loads 47

2.6. Effect of wettability of surfaces 49

2.6.1. Wenzel model 50

2.6.2. Cassie-Baxter model 51

2.7. Numerical simulations of textured surfaces 52

2.8. Summary of research strategy 53

Chapter 3 Materials, Test Equipment, and Experimental

Methodologies 55

3.1. Materials 56

3.1.1. Silicon 56

3.1.2. SU-8 56

Table of Contents

vii

3.1.3. Polydimethylsiloxane (PDMS) 58

3.1.4. Perfluoropolyether (PFPE) 59

3.1.5. SU-8 developer 59

3.1.6. Silicon nitride (Si3N4) 60

3.2. Process flow 61

3.2.1. Process flow chart of the fabrication

process 61

3.2.1.1. Process flow chart of the

fabrication process for the 61

SU-8 textured surface

3.2.1.2. Process flow chart of the

fabrication process for the 61

PDMS textured surface

3.2.2. Process flow chart of the experimental

tribological characterization 63

3.3. Sample preparation and fabrication methods 63

3.3.1. Preparing of silicon (Si) substrates 64

3.3.2. Cleaning process for silicon substrates 64

3.3.3. Spin-coating of SU-8 on silicon substrate 65

3.3.4. The curing process for SU-8 on Si 66

substrate

3.3.4.1. Pre-baking process 66

3.3.4.2. Post-baking process 67

3.3.5. Exposing SU-8 to ultraviolet light 67

3.3.6. Preparation of the flat SU-8 surface 68

3.3.7. Preparation of SU-8 honeycomb textures

on Si/SU-8 surface 69

3.3.8. Preparation of SU-8 micro-dot textures

Table of Contents

viii

on Si/SU-8 surface 71

3.3.9. Preparation of SU-8 negative fingerprint

textures on Si/SU-8 surface 73

3.3.10. Preparation of SU-8 positive fingerprint

textures on Si/SU-8 surface 74

3.3.11. Preparation of the PDMS positive fingerprint

pattern 75

3.3.12. Preparation of the PDMS negative fingerprint

pattern 77

3.3.13. PFPE lubricant 77

3.4. Surface analysis equipment and techniques 79

3.4.1. Optical microscopy 79

3.4.2. Optical profiler 79

3.4.3. Field emission scanning electron

microscopy (FESEM) 80

3.4.4. Contact analysis techniques 82

3.4.5. Tribological characterization 83

3.5. Summary 86

Chapter 4 Fabrication Method and Tribological Analysis of the

Negative Fingerprint and Positive Fingerprint Patterns on

SU-8 (with and without nano-lubricant) 87

4.1. Introduction and objectives 88

4.2. Sample preparation process for the flat SU-8

surface and the SU-8 textured surfaces 89

4.2.1. Fabrication of the flat SU-8 surface 89

4.2.2. Fabrication of 3D SU-8 negative

fingerprint textures on Si/SU-8 surface 91

4.2.3. Fabrication of 3D SU-8 positive

fingerprint textures on Si/SU-8 surface 93

Table of Contents

ix

4.2.4. PFPE coating on Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP 95

4.3. Experimental procedures 95

4.4. Results and discussion 98

4.4.1. Surface morphology 98

4.4.2. Water contact angle measurements 100

4.4.3. Friction tests on Si/SU-8, Si/SU-8/NFP

and Si/SU-8/PFP 103

4.4.3.1. Analysis of the effect of contact

area versus the spatial density of the

textures on friction 104

4.4.3.2 . Analysis of wear mechanisms for

the friction tests 107

4.4.4. Wear tests on Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP (with and without PFPE) 109

4.4.4.1. Analysis of wear mechanisms for the

wear tests 111

4.4.5. Effects of different normal loads on friction

for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP 115

4.5. Summary 118

Chapter 5 Fabrication Methods and Tribological Analysis of the

Micro-dot Pattern and the Honeycomb Pattern on

Si/SU-8 (with and without PFPE nano-lubricant) 121


5.2. Sample preparations 123

5.2.1. Fabrication of 3D SU-8 micro-dot textures

on Si/SU-8 surfaces 123

5.2.2. Fabrication of 3D honeycomb patterns

On Si/SU-8 surfaces 125

Table of Contents

x

5.2.3. PFPE coating on 3D SU-8 micro-dot and

honeycomb patterns 129




5.4.2. Water contact angle measurements 133

5.4.3. Friction tests on Si/SU-8/MD with

different pitch lengths 136

5.4.3.1. Analysis of contact area versus

the spatial density of textures on

friction 138

5.4.3.2. Analysis of water mechanisms

for the friction tests 142

5.4.4. Wear tests on Si/SU-8/HC and Si/SU-8/MD

with different pitch lengths 144

5.4.5. Effects of different normal loads on

friction for Si/SU-8/HC and

Si/SU-8/MD (400 μm) 149

5.5. Summary 152

Chapter 6 The Effect of Different Size Counterface Balls on SU-8

Textured Surfaces 155


6.2. Materials and sample preparation 158

6.3. Experimental procedure 159


6.4.1. Effects of different size silicon nitride

counterface balls on friction for Si/SU-8,

Si/SU-8/NFP, and Si/SU-8.PFP 161

6.4.1.1. Analysis of the dependence

of coefficient of friction on contact

Table of Contents

xi

area with spatial texture density 164

6.4.2. Wear tests with different size counterface

balls on Si/SU-8, Si/SU-8/NFP and

Si/SU-8/PFP under dry conditions

(without PFPE nano-lubricant) 169

6.4.2.1. Analysis of wear mechanisms

under dry conditions with different

size counterface balls 173

6.4.3. Wear tests with different size counterface

balls on Si/SU-8, Si/SU-8/NFP and

Si/SU-8/PFP in lubricated conditions 176

6.4.3.1. Analysis of wear mechanisms in

lubricated conditions under

different size counterface balls 176

6.5. Summary 181

Chapter 7 Fabrication Methods and Tribological Analysis of

Positive and Negative Patterns on PDMS (with and

without PFPE nano-lubricant) 185


7.2. Sample preparation of the PDMS flat surface

and the PDMS textured surfaces 188

7.2.1. Fabrication process for the 3D PDMS

positive fingerprint textures 188

7.2.2. Fabrication process for 3D PDMS

negative fingerprint textures 189

7.2.3. Fabrication process for the PDMS flat

surface 190

7.2.4. PFPE coating on the PDMS flat surface,

3D PDMS positive fingerprint patterns

and 3D PDMS negative fingerprint patterns 190


7.3.1. Surface characterizations 192

Table of Contents

xii

7.3.2. Tribological characterizations 192

7.3.3. Different normal loads tests on the PDMS

textured surfaces 194



7.4.2. Surface analysis 194

7.4.3. Friction behaviors of the PDMS flat

surface and the PDMS textured surfaces 198

7.4.4. Wear behavior on the PDMS flat surface

and the PDMS textured surfaces under

dry and lubricated conditions 203

7.4.4.1. Wear behavior under dry

conditions 203

7.4.4.2. Wear behavior under lubricated

conditions 207

7.4.5. Effects of different normal loads 211

7.5. Summary 214

Chapter 8 Contact Area Analysis of the Textured Surfaces 217


8.2. Calculation of the contact radius (a) and the

area of contact (A) on Si/SU-8 218

8.2.1. Analysis of the contact radius (a) and the

area of contact (A) on the negative

fingerprint pattern of Si/SU-8 221


area of contact (A) on the positive

fingerprint pattern of Si/SU-8 221


area of contact (A) on the honeycomb

pattern of Si/SU-8 224

Table of Contents

xiii


area of contact (A) on the micro-dot

pattern of Si/SU-8 225

8.3. Calculation of the contact radius (a) and the

area of contact (A) on PDMS 228


area of contact (A) on the PDMS flat

surface 228


area of contact (A) on the PDMS positive

fingerprint surface 229


area of contact (A) on the PDMS negative

fingerprint surface 231

8.4. Different counterface effects on the contact

Radius (a) and the area of contact (A) on

Si/SU-8/NFP and Si/SU-8/PFP 234

Chapter 9 Numerical Simulations for the SU-8 Textured

Surfaces 236


9.2. Finite element method 240

9.3. Modeling procedures 242

9.3.1. Geometry of 3D fingerprint model 243

9.3.2. Material properties 244

9.3.3. Assembly and contact interaction

between the fingerprint pattern and

the slider 244

9.3.4. Loading and boundary condition 245

9.3.5. Meshing process of models 246

9.4. Results 247

9.4.1. Analysis of coefficient of friction on

Table of Contents

xiv

the fingerprint textured surface 247

9.4.1.1. Sliding up the step/the ridge 248

9.4.1.2. Sliding down the step/the ridge 248

9.4.1.3. Sliding on the top surface 249

9.5. Discussion 250

9.6. Conclusion of the numerical simulation 251

Chapter 10 Conclusions and Future Recommendations 253

10.1. Conclusions 253

10.2. Future work 257

References 259

Summary

xv

Summary

Polymer coating is a vital role to control the friction and wear

behaviors that are more critical in the content of Micro-Electro-Mechanical

Systems (MEMS) devices where silicon (Si) is widely used as a structural

material. However, due to the poor tribological and mechanical properties of

Si, SU-8 and polydimethylsiloxane (PDMS) are attractive for the development

of MEMS/bio-MEMS devices due to their excellent stability. Recently,

surface texturing is a great momentum for interfaces to improve surface

tribological issues on SU-8 and PDMS.

This research project focuses on the tribological characteristics of

substrates with patterned surfaces, namely, fingerprint patterns that have

rounded corners on SU-8 and polydimethylsiloxane (PDMS). The

protuberances (positive asperity) on the surface can lead to a significant

reduction in friction and wear when compared to flat featureless polymer

surfaces. Different surface patterns with positive asperities (fingerprint and

micro-dot patterns) and negative asperities (honeycomb pattern) were also

investigated for their tribological behaviors. Among the different surface

patterns on SU-8, the negative fingerprint pattern exhibited the best

tribological behaviors. The friction and wear behaviors were affected by

changes in contact area and spatial texture density that is related to multiple

asperities. Aiming to further reduce friction and to enhance wear durability on

the SU-8 textured surfaces, perfluoropolyether (PFPE) was overcoated and it

served as an effective protective layer on the SU-8 textured surfaces. To study

the effects of the contact area, tribological tests using different size

Summary

xvi

counterface balls and different normal loads were conducted on the textured

surfaces. Numerical simulations using the finite element method were also

performed in an attempt to understand the improvement in wear resistance.

Extensive experiments were conducted on PDMS, a silicone

elastomeric material, by developing the fingerprint pattern. It was also

observed that the spatial texture density of fingerprint pattern had significant

effect on the tribological behaviors of PDMS. However, PFPE nano-lubricant

was not suitable to coat on PDMS because fluorocarbons did not swell PDMS.

List of Tables

xvii

List of Tables

Table 2.1. Advantages and disadvantages of polymers over

metals 18

Table 2.2. The effect of normal load on the coefficient of friction 49

Table 3.1. The physical properties of Grade 2050 and Grade

2000.5 of SU-8 57

Table 3.2. The physical properties of Sylgard 184 silicone

elastomer 59

Table 3.3. The physical properties of PFPE (Fombin Zdol 4000) 60

Table 3.4. Development times for SU-8 developer 60

Table 4.1. Sample nomenclature of the tested surfaces 100

Table 4.2. The measured dimensions of the tested surfaces

(Si, Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP) 100

Table 4.3. The measured contact angles (WCA) of the tested

surfaces with and without PFEP 100

Table 4.4. The dimensions and the spatial texture density of

Si/SU-8/NFP and Si/SU-8/PFP 103

Table 4.5. The coefficients of friction and contact area (contact

area is calculated using Hertz’s equation by

considering a single asperity contact and the static

condition) for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP

under a normal load of 100 mN 118




condition) versus the applied normal load for Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP. 121

Table 5.1. The measured dimensions of the tested surfaces

(Si/SU-8/MD and Si/SU-8/HC) 132

Table 5.2. Sample nomenclature and the measured water contact

angles (WCA) for the SU-8 honeycomb patterns and

the SU-8 micro-dot patterns with different pitch length 135

Table 5.3. The dimensions, spatial texture densities, coefficient of

frition (COF), and calculated contact area based on the

Hertz contact equation for the tested specimens under

a normal load of 100 mN 140




condition) versus the applied normal loads for Si/SU-8/HC

and Si/SU-8/MD (400 µm) 152

Table 6.1. The coefficients of friction versus the spatial texture

density of the tested specimens (Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP) 168

Table 6.2. The calculated contact area on the tested specimens

(Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP) 169

Table 7.1. Sample nomenclature and the measured water contact

List of Tables

xviii

angles (WCA) for the PDMS flat surface and the

PDMS textured surfaces with and without PFPE 196

Table 7.2. The measured dimensions (width, length, and depth),

spatial density, and calculated contact area that were

based on a single asperity using the Hertz model for

PDMS, PDMS/PFP, and PDMS/NFP 200

Table 7.3. The calculated contact area that were based on a single

asperity using the Hertzian model for PDMS, PDMS/PFP,

and PDMS/NFP 200

Table 8.1. The properties of SU-8 material, PDMS material, and the

Si3N4 counterface ball 220

Table 8.2. The radius of contact and area of contact for single

asperity contact under the static condition using

Hertz’s equation versus the applied normal loads for

Si/SU-8/NFP 221




Si/SU-8/PFP 223




Si/SU-8/HC 225




Si/SU-8/MD (400 µm) 226


asperity contact at the static condition using Hertz’s

equation versus the applied normal loads for PDMS 229

Table 8.7. The radius of contact and the area of contact for single


Hertz’s equation versus the applied normal loads

for PDMS/PFP 230

Table 8.8. The radius of contact and the area of contact for single


Hertz’s equation versus the applied normal loads

for PDMS/NFP 233



Hertz’s equation versus the applied normal loads with

different counterface balls (2 mm, 4 mm, and 6 mm)

for Si/SU-8/NFP 235



Hertz’s equation versus the applied normal loads with

different counterface balls (2 mm, 4 mm, and 6 mm)

List of Tables

xix

for Si/SU-8/PFP 235

Table 9.1. Mechanical properties of SU-8 and Si3N4 244

List of figures

xx

List of Figures

Fig. 1.1. The profile view of (a) a positive asperity and (b) a

negative asperity 9

Fig. 1.2. The research methodology 12

Fig. 2.1. Different surface textures: (a) directional vs. non-

directional, (b) smooth vs. rough, (c) coarse vs. fine,

and (d) regular vs. irregular 20

Fig. 2.2. The shapes of features and classification of surface

textures (a) discrete textures and (b) continuous textures 20

Fig. 2.3. General classifications of surface textures 21

Fig. 2.4. Different surface topography of positive asperity and

negative asperity: (a) lotus leaf (b) circular pillar

(c) square pillar (d) honeycomb and (e) groove 23

Fig. 2.5. Various fabrication methods for micro/nano surface

textures 25

Fig. 2.6. A single channel of micro-jet printing technology (Drop-

on-Demand mode) with 50 µm diameter 26

Fig. 2.7. Contact configurations for (a) ball-on-disc configuration

and (b)flat-on-disc configuration 39

Fig. 2.8. The contact radius )(a and stress distribution and on the

contact surface occurs whenever a normal load )( nF

is applied on the contact region based on Hertzian contact 40

Fig. 2.9. Friction force versus real contact area 42

Fig. 2.10. Classifications of wear of polymers 43

Fig. 2.11. Schematic diagram of adhesive wear 44

Fig. 2.12. A schematic diagram of abrasive wear 46

Fig. 2.13. Outcomes of fatigue/ rolling wear 47

Fig. 2.14. The contact angle between a liquid and a solid

surface for (a) hydrophilic and (b) hydrophobic surfaces 50

Fig. 2.15. A schematic illustration of the homogeneous (Wenzel)

wetting 51

Fig. 2.16. A schematic illustration of the heterogeneous (Cassie-

Baxter) wetting 52

Fig. 3.1. The molecular structure of an EPON SU-8 epoxy resin

material 57

Fig. 3.2. The chemical formula of PDMS 58

Fig. 3.3. Process flow chart of the fabrication methods for the

SU-8 textured surfaces 62

Fig. 3.4. Process flow chart of the fabrication methods for the

PDMS textured surfaces 63

Fig. 3.5. Process flow chart of the tribological characterization

of the SU-8 textured specimens 64

Fig. 3.6. Preparation of a silicon sample 64

Fig. 3.7. Harrick Plasma – PDC32G macine 65

List of figures

xxi

Fig. 3.8. The process of spin-coating SU-8 on a silicon substrate 66

Fig. 3.9. The relationship between the thickness of SU-8 Grade 2050

thin film and the spin speed 66

Fig. 3.10. Hot plate used for pre-baking and post-baking processes 67

Fig. 3.11. Mask aligner machine for UV light treatment 68

Fig. 3.12. The fabrication process for the flat SU-8 surface 69

Fig. 3.13. The fabrication process of the SU-8 honeycomb pattern 70

Fig. 3.14. (a) Polymer jet printer and (b) the fabrication process of

the desired SU-8 micro-dot patterns using a drop-on-

demand device 72

Fig. 3.15. The fabrication process for the desired SU-8 negative

fingerprint pattern on Si/SU-8 74

Fig. 3.16. The fabrication process of the desired SU-8 positive

fingerprint pattern 75

Fig. 3.17. The fabrication process for the desired PDMS positive


Fig. 3.18. The fabrication process for the desired PDMS negative


Fig. 3.19. PFPE lubricant on the tested specimen using the dip-

coating method 78

Fig. 3.20. The optical microscope 79

Fig. 3.21. A Wyko NT1100 optical profiler 81

Fig. 3.22. A field emission scanning electron microscope FESEM

(Hitachi S4300) 81

Fig. 3.23. A schematic of a liquid drop showing the quantities in

Young’s equation 82

Fig. 3.24. A VCA optima contact angle system 83

Fig. 3.25. A custom-built ball on disc tribometer 85

Fig. 3.26. A schematic diagram of a custom-built ball on disc

tribometer 85

Fig. 4.1. Fabrication procedure for the flat SU-8 surface(Si/SU-8) 91

Fig. 4.2. Fabrication procedure for the 3D SU-8 negative

fingerprint structure on Si/SU-8 (Si/SU-8/NFP) 93

Fig. 4.3. Fabrication procedures for the 3D SU-8 positive

fingerprint pattern on Si/SU-8 (Si/SU-8/PFP) 94

Fig. 4.4. The coated layer of PFPE for (a) Si/SU-8/PFPE,

(b) Si/SU-8/NFP/PFPE, and (c) Si/SU-8/PFP/PFPE 96

Fig. 4.5. An optical image and a 3D image for (a) Si/SU-8/NFP

(Ra = 814 nm) and (b) Si/SU-8/PFP (Ra = 599 nm) 99

Fig. 4.6. The schematic diagrams (not to scale) of

(a) Si/SU-8/NFP and (b) Si/SU-8/PFP 99

Fig. 4.7. Graph of the coefficients of friction for Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP in dry sliding tests

under a normal load of 100 mN and a rotational speed

of 2 rpm 104

List of figures

xxii

Fig. 4.8. The optical images of debris on the counterface ball and

wear track on the samples for (a) Si/SU-8,

(b) Si/SU-8/NFP, and (c) Si/SU-8/PFP 108

Fig. 4.9. The coefficients of friction for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP with and without PFPE nano-lubricant


of 500 rpm 111

Fig. 4.10. The wear life cycles of different specimens (Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP) with and without PFPE

nano-lubricant under a normal load of 100 mN and a

rotational speed of 500 rpm 112

Fig. 4.11. The optical images of wear debris on the Si3N4 counter-

face ball and wear track on the samples of Si/SU-8

without PFPE (300 cycles), (b) Si/SU-8 with PFPE

(5,000 cycles), (c) Si/SU-8/NFP without PFPE

(2,000 cycles), (d) Si/SU-8/NFP with PFPE (60,000 cycles),

(e) Si/SU-8/PFP without PFPE (1,800 cycles), and

(f) Si/SU-8/PFP with PFPE (30,000 cycles) 114

Fig. 4.12. The resulting coefficients of friction versus the

different normal loads for Si/SU-8, Si/SU-8/NFP, and

Si/SU-8/PFP 118

Fig. 5.1. Fabrication procedures for 3D micro-dot patterns on

Si/SU-8 surfaces (using PJP machine) 127

Fig. 5.2. Fabrication procedure for 3D honeycomb patterns on

Si/SU- surface 128

Fig. 5.3. The coated layer of PFPE for (a) Si/SU-8/MD/PFPE

and (b) Si/SU-8/HC/PFPE 129

Fig. 5.4. An optical image and a 3D image for (a) Si/SU-8/HC

(Ra = 28 µm) and (b) Si/SU-8/MD (Ra = 92 nm) 131

Fig. 5.5. Schematic diagrams (not to scale) of (a) Si/SU-8/HC

and (b) Si/SU-8/MD 132

Fig. 5.6. The summarized results of the coefficients of friction

for the SU-8 micro-dot patterns with different pitch

lengths (200 µm, 300 µm, 400 µm, 500 µm, and 600 µm)

and the SU-8 honeycomb pattern in dry sliding tests


of 2 rpm 141

Fig. 5.7. The coefficients of friction and the spatial texture

densities for the honeycomb patterns and the micro-dot

patterns with different pitch lengths 141

Fig. 5.8. The optical images of debris on the counterface ball

and wear track on Si/SU-8/MD (200 µm), Si/SU-8/MD

(300 µm), Si/SU-8/MD(400 µm), Si/SU-8/MD

(500 µm), Si/SU-8/MD(600 µm), and Si/SU-8/HC

(from FESEM) in dry sliding tests under a normal load

of 100 mN and a rotational speed of 2 rpm 144

List of figures

xxiii

Fig. 5.9. The coefficients of friction for Si/SU-8/MD (200 µm),

Si/SU-8/MD (300 µm), Si/SU-8/MD (400 µm),

Si/SU-8/MD (500 µm), Si/SU-8/MD (600 µm), and

Si/SU-8/HC with and without PFPE nano-lubricant

under a normal load of 100 mN and a rotational

speed of 500 rpm 147

Fig. 5.10. The wear life cycles for different specimens

(Si/SU-8/MD (200 µm), Si/SU-8/MD (300 µm),

Si/SU-8/MD (400 µm), Si/SU-8/MD (500 µm),

Si/SU-8/MD (600 µm), and Si/SU-8/HC) with and

without PFPE nano-lubricant under a normal load


Fig. 5.11. The optical images of wear debris on the counterface

ball and wear track on the tested samples with and

without PFPE nano-lubricant under a normal load of

100 mN and a high rotational speed of 500 rpm 149

Fig. 5.12. The resulting coefficient of friction versus the

different normal loads for Si/SU-8/HC and

Si/SU-8/MD (400 µm) 151

Fig. 6.1. The coefficients of friction versus different counterface

balls (2 mm, 4 mm, and 6 mm) on the flat SU-8 surface

(Si/SU-8) under a normal load of 100 mN and a rotat-

ional speed of 2 rpm 162


balls on the SU-8 negative fingerprint pattern

(Si/SU-8/NFP) under a normal load of 100 mN and a



balls on the SU-8 positive fingerprint pattern

(Si/SU-8/PFP) under a normal load of 100 mN and a


Fig. 6.4. The summarized results of coefficients of friction

versus different counterface balls on Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP under a normal load of

100 mN and a rotational speed of 2 rpm 168


balls such as 2 mm, 4 mm, and 6 mm for Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP under dry conditions

(without PFPE nano-lubricant) with a normal load of

100 mN and a rotational speed of 500 rpm 172

Fig. 6.6. The wear durability versus different counterface balls

such as 2 mm, 4 mm, and 6 mm for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP under a dry condition with a normal

load of 100 mN and a rotational speed of 500 rpm 173

Fig. 6.7. The optical images of debris on the counterface balls

and wear track on the samples for (a) Si/SU-8,

(b) Si/SU-8/NFP, and (c) Si/SU-8/PFP under 2 mm,

List of figures

xxiv

4 mm, and 6 mm counterface balls 175


balls such as 2 mm, 4 mm, and 6 mm for Si/SU-8/PFPE,

Si/SU-8/NFP/PFPE, and Si/SU-8/PFP/PFPE under

lubricated conditions (with PFPE nano-lubricant) with a

normal load of 100 mN and a rotational speed of 500 rpm 176

Fig. 6.9. The wear durability versus different counterface balls

such as 2 mm, 4 mm, and 6 mm for Si/SU-8/PFPE,

Si/SU-8/NFP/PFPE, and Si/SU-8/PFP/PFPE in

lubricated conditions with a normal load of 100 mN

and a rotational speed of 500 rpm 178

Fig. 6.10. The optical images of debris on the counterface ball

and wear track on the samples for (a) Si/SU-8/PFPE,

(b) Si/SU-8/NFP/PFPE, and (c) Si/SU-8/PFP/PFPE

under 2 mm, 4 mm, and 6 mm counterface balls 180

Fig. 7.1. Fabrication procedures for (a) 3D PDMS positive

fingerprint structure and (b) 3D PDMS negative

fingerprint structure 191

Fig. 7.2. The PFPE coating layer (not to scale) of (a) PDMS/PFPE,

(b) PDMS/NFP/PFPE, and (c) PDMS/PFP/PFPE 192

Fig. 7.3. An optical image and a schematic diagram with full

dimensions (not to scale) of (a) the positive fingerprint

pattern on PDMS (PDMS/PFP) and (b) the negative

fingerprint pattern on PDMS (PDMS/NFP) 195

Fig. 7.4. Optical images of the water contact angles for

PDMS, PDMS/PFP, and PDMS/NFP 196

Fig. 7.5. The water contact angles and surface energies of the

PDMS flat surface and the PDMS negative fingerprint

and positive fingerprint patterns 198

Fig. 7.6. The resulting coefficients of friction for the tested

specimens (PDMS, PDMS/NFP, and PDMS/PFP) 201

Fig. 7.7. The coefficients of friction versus the number of cycles

for PDMS, PDMS/PFP, and PDMS/NFP under a normal

load of 100 mN and a rotational speed of 2 rpm. Distinct

stick-slip behaviors were observed for PDMS and

PDMS/NFP 202

Fig. 7.8. The dimensions (µm) of wear tracks for the tested

specimens (PDMS, PDMS/PFP, and PDMS/NFP) for

rotational tests of 50,000 cycles, 100,000 cycles, and

200,000 cycles.The experimental tests were carried out


of 500 rpm 205

Fig. 7.9. Optical images of wear tracks on the tested specimens

(PDMS, PDMS/PFP, and PDMS/NFP) and debris on the

silicon nitride counterface ball after 50,000 cycles,

100,000 cycles, and 200,000 cycles.The experimental

tests were carried out under a normal load of 100 mN

List of figures

xxv


Fig. 7.10. Optical images of wear tracks on the tested specimens

with PFPE (PDMS/PFPE, PDMS/PFP/PFPE ,

and PDMS/NFP/PFPE) and debris on the silicon

nitride counterface ball after 50,000 cycles,

100,000 cycles, and 200,000 cycles. The experimental

tests were carried out under a normal load of 100 mN


Fig. 7.11. Dimensions (µm) of wear tracks for the tested specimens

with PFPE (PDMS/PFPE, PDMS/PFP/PFPE,

and PDMS/NFP/PFPE) for the rotational tests of 50,000

cycles, 100,000 cycles, and 200,000 cycles.The

experimental tests were carried out under a normal load


Fig. 7.12. XPS wide spectrum for (a) PDMS/PFPE,

(b) PDMS/NFP/PFPE, and (c) PDMS/PFP/PFPE 211

Fig 7.13. The coefficients of friction for the negative and

positive fingerprint patterns on PDMS versus the

applied normal loads (50 mN, 100 mN, 150 mN,

300 mN, and 400 mN) at a constant speed of 2 rpm.

The measurements were taken at 20 sliding cycles. 214

Fig. 8.1. A schematic presentation of single asperity contact

between the negative fingerprint pattern and the

counterface ball (not to scale). The radius of curvature

was obtained by fitting a circle to the ridge geometry. 220

Fig. 8.2. A schematic presentation of a single asperity contact

between the positive fingerprint pattern and the counter-

face ball (not to scale). The radius of curvature was

obtained by fitting a circle to the ridge geometry. 223


between the honeycomb pattern and the counterface

ball (not to scale) 224

Fig. 8.4. A schematic presentation of a single asperity contact

between Si/SU-8/MD (400µm) and the counterface

ball (not to scale) 227


between the PDMS flat surface and the counterface

ball (not to scale). 228


between the PDMS positive fingerprint pattern

and the counterface ball (not to scale). 231


between the PDMS negative fingerprint pattern

and the counterface ball (not to scale). 233

Fig. 9.1. Contact problem geometry 243

Fig. 9.2. Three dimensional fingerprint patterns on SU-8 244

Fig. 9.3. Three dimensional discrete rigid sphere 244

List of figures

xxvi

Fig. 9.4. The assembly and contact behavior between the

fingerprint pattern and the indenter 245

Fig. 9.5. Three dimensional finite element mesh for the finger-

print textured pattern and the slider 246

Fig. 9.6. The coefficient of friction and stress concentration

curves for the condition of sliding up the ridge 249

Fig. 9.7. The coefficient of friction and stress concentration curve

for the conditon of sliding down the ridge 249

Fig. 9.8. The coefficient of friction curve for the condition on

the top surface of finger ridge 250

List of Abbreviations

xxvii


3D Three dimension

DI Deionized

DOD Drop-on-demand

EDS Energy dispersive spectrometer

FEM Finite element method

FESEM Field emission scanning electron microscopy

FP Fingerprint

GBL Gamma-butyrolactone

HC Honeycomb

HAR High aspect ratio

IC Integrated circuits

LOC Lab-on-chip

LST Laser surface texture

MD Micro-dot

MEMS Micro-Electro-Mechanical System

NFP Negative fingerprint

PDMS Polydimethylsiloxne


xxviii

PEB Post exposure baking

PFPE Perfluoropolyether

PFP Positive fingerprint

PJP Polymer jet printer

PSI Phase shifting interferometry

Si Silicon

Si3N4 Silicon nitride

SWCNTs Single-walled carbon nanotubes

UV Ultraviolet

UHMWPE Ultra-high-molecular weight polyethylene

VSI Vertical scanning interferometry

List of Symbols

xxix

List of Symbols

a Contact radius

rA Real contact area

e Sliding velocity

E Young Modulus

F Friction force

ndeformatioF Mechanical deformation

adhesiveF Adhesive interaction force

sF Intermolecular adhesive force

nF Applied external normal load

P Contact pressure

mP Mean pressure

Q Volume of wear debris

maxQ Maximum tangential force

R Radius of curvature

maxP Maximum pressure

S Ultimate tensile stress

V Wear volume

List of Symbols

xxx

W Actual contact load

Shear stress

o Initial shear stress

Coefficient of friction

Contact angle

Poisson Ratio

Chapter 1

1

Chapter 1

Introduction

This chapter introduces the concepts of “Tribology”, the general

concepts of surface texture, the effect of surface texture in tribology, the

choice of polymer and concludes with the research objectives and

methodology.

Chapter 1

2

1.1. Introduction to tribology

Tribology is defined as the science and technology of interacting

surfaces in relative motion and of related subjects and particles. It is derived

from the Greek word “Τριβοσ” (“TRIBOS”: meaning “rubbing” or “to rub”)

and “λογοσ” (“logos”: meaning “principle” or “logic”): so that “tribology”

would be the science of rubbing. Jost [1966] has also pointed out that the

meaning of tribology is the study of friction, wear and lubrication of

interacting surfaces in relative motion. Tribology is multidisciplinary in nature,

and it involves materials science, mechanical engineering, surface technology,

coatings and the chemistry of lubricants and additives. The study of friction,

wear and lubrication are of great significance in industrial application.

Industries of today need to control friction and improve wear durability of

products. Good tribological behaviors of products minimize consumption of

energy and resources and strengthen the competitiveness of the manufacturing

industries. Controlling friction and wear debris can improve safety, energy

efficiency, the economic viability and technical feasibility of a product. Since

the report by Jost [1990], economic savings of between 1.3% and 1.6% of the

GNP (Gross National Product) were obtained due to the proper attention to

tribology.

One great interest in many applications is the friction force that can

occur between two sliding surfaces. At first, an attractive adhesive force

persists between two surfaces when these surfaces are loaded together, and

then the attraction force between two surfaces can be high due to the

formation of high surface tension or surface energy. Friction occurs when two

Chapter 1

3

surfaces in contact start moving relative to each other. Shear stresses also arise

during sliding. The adhesive and shear forces can vary depending on the

attraction and contact behaviors between the two surfaces. Friction exists in

many applications of our daily lives and our environment. Depending on the

applications, high friction is needed for a particular purpose as in vehicle tires,

brakes and frictional power transmission systems. However, the reduction of

friction is normally desired in some applications to improve component

durability and system reliability. There are various ways of investigation in

tribology viz. coating, lubrication and surface roughness/textures in order to

control friction on the surface.

Surface coating applied onto a substrate is one of the main ways of

reducing friction. In recent years, soft polymer coatings have emerged as

durable wear protective layers and polymeric coatings have been widely used

in many applications such as buildings, bridges, automobiles, electronic

equipment, MEMS/NEMS (micro/nano electo-mechanical systems) and

BioMEMS/BioNEMS devices for both functional and aesthetic purposes due

to their ease of fabrication, cost effectiveness, low friction, good corrosion

resistance, impact absorption capabilities and highly tailored production

[Nicholaos and Polycarpou 2008]. However, the challenges of surface

interactions, debonding and fracture of the coating layer are still areas of

research in biotechnology and nanotechnology. Therefore, alternative

approaches should be investigated to control friction.

Chapter 1

4

Surface modification is another effective mean of reducing friction and

wears [Burton et al. 2005, Singh et al. 2007 and 2011 and Tay et al. 2011].

Many different surface patterns/textures can be observed in our daily

environment. The foot of geckos is one good example of adhesive systems

[Irschick et al. 1996]. Another well-known example is the patterned skin of

some fish that can reduce the drag forces when they are swimming. In the

magnetic disk industry, laser textured surfaces are well established for

reducing stiction and wear [Tam et al. 1989, Ito et al. 1990 and Aronoff et al.

1992]. Nowadays, it is common to add a controlled texture to one surface to

introduce many positive effects in tribology. It was found in earlier research

investigations that surface textures can reduce friction, wear debris and the

consumption of lubricant.

Surface asperity and surface roughness can reduce or increase contact

area and, consequently, the friction can vary by changing the slope of surface

conditions. The contact area between two surfaces is a critical parameter

determining the frictional behaviors of materials. Generally, high adhesion and

friction lead to an increase in wear of materials. Reducing adhesion and

friction can reduce wear and debris particles generated on the surfaces.

Lubrication is one of the solutions to reduce high friction and wear.

In this research investigation, the effects of surface textures

(fingerprint, micro-dot, and honeycomb) of polymer film on friction and wear

are presented. Different fabrication techniques are employed on the substrate

in order to impart different micro-scale textures. In addition, a nano-lubricated

Chapter 1

5

layer of perfluoropolyether (PFPE) is overcoated on the textured surfaces to

examine the effect of nano-lubricant on friction and wear of these textured

surfaces. Experiments on and analytical solutions of surface texture on friction

are a major part of the main research.

1.2. Effect of surface texture in tribology

The improvements in tribology have been historically achieved by

design changes, selection of materials and lubricants. Surface property is the

most important factor influencing functional properties, such as friction and

wear. Product lifetimes can be changed due to changes in the behavior of

surfaces. Even with smooth surfaces, there exists micro/nanometer sized

surface roughness or surface textures on these surfaces. The

roughness/textures on the surface play an important role in determining

surface properties such as adhesion, friction and wear. In recent years, surface

texturing has been a tool of focus in the variation of friction. The expression of

surface texture is a sub-discipline of material science that deals with the

modification of surface phases.

Based on earlier research work, friction can also be increased through

surface texturing. A high friction is needed in numerous applications such as

vehicle tires, fixtures and bolted joints. A high friction can be observed when a

rough surface slides against a softer surface. Hammerstrom and Jacobson

[2008] showed that high coefficient of friction was observed with pyramid

textured surfaces of lubricated poly alfa olefin (PAO). Stainless steel (grade

304) surfaces with high friction can be fabricated using nanosecond pulse

Chapter 1

6

energy [Andrew et al. 2014]. Singh showed that low frictional properties and

excellent wear durability were achieved due to the bio-inspired polymeric

patterns on SU-8 polymeric film [Singh et al. 2011]. The presence of

protuberances, like those on lotus leaves, exhibited excellent tribological

behaviors [Yoon et al. 2006 and Singh et al. 2011 and 2012]. Kovalchenko et

al. [2005] investigated laser surface texture (LST) effects with lubricants of

different viscosities (54.8 and 124.7 cSt at 40oC) under a pin-on-disk

apparatus. The friction of laser textured surfaces was determined under

different sliding speeds (in the range of 0.015 – 0.75 m/s) and nominal contact

pressure (in the range of 0.16 – 1.6 MPa). The reduction in friction offered by

laser surface texture on friction was observed at high speeds and loads and

with high viscosity oil. Moreover, the improvements in wear lives of the

samples were eight times higher than that of the unstructured specimens due to

the presence of the micro dimples on the steel disks [Dumitru et al. 2000].

Different designed textures can be developed in a range of applications

requiring low friction and high friction. The effectiveness of surface texture

can be achieved in tribology by changing different texture parameters such as

pattern, size, density, depth and orientation. Deterministic asperities, such as

the positive (protruding) and negative (recessed) asperities, are well developed

on the surface. Protuberances on surfaces can lead to a reduction in contact

area and friction. Improving tribological behaviors in dry conditions can be

achieved by changing the contact conditions using different surface patterns.

Surface textures can also increase the load carrying capacity of the surface.

Chapter 1

7

Researchers have shown that different patterns on surfaces, such as

mirco dimples or grooves, can significantly reduce friction [Etsion et al. 1996

and 1999, Varenberg et al. 2002, Pettersson et al. 2003, Wakuda et al. 2003,

Wang et al. 2003 and Etsion 2004 and 2005]. Moreover, the improvement in

seal performance was observed due to the presence of micro-dimples on one

of the mating seal surfaces [Etsion 2004 and 2005]. The influence of surface

texture on the tribological properties is widely studied for a large number of

other applications - increasing adhesion for mechanical interlocking

[Crowninshield et al. 1989], increasing friction and gripping action, reduction

of stiction in magnetic recording [Baumgart et al. 1995] and control of

wettability [Parry et al. 1996]. One particularly relevant application of

tribological surface texturing is for MEMS devices.

1.3. Application of surface texture

There are many different surface textures in our environment and these

surface patterns have been applied in various applications. The main purpose

for surface texturing in tribology is to reduce friction and wear debris.

However, high friction is sometimes required in some applications and surface

texturing can also be used to achieve this. Some applications for which surface

texturing are well developed are electronics [Mack 1985], micro-

electromechanical systems (MEMS) and BioMEMS, automotive components

and metal forming processes.

Surface patterning and shaping for interfaces in MEMS is critical in

order to achieve durability by controlling both adhesion and friction [Tayebi et

Chapter 1

8

al. 2005]. The influence of surface texture in the cylinder bores of a

combustion engine has been reported by Willis [1985]. It has also been

reported that the stiction of head-disk interface can be reduced by introducing

discrete round bump topography using laser surface texturing [Baumgart et al.

1995 and Wang et al. 1997]. Tagawa and David [2002] observed very

significant effects on the air film damping of a slider while the micro-texture

was well developed. In polycrystalline silicon MEMS systems, a reduction in

adhesion and friction was observed due to the presence of surface roughness,

asymmetry and peakiness on these surfaces [Tayebi et al. 2005]. Surface

patterns were also applied in biomedical fields such as the attachment of

osteoprogenitor cells on Mylar film and astroglial cells on silicon [Turner et al.

2000 and Duncan et al. 2002]. In addition, surface patterns have been used in

the automotive components. The laser surface textures were used on the

mechanical seal rings in order to obtain the reduction in friction [Etsion et al.

1994, 1996, 1999, 2002 and Yu et al. 2002].

1.4. Glossary and choice of surface textures

The textures on surfaces have a vital role in the control of friction. This

section explains commonly used terms in this thesis. Fig. 1.1 shows the profile

view of the positive and negative asperities.

Positive asperities: there are in the form of protrusion on a surface.

Negative asperities: there are in the form of cavities on a surface.

Chapter 1

9

(a) Profile view of a positive asperity (b) Profile view of a negative asperity

Fig. 1.1. The profile view of (a) a positive asperity and (b) a negative asperity.

Surface textures can control interfacial properties such as

(a) Area of contact

(b) Contact pressure

(c) Stress concentration

(d) Interfacial temperatures

(e) Frictional resistance

In our research, fingerprint patterns and micro-dot patterns are used for

surfaces with positive asperities and honeycomb patterns are used for negative

asperities on polymer surfaces.

1.5. Choice of polymers

In recent years, silicon (Si) is widely used as a structural material in

consumer electronics and MEMS systems. Si has a number of desirable

attributes in the manufacture of micro-electronic integrated circuits (IC), such

as its low cost and scalability in production. However, silicon does not have

optimal tribological behaviors and cannot support high load [Liu et al. 2001,

oh

1h

Slider Slider

oh

1h

Chapter 1

10

Yoon et al. 2005 and Williams et al. 2006]. In order to improve the

tribological behaviors of Si, soft polymer wear resistant coatings have been

developed to improve durability for MEMS/NEMS applications because

tribology plays an important role in the performance of MEMS materials.

Polymers are capable of providing low friction and have low shear strength

[Tanaka 1984]. Among the different types of polymers, SU-8 and

polydimethylsiloxane (PDMS) have been adopted for tribological applications

in biological micro-devices. There is still potential to further develop these

two materials for to improve their tribological properties.

SU-8, an epoxy-based negative photoresist material, exhibits good

structural strength, thermal resistance and chemical resistance [Pan et al. 2002

and Yu et al. 2006]. SU-8 is compatible with many micro-fabrication

techniques (photolithography) and involves much less number of processing

steps than those for silicon (Si). SU-8 is commonly used as a structural

material in MEMS/BioMEMS application such as neuroprobe, and integrated

circuits (IC) because it is easy to process [Lorenz et al. 1997 and Abgrall et al.

2007]. SU-8 also possesses good mechanical and biocompatible properties and

it has sufficient compliance (Young’s modulus of ~4 GPa) for flexible designs.

Based on aqueous physiochemical tests, SU-8 leaches detectable nonvolatile

residues in isopropyl alcohol. It is also considered for implanted and drug

delivery devices.

Another common tribological material is polydimethylsiloxane

(PDMS), a silicone elastomeric material that is especially attractive for the

Chapter 1

11

development of MEMS/BioMEMS application. It is biocompatible [Belanger

et al. 2001], affordable, optically transparent [Piruska et al. 2005], non-

fluorescent and nontoxic. PDMS is chemically inert, thermally stable and

exhibits isotropic and homogeneous properties. PDMS elastomer has low

surface energy and non-wetting (hydrophobicity) properties [Chaudhury et al.

1991] and it is easy to modify different surface patterns effectively using soft

lithography and replica technique. Moreover, PDMS polymeric materials do

not need stabilizers due to their intrinsic stability. PDMS is well established in

biomedical and medical applications such as catheters, implants, valves and

micro gaskets.

1.6. Research objectives

The main objective of this doctoral research is to reduce friction

significantly and improve wear behaviors on SU-8 and PDMS with a specific

significant surface pattern under dry and lubricated sliding. This research

requires to meet through a development of the positive asperity of surface

texture with the rounded corner (fingerprint pattern), an analysis of surface

contact area and surface texture density in dry sliding, an investigation with

the help of lubricant (perfluoropolyether) on surface texture, an evaluation on

the effect of different size counterface sliding balls and an extensive study

with numerical simulation using the finite element method on the fingerprint

pattern.

Chapter 1

12

1.7. Research methodology

There are four main tasks to accomplish the above mentioned

objectives as shown in Fig. 1.2. The first part focuses on fabrication methods

of different surface patterns on Si/SU-8 and PDMS. Positive asperities

(fingerprint and micro-dot patterns) and negative asperities (honeycomb

pattern) are developed for Si/SU-8 and PDMS.

Fig. 1.2. The research methodology

The new fabrication approach undertaken as follows:

(1) Positive asperity on Si/SU-8 and PDMS

(a) Negative fingerprint patterns:

Fabrication of negative fingerprint patterns on Si/SU-8

using a subject’s finger,

Fabrication of negative fingerprint patterns on PDMS using

a PDMS positive fingerprint mould,

(b) Positive fingerprint patterns:

Fabrication of positive fingerprint patterns on Si/SU-8

using a PDMS negative fingerprint pattern as a mould,

Research Methodology

Fabrication

methods

Tribological

characterizations

Analytical

calculation based

on Hertzian

method

Analytical

simulation

Chapter 1

13

Fabrication of positive fingerprint patterns on PDMS using

a Si/SU-8 negative fingerprint mould,

(c) Micro-dot patterns:

Fabrication of micro-dot patterns on Si/SU-8 using polymer

jet printer,

(2) Negative asperity on Si/SU-8

(a) Honeycomb patterns:

Fabrication of honeycomb patterns on Si/SU-8 using a glass

honeycomb mask.

The second part involves the investigation of surface properties and

tribological behaviors of different surface textures using ball-on-disc

tribometer and these are evaluated in terms of:

(a) Water contact angle and surface characterization

(b) Coefficient of friction and wear durability under

a. Different normal loads effects, and

b. Different counterface ball effects

Different normal loads are used to evaluate load carrying capacity of the

surface textures. Then, the correlation between nano-lubrication of

perfluropolyether (PFPE) and surface textures on Si/SU-8 has inspired their

application for friction and wear.

The third part presents the analytical calculation for the contact area

and pressure on a single asperity using Hertzian method to analyze the effect

of contact behavior on friction. The fourth part includes the analytical

Chapter 1

14

investigations of coefficient of friction, and shear stress of fingerprint patterns

on Si/SU-8 using the finite element analysis with ABAQUS simulation

software. The approach can support to the experimental parts of tribology.

1.8. Thesis content layout

The next chapter, Chapter 2, presents an overview of surface texturing

and tribological characterizations. Polymeric materials and the fabrication

methods of micro-scale patterning on polymers are also introduced.

Chapter 3 describes the materials and lists of experimental apparatus

for surface characterizations and tribological characterizations. The fabrication

process flow and test process flow are described in detail.

Chapter 4 presents the fabrication methods for the negative fingerprint

and positive fingerprint patterns on Si/SU-8. The water contact angle

measurements and surface characterizations are investigated in detail. The

comparative results on the tribological characterizations between the flat SU-8

surface and SU-8 textured surfaces (negative fingerprint pattern and positive

fingerprint pattern) are studied with and without PFPE nano-lubricant.

Chapter 5 describes the fabrication methods for the micro-dot pattern

with different pitch lengths and honeycomb pattern on Si/SU-8. The results

from the tribological characterizations for both positive asperity (micro-dot

pattern) and negative asperity (honeycomb pattern) are analyzed. The nano-

lubricant layer is developed on the tested specimens to analyze for long term

wear durability.

Chapter 1

15

Chapter 6 presents the effects of different size counterface balls on the

SU-8 negative fingerprint pattern and SU-8 positive fingerprint pattern.

Chapter 7 presents the investigations on the effects of tribological

properties when the fingerprint patterns are introduced on PDMS surface. The

fabrication method for PDMS textured surfaces is also described in detail.

Chapter 8 describes the analysis of contact area on the textured

surfaces using Hertzian contact equation. The calculation Hertzian method is

based on a single asperity of surface texture.

Chapter 9 presents the stress analysis on the textured surfaces using

numerical simulation with the finite element method (ABAQUS software).

Chapter 10 contains the conclusion and results for improvement and

recommendation.

Chapter 2

16

Chapter 2

Literature Review

This chapter presents a literature review of the properties of polymers

and different structures of surface patterns. The current methods of obtaining

different patterns on polymers are involved. The main mechanisms in

tribology and the concept of tribology for bulk polymer and polymer thin films

are also presented in an attempt to understand the behaviors of tribology and

reduce overall friction and wear debris. In addition, the actual area of contact,

the main factor affecting fiction, is discussed.

Chapter 2

17

2.1. General properties of polymers

Polymers (poly – ‘many’ and mer – ‘parts’) are composed of many

repeated molecules, to form organic compounds. The molecules of polymer

are formed by combining large numbers of parts (mers). Polymers are not as

stiff and hard as metals, but not as soft as liquids. Polymers are durable and

flexible and then they can be easily molded into any shape. Basically, there are

two types of polymers such as natural polymers and synthetic polymers. Both

natural and synthetic polymers play an important role in the environment.

Natural polymers such as silk, wood, natural rubber, cotton and proteins, etc.,

have been widely used in everyday life for many centuries. Synthetic polymers,

human made polymers, are used in the applications of ball bearings, seals of

shafts, sliders, tires and in a variety of consumer products. A wide range of

synthetic materials are widely found in biomedical devices such as blood bags,

coronary stents, catheters, contact lenses and etc. Both natural and synthetic

polymers can be created with a wide range of density, stiffness, and strength

and heat resistance that may develop in various applications.

Based on the literature review, several advantages and disadvantages

of polymers over metals in tribological applications are as follows in Table 2.1.

Moreover, there are a number of challenges in improving the tribological

properties of polymers in many applications. Many factors can change the

tribological performances of polymers. The understanding of friction and wear

behaviors of polymers is critical in the selection of the appropriate material for

various industries. Recently, many tribologists have shown a strong interest in

Chapter 2

18

polymer tribology. Investigations into polymer tribology have been reported

by Bowden and Tabor [1950], Briscoe and Tabor [1978], Briscoe [1981] and

Friedrich [1986].

Table 2.1. Advantages and disadvantages of polymers over metals

Advantages Disadvantages

1. Low cost and light weight

2. Easy to fabricate and modify

surface pattern

3. High-aspect-ratio

microstructures and impact

resistance

4. Good chemical and corrosion

resistance

5. Low friction under dry

condition and high wear

durability

6. Biocompatibility

1. Low thermal resistance

2. Low thermal conductivity

3. High coefficient of thermal

expansion

4. Low stiffness

5. Poor tensile strength

Among the different types of polymers, epoxy based negative

photoresist SU-8 is one of the most well developed in the fabrication of

numerous high aspect ratio micro-electro-mechanical systems (MEMS),

biological micro-electro-mechanical systems (BioMEMS) or lab-on-chip

(LOC) devices due to its excellent mechanical properties, thermal and

chemical stability and optical transparency [Abgrall et al. 2007 and Campo et

al. 2007]. SU-8 is attractive as a functional material because it is fully cross-

linked. An epoxy based negative photoresist of SU-8 is a viscous polymer that

can be spread over the substrate with thickness ranging from below 1 µm to

above 300 mm. SU-8 photoresist can be used at high aspect ratios of above 20

with standard contact lithography techniques. SU-8 is most commonly used

with ultraviolet (UV) light radiation of 365 nm. SU-8 is highly hydrophobic

and it has a low surface energy making it suitable for permanent applications.

Chapter 2

19

Other useful tribological materials include silicone based materials that

are widely used in medical applications because of their biocompatibility. As

commercial materials, they have been available since 1946. Silicone elastomer

has been developed for biliary surgery [Lahey 1946], life-saving medical

devices like pacemakers [Curtis et al. 2004] and pharmaceutical applications.

Polydimethylsiloxane (PDMS), a group of polymeric organosilicon

compounds, is becoming more popular in a wide range of biomedical

applications. The advantages of PDMS are their low cost, durability,

flexibility and transparency. However, it has a low glass transition

temperature. Due to the low reactivity, PDMS is a stable and non-toxic

material and has high gas permeability [Lotters et al. 1997].

2.2. Surface textures on polymers

Surface texturing, one of the methods to control friction, affects

surface characteristics that are used to identify and recognize objects. Surface

texturing involves the modification of surface topography such as the creation

of regular or irregular shaped asperities. The textures and repetition of image

patterns may be directional or non-directional, smooth or rough, coarse or fine,

regular or irregular, etc as shown in Fig. 2.1. Different surface textures have

been observed in nature and it is very easy to recognize naturally textured

surfaces.

There are many theoretical studies on the geometric parameters of

textures [Etsion et al. 1996 and 1999 and Wang et al. 2005]. These parameters

affect tribological properties. Surface feature may be discrete textures (e.g.

Chapter 2

20

circle, rectangular, triangle, honeycomb and etc.) or continuous textures (e.g.

an array of straight line or curved line in parallel or crosshatch form). Fig. 2.2

shows the shapes of features and classifications of surface textures with

discrete textures and continuous textures. The different surface textures can

change the contact area and contact pressure contours inside the contact. The

general classifications of surface textures are shown in Fig. 2.3.

Fig. 2.1. Different surface textures: (a) directional vs. non-directional, (b) smooth vs.

rough, (c) coarse vs. fine, and (d) regular vs. irregular.

Fig. 2.2. The shapes of features and classification of surface textures (a) discrete

textures and (b) continuous textures [Suh et al.2010].

(a) (b)

(a) (b)

(c) (d)

Chapter 2

21

Fig. 2.3. General classifications of surface textures.

2.2.1. Patterning of micro/nano structures in tribology

Even though surfaces may seem smooth, micro/nano-meter sized

surface roughness or surface textures can exist on the surfaces. The

roughness/asperities of the texture on the surface play a vital role in

determining adhesion, friction and wear on polymer surfaces in order to

control friction and wear [Burton et al. 2005, Yoon et al. 2006, Pham 2009

and Sigh et al.2011]. Different textures on the surfaces have been tested on

magnetic hard disks [Ranjan et al. 1991, Ishihara et al. 1994and Zhou et

al.2000] and mechanical components [Etsion 2004].

The texturing of gecko feet is also of great interest in tribology because

it results in a very strong adhesion force on surfaces. The micro scale structure

of lotus leaves reduces friction and gives rise to super-hydrophobicity. Of

particular interest is the impact of positive (protruding) and negative

(recessed) asperities on friction and wear properties. These have been

developed in MEMS systems by different fabrication methods. Tribologists

have investigated a variety of surfaces with positive asperity (e.g. lotus leaf

Surface Textures

Directional

Continuous

Un-directional

Smooth

Roughness

Coarse

Fine

Regular

Continuous

Discrete

Irregular

Discrete

Chapter 2

22

and pillar structures) and negative asperity (e.g. groove and honeycomb

structures) as shown in Fig. 2.4.

Dumitru et al. [2000] showed that micro dimples on steel disks

improved the lifetime of the samples by eight times. Yu et al. [2002]

examined the effect of the laser-textured mechanical seals with silicon carbide

(SiC) for the stationary mating ring material and carbon impregnated with

resin for the rotating seal ring. They obtained a significant reduction in the

coefficient of friction by using a laser-textured seal face with micro pores

under closing forces of 68 N and 136 N. Pettersson and Jacobson [2003]

investigated the effect of two surface patterns - grooves and square dimples -

on both TiN and DLC coated silicon surfaces. Experimental tests were carried

out with 10 mm diameter of steel ball under boundary lubricated and

unlubricated sliding conditions. Under dry reciprocating sliding condition, low

friction was observed on the TiN textured samples with both groove and

square dimple patterns. However, the friction on the DLC coated textured

surface was higher and more unstable than that of the flat surface. Although

the groove texture on the DLC coated layer showed high friction, the square

dimples on the DLC coated layer was stable and low under boundary

lubrication reciprocating sliding conditions. All textures on TiN showed

unstable friction.

Wakuda et al. [2003] verified that micro-dimple ceramic texture was

an effective way to reduce friction. The micro-dimple size was approximately

100 µm at a density of 5 – 20%. In addition, Bhushan and Jung [2007]

Chapter 2

23

investigated the silicon surfaces patterned with pillars and obtained a decrease

in coefficient of friction as the pitch length between pillars increased up to 10

µm. He et al. [2008] investigated the friction of poly(dimethylsiloxane)

(PDMS) elastomeric materials with surface textures such as pillars and

grooves. They observed that the coefficient of friction on the pillar-textured

surface reduced by approximately 59% for macro-scale tests and 38% for

micro-scale tests compared to untextured surface.

Fig. 2.4. Different surface topography of positive asperity and negative asperity: (a)

lotus leaf (b) circular pillar (c) square pillar (d) honeycomb and (e) groove.

Wang et al. [2010] studied a friction reduction due to the presence of

micro-dimples on various surfaces including silicon carbide and PDMS. Tay

et al. [2011] investigated the tribological behaviors of micro-dot (108 µm

diameter) patterns of SU-8 on Si wafer. Better frictional properties were

obtained for the pitch lengths between 100 µm and 200 µm. Wear lives of SU-

8 patterned Si wafer increased to 100,000 cycles for optimized micro-dot

texture from only a few cycles for bare Si or SU-8 under the same contact

(a) (b) (c)

(d) (e)

Chapter 2

24

conditions [Tay et al. 2011]. Singh et al. [2011] investigated the friction and

wear behaviors on bio-inspired polymeric patterns and observed low friction

and excellent wear durability due to the topographical modification. It is well

known from these studies that surface patterns can offer better tribological

behaviors.

2.2.2. Fabrication methods of surface texture

Fig. 2.5 shows the various fabrication methods for micro/nano textures.

In order to prepare different textures, lithography and non-lithography

methods have evolved. The printing process which creates a lithograph

includes the masks. The surface patterns transfer from the masks onto the thin

films on the wafer substrates using either light source of UV or X-ray

photoresists. The masks are made of transparent materials such as plastic,

glass and quartz. The contact masks are used as a soft contact and hard contact.

In addition, developing surface patterns on the thin films are carried out by the

dry and/or chemical etching.

Non-lithography methods are also used to achieve surface patterns.

Contact (soft lithography) and non-contact printing methods are involved in

the printing technology of non-lithography. In the soft lithography process, the

replication of surface texture is obtained from a master piece mould of the

desired surface patterns. The mould can be obtained using lithography method

[John and Ralph 2005]. It is also possible to replicate the master piece multiple

times with the mould. Polydimethylsiloxane (PDMS) has been one of the most

actively developed polymers using soft lithography method [McDonald and

Chapter 2

25

Whitesides 2002]. Soft lithography method is simple and inexpensive and then

the fabrication process is rapid.

Fig. 2.5. Various fabrication methods for micro/nano surface textures.

The non-contact printing method is well developed. Future

development of surface texturing on polymer is obtained using jet printing.

The advantage of non-contact printing method is that it is maskless and cost

effective. The polymeric material is dispensed directly onto the substrates

through the nozzle/drop to obtain fine textures when using high resolution of

drop-on-demand jet printer.

2.2.2.1. Polymer jet printer

Ink-jet printers may operate continuously or in drop-on-demand (DOD)

mode. The DOD mode is chosen because of high accuracy of placement. In

our research study, the piezoelectric DOD printing technology is used. The

Fabrication Methods

Micro/Nano patterns

Lithography Non-lithography

Contact printing Non-contact printing

Micro-contact

printing

Micro-molding

Embossing

Ink-jet printing

Dip-pen

printing

Chapter 2

26

polymer jet printer is supplied by MicroFab Technologies Inc, Plano, TX. It is

also capable of producing and placing droplets of solders, polymer, particle-

laden fluids, and etc. This micro-jet printing method is low cost (no tooling

required), noncontact, flexible and data-driven. It is also capable of printing

directly from CAD information.

Fig. 2.6 shows a single 50 µm diameter channel of micro-jet printing

technology. In order to obtain a single droplet, a novel drive waveform

technology is used. The size of polymer drop (bump) is changed under

software control.

Fig. 2.6. A single channel of micro-jet printing technology (Drop-on-Demand mode)

with 50 µm diameter.

2.3. Tribology of polymers

Polymers are available in solid, semi-solid and liquid forms at room

temperature. They can be used as bulk solids and films. Depending on the

properties of polymer materials, the polymer surfaces can be elastic, plastic

and visco-elastic. Three basic effects on the tribology of polymers are (i)

adhesive junctions, (ii) shear strength of two rubbing contacts, and (iii) real

contact area [Bowden and Tabor 1950, Kragelskii 1982 and Myshkin et al.

1997].

Chapter 2

27

2.3.1. Tribology of bulk polymer

The tribology of bulk polymer started with rubber that in the modern

automotive industry. Rubber is an example of an elastomer which deforms

elastically under small loads. Some common characters such as high elasticity,

viscoelasticity and glass transition temperature far below room temperature

are observed in rubber. Rubber has some favorable properties such as high

wearability and oil-resistance that are lacking in metal and other polymeric

materials. Therefore, rubber is also widely used outside the automobile

industry.

Since the 1970’s, rubber has been used in vehicle tires, sealing rings,

soft limited journal bearings, water lubricated bearings etc. When rubber slides

on a hard and rough surface, the oscillating forces and the energy dissipation

occur on the surface. Persson [1998] investigated the friction force acting on a

hard cylinder and spherical ball rolling on a flat viscoelastic solid and he

observed that the adhesion of rubber played a major role. Because of its low

elastic modulus, the rubber was deformed due to the adhesion force.

Shooter and Tabor [1952] studied the tribological properties of linear

polymers such as polyvinyl chloride, polyethylene, polytetrafluoroethylene,

nylon and polymethylmethacrylate at loads ranging from tens of milligrams

(0.04 g) to kilograms (10 kg) at slow sliding speeds. Although the coefficients

of friction were high at light loads, the coefficients of friction were nearly

constant at loads above 100g. However, the frictional behavior of

polytetrafluoroethylene was different from that of the other plastics. The

Chapter 2

28

investigations on the frictional behaviors of thermoplastic polymers were

widely studied in 1960s and 1970s [Ludema and Tabor 1966, Pooley and

Tabor 1972, Tanaka and Miyata 1977]. In 1966, Ludmea and Tabor showed

the frictional properties of polymers with viscoelastic characteristics. Santner

and Czichos [1989] carried investigations on thermoplastics (e.g. HDPE, PP,

PTFE, PA6, PA 66, POM, PETP, PBTP, PI and some reinforced and

composite materials) that are suitable for practical applications. Fritzson [1990]

investigated the coefficient of friction of elastomer composite siding against a

cast iron surface. All results showed that the coefficient of friction is highly

dependent on the surface structure, pressure, temperature and sliding velocity.

2.3.2. Tribology of polymer thin film

Thin film polymer coatings have been highly investigated because they

can be used as protective layers to extend product-life span and decrease

friction. However, the load carrying capacity of soft thin polymer films is low.

In 1958, Fitzsimmons and Zisman studied the tribological performance of

PTFE (Teflon) polymer films 15.2 µm to 17.8 µm thick. It was commonly

coated on steel, brass and aluminum and it increased corrosion resistance.

Kitoh and Honda [1995] observed good tribological performance on

the polymide films that were sputtered onto silicon wafers. The film thickness

of sputtered polymide on Si surface was 20 nm. Both the coefficient of friction

and abrasion resistance of the polymide film were adequate and the abrasion

resistance of the polymide film was three to five times higher than that of

PTFE film. Liu et al. [2002] investigated the tribological behaviors of 180 nm

Chapter 2

29

polydimethylsiloxane (PDMS) films on silicon substrates. The most stable

coefficient of friction and the longest wear durability were observed on the

PDMS films on hydrophilic substrates under a 3 mm diameter of ASIS-52100

steel ball. Based on the experimental results of multi PDMS layers by Yamada

[2003], the magnitude of friction in the tests of two-layer PDMS films was six

to eight times higher than that of the films with a mobile middle layers. Then,

Satyanarayana et al. [2009] also studied the tribological behavior of a

chemisorbed UHMWPE (ultra-high-molecular-weight polyethylene) film with

a thickness of approximately 1.4 µm on Si surface. They observed low

coefficients of friction and high wear durability under a normal load of 0.3 N

and a sliding speed of 0.042 m/s.

An investigation of UHMWPE films on Si surfaces [Minn and Sinha

2009], showed that the coefficient of friction was reduced to 0.18 and the wear

life cycles were increased to 20,000 cycles after the Si surfaces were coated.

In addition, the thin layer of UHMWPE on Si surface decreased the coefficient

of friction further to 0.13 and improved the wear lives to 100,000 cycles with

the presence of an intermediate layer of DLC. The improvements of wear life

were obtained by controlling the thickness of UHMWPE top thin layer. Samad

et al. [2010] determined the tribological performances of a nanocomposite

polymer coating of ultra-high molecular weight polyethylene (UHMWPE)

reinforced with 0.1 wt% of single-walled carbon nanotubes (SWCNTs). Singh

et al. [2012] developed thin films of SU-8 (thickness ~ 1.8 µm) on Si

substrates and observed that the friction was reduced considerably. PFPE thin

films with oxygen plasma treatment on the SU-8 thin film reduced the

Chapter 2

30

coefficient of friction significantly. SU-8 spin-coated layer was fabricated on

Si surface and better tribological behaviors were observed when compared to

the Si wafer [Sandar et al. 2013]. The wear durability of Si/SU-8 was

approximately 300 cycles and the coefficient of friction of Si/SU-8 was 0.2. In

addition, the thin layer of PFPE on SU-8 increased the wear lives to 5,000

cycles and wear particles were significantly reduced on the counterface ball

and the tested samples. Saravanan et al. [2013] studied a stand-alone laminate

film; SU-8 + PFPE composite lub-tape film. Due to the presence of the lub-

tape film, the initial coefficient of friction was reduced by nearly seven times

and the wear life was increased by more than five orders when compared with

SU-8.

Based on literature review, deposition of thin polymer films is well

developed. Both polymer and polymer composite can be successfully coated

as the thin films on various substrates. The thin film can serve as a protective

layer and it is very useful for various applications. Compared to bulk polymer,

thin films can lead to better tribological behaviors - reduced coefficient of

friction and increased wear durability. The surface of thin film has been

developed with micro/nano textures to improve friction and wear behaviors of

substrates. Different techniques can be used to employ thin polymer films and

polymer composite on various substrates.

2.3.2.1. Polymer film coating

The primary function of a polymer film coating is to protect the surface

of an object from the environment. In a solution based coating process,

Chapter 2

31

polymer molecules are adsorbed from the solution and the solvent evaporates

while curing and drying. Different types of polymer coating techniques are

widely used in applications [Advincula et al. 2004]. The different techniques

are

Spray coating

Vapor deposited coating

Spin coating

Dip coating

Doctor blading

The small thickness of thin films (nano/micro meter) can be achieved

using spin coating or dip coating process with good control of sliding speeds

and dipping speeds. Depending on the coating techniques, polymer thin films

can be multilayer coatings on the substrates.

2.4. Important mechanisms in tribology

Tribological performances of polymers have been investigated since

the mid-20th

century. The mechanisms of friction and wear conditions of

polymers will be reviewed in this section.

2.4.1. Friction mechanism

Interfaces between two moving surfaces are important in tribology.

Many factors such as adhesion, friction, wear, lubrication and/or

contamination can greatly influence the interacting materials. Friction is the

resistance to relative motion when one surface slides over another surface first

Chapter 2

32

discovered by Leonardo da Vinci and presented as the first law of friction in

the middle of fifteenth century. It was concluded that the friction force is

proportional to the normal load. Then, in 1699, Guilliaume Amontons defined

the two basic ‘laws’ of friction. These two laws are denoted as ‘Amontons

Laws’ in 1699.

Amonton’s First Law: The friction force is directly proportional to the

applied normal load between the bodies.

Amonton’s Second Law: The friction force is independent of the

apparent area of contact.

In 1706, Leibnitz studied a distinction between rolling and sliding

friction. In 1748, Euler explained the most important outcome of kinetic

friction from static friction. Coulomb in 1785 introduced a third law of friction

or Coulomb’s law of friction:

Kinetic friction is nearly independent of the sliding speed and is much

lower than the static friction.

Friction is a common phenomenon in daily life. In 1942, Bowden and

Tabor described the behaviors of friction under dry sliding conditions that are

governed by two mechanisms, namely, adhesion and ploughing. The friction

force )( frictionF between a hard slider and a plane surface of the soft material

is the sum of the ploughing actions/mechanical deformation )( ndeformatioF and

the adhesive interactions )( adhesiveF . The total friction force is expressed as;

adhesiivendeformatiofriction FFF (2.1)

Chapter 2

33

Ploughing action/mechanical deformation occurs when a hard surface slides

over a soft surface. The adhesive interaction is due to the presence of

interfacial shear stress between the two surfaces of hard asperities. The

adhesion depends on the interfacial shear stress. In 1981, Tabor mentioned

three main elements that are involved in dry sliding [Tabor 1981]:

(1) The real contact area between two sliding surfaces,

(2) The strength of bonds at the contact interfaces of the two

sliding surfaces,

(3) The way the material at and around the regions of contacts are

sheared or ruptured.

The shear stress, , is directly proportional to the contact pressure, P , when

the velocity and temperature are kept constant and it is expressed as [Bowden

and Tabor 1986]

Po (2.2)

where o

and are the constants of the initial shear property and the pressure

coefficient, respectively. The shear stress at the junctions is the ratio of friction

force, frictionF , to the real contact area, rA :

r

friction

A

F (2.3)

Chapter 2

34

The ratio of the applied external normal load, n

F , and the real contact area,

rA , is defined as the contact pressure, P , i.e.

r

n

A

FP (2.4)

By substituting equation (2.3) and (2.4) into equation (2.2), we obtain equation

(2.5),

r

no

r

friction

A

F

A

F (2.5)

The initial shear stress,o

, is neglected when the applied normal load is high.

Therefore, we obtain equation (2.6):

nfriction FF (2.6)

When and is defined as the coefficient of friction, so the coefficient

of friction is the tangential force )( frictionF divided by the applied external

normal force )( nF and therefore,

n

friction

F

F (2.7)

Based on the definition of friction force by Tabor, Chang et al. [1987]

introduced the theoretical model of static coefficient of friction, static , that is

related to the shearing of the contact. The contact load is related to the true

contact area. The static coefficient of friction is as follow;

Chapter 2

35

sn

staticFW

Q

F

Q

maxmax (2.8)

where maxQ is the maximum tangential force that relates to the shearing of the

contact at the interface, and nF is the applied external normal force that

relates to the actual contact load (W ) and the intermolecular adhesion force,

( sF ). The actual contact load (W ) is related to the real contact area at the

contacting points and the deformation of the contact whether it is elastic,

elastic-plastic or fully plastic. The intermolecular adhesion force ( sF ) is

related to the strength of the bonds that is formed at the junctions of the

contacts.

Following the reasoning of Green [1955] and Halling [1976], in

particular situations more than one asperity is in contact between two surfaces,

many asperities are performing the same deformation cycle and these cycles

are independent. The coefficient of friction can be calculated based on the

ratio of the average shear force and the average applied normal force over one

cycle multiplied by the number of asperities [Tangena and Wijnhoven, 1985]

and it can be described as

m

n

m

friction

n

friction

in

iifriction

dxFm

dxFm

Fn

Fn

F

F

0

0

)(

)(

1

1

(2.9)

where is )(ifrictionF is the shear stress of the ith individual asperity, nF is the

normal force of the ith individual asperity, n is the total number of asperities,

Chapter 2

36

frictionF is the average shear force, nF is the average normal force, x is the

horizontal axis and m is the length of deformation cycle.

2.4.2. Adhesion in friction

When two surfaces are loaded against each other, contact is made at

the tips of asperities and the interatomic forces contribute to friction. At low

shear stress, the slider against the surface is unable to move when the

interatomic forces between the atoms are not overcome. The contact between

two surfaces may then deform. However, the slider can move easily if the

applied force exceeds the sum of all the interatomic forces in the region of

contact. Based on the development of the basic model of adhesion by Bowden

and Tabor, when the contact area between two surfaces is small, the contact

pressure over the asperities at the contact points is high and plastic

deformation can occur. The large deformation leads to an increase in the real

contact area. The shear stress is necessary to start and the friction force per

unit area of contact can be expressed as

rfriction AF (2.10)

where frictionF is the tangential friction force, rA is the real contact area and

is the shear stress in contact. McFarlane and Tabor [1950] described an

investigation into the adhesion and friction of steel sliding on indium in air.

Experiments showed that an increase in real contact area was observed while

the normal and tangential loading increased at the junctions. There was a

continuous increase in the friction force and adhesion. Lower real contact area

Chapter 2

37

and lower adhesion was observed for materials with higher roughness.

Bhushan [1999] mentioned that adhesion force is determined by real contact

area, the normal load, surface roughness and mechanical properties.

Johnson, Kendall and Robert [1971] developed the JKR model that is

based on Hertzian contact analysis. The contact in the JKR theory is

considered to be adhesive and correlates to the contact area and the interfacial

interaction strength of the elastic surfaces. Based on the JKR model, the

contact radius at light loads is greater than that of the Hertz’s model. The

Hertz equation, modified to take into account the surface energy, is given by

2

*

3 3634

3RRFRF

E

Ra nn (2.11)

Based on the approach of the JKR model, Derjaguin, Muller and

Toporov [1975] developed the DMT model. The contact profile remains the

same as in Hertzian contact but the adhesion force acts beyond the contact

area with additional attractive interactions. The influence of the contact

deformation and the molecular attraction such as the van der Waal’s force

increase the contact area between the spherical slider and the plane surface.

The contact area between two surfaces from DMT theory is

RFE

Ra n 4

4

3*

3 (2.12)

2.4.3. Contact area and contact pressure

An understanding of the mechanics contact of solid bodies is a main

part in tribology. The nature of deformations of and stress distributions over

Chapter 2

38

contacting surfaces affect friction and wear. Either elastic or plastic

deformation can be observed due to normal load and contact pressure. With

plastic deformation, materials eroded from the contacting surface act as wear

debris which increases the wear rate.

The contact area plays a vital role when the asperities and type of

contact configuration are considered, namely ball-on-disc (spheres) and flat-

on-flat (cylinder) configurations as shown in Fig. 2.7. When two surfaces with

microroughness are placed in contact, the stress at and deformation of the

asperities have an influence on the frictional behaviors. The real (true) contact

area is a small fraction of the apparent (nominal) area of contact. The real

contact area can change due to the surface topography, material properties and

loading condition. Fig. 2.8 shows a contact of radius )(a and the stress

distribution that occurs over the contact surface when a normal load )( nF is

applied.

For surface contact between asperities, the Hertzian contact solution

[Hertz, 1882] is often referred to for the analysis of elastic contact deformation

for circular arcs shown in Fig. 2.8. Based on Hertzian contact, a deformation

contact radius )(a is assumed when two spheres come in contact with any

given load. The width of contact zone between two elastic spheres can be

increased by increasing the applied normal load )( nF .The contact radius )(a ,

the contact area )( rA , and both the maximum pressure )( maxP and the mean

pressure )( mP can be obtained as:

Chapter 2

39

Hertz contact radius: 3

1

*

*

4

3

E

RFa n (2.13)

Hertz contact area: 3

2

*

*2

4

3

E

RFaA n

r (2.14)

where 21

*

1

1

1

RRR and

2

2

2

1

2

1

*

- 1

- 1

1

EEE

Mean contact pressure (normal pressure):

31

32

*

*

3

41 nmean F

R

EP

(2.15)

Hertz maximum contact pressure: 31

32

*

*

max 3

4

2

3

2

3 nmean F

R

EPP

(2.16)

1E and 2E are the Young’s moduli of the slider and the surface. 1 and 2

are the poisson’s ratio of the slider and the surface.

Fig. 2.7. Contact configurations for (a) ball-on-disc configuration and (b) flat-on-

disc configuration.

nF

(a)

nF

(b)

Chapter 2

40

Fig. 2.8. The contact radius and stress distribution and on the contact surface

occurs whenever a normal load is applied on the contact region based on

Hertzian contact.

2.4.4. Friction force versus real contact area

Based on the literature survey, the real contact area is the main

determinant of frictional behaviors. The friction force occurs due to the

adhesion force and the ploughing action/mechanical deformation. As the real

contact area increases, friction force due to adhesion increases asymptotically

while the friction force due to the presence of mechanical deformation

decreases. Accordingly, the overall friction force can be represented as shown

in Fig. 2.9. Based on the analytical proof of friction force versus real contact

area and mechanical deformation, large mechanical deformation occurs at the

small real contact area while the adhesion force is small. A minimum overall

friction force is observed at a certain value of a real contact area.

It can be seen from Fig. 2.9 that adhesion force increases with the real

contact area but ploughing action/mechanical deformation is inversely

)(a

)( nF

nF

Chapter 2

41

proportional to the real contact area, rA . Hence, the friction force is related to

the real contact area,

radhesiverndeformatiorfriction AFAFAF (2.17)

When the real contact area 1A > 2A ,

1AFadhesive > 2AFadhesive (2.18)

for the ploughing action/deformation action;

1AF ndeformatio < 2AF ndeformatio (2.19)

Therefore,

r

adhesive

A

F

>0, (2.20)

r

ndeformatio

A

F

< 0, (2.21)

Based on these equations, the minimum friction force is obtained at a certain

non-zero contact area.

Chapter 2

42

Fig. 2.9. Friction force versus real contact area.

2.4.5. Wear mechanism

Wear can be defined as the loss of material from one or both of the

contacting surfaces sliding over each other. The surface damage can be caused

by mechanical forces, frictional work, impact forces, capillary forces, van der

Waals forces, etc. The loss of material is measured by the mass or volume of a

slider. There are four main wear mechanisms [Burwell 1957 and Opondo and

Bessell 1982] viz. adhesive wear, abrasive wear, corrosive wear and surface

fatigue. Fig. 2.10 shows the classification of wear models [Briscoe and Sinha

2002].

Force

Mechanical deformation/

ploughing

Adhesion force

Small real contact area

Large real contact area

Minimum point Overall friction force

Real contact area

Chapter 2

43

Fig. 2.10. Classifications of wear of polymers.

2.4.5.1. Adhesive wear

Adhesive wear is a basic phenomenon that takes place between two

contacting surfaces rubbing against each other [Rabinowicz 1976]. In adhesive

wear, plastic deformation leads to the transfer of materials from the worn parts.

When two surfaces are brought together and a normal load is applied onto the

Material Response Approach

Polymer class model:

Elastomers

Thermosets

Glassy polymers

Semi-crystalline

polymers

Classifications

of wear of

polymers

Phenomenological Approach

Origin of wear process model:

Abrasive wear

Adhesive wear

Transfer wear

Chemical wear

Fatigue wear

Erosion

Delamination wear

Generic Scaling Approach

Two-term interacting

model:

Cohesive wear

Interfacial wear

Chapter 2

44

contact asperities, the local contact pressure is extremely high so that the yield

stress at the contact is exceeded. As a result of these high local contact

pressures, the junction of surface contact is sheared and new junctions are

formed. This leads to wear, i.e., the transfer of material from the softer to the

other harder surface. The amount of wear is based on the position at which the

junction of two surfaces is sheared and damaged. Due to the effect of the

material transfer and the ambient conditions, the steady-state coefficient of

friction and the wear rate can be varied. The schematic diagram of adhesive

wear is shown in Fig. 2.11.

Fig. 2.11. Schematic diagram of adhesive wear.

In Archard equation for adhesive wear [Arachard 1953], the total

volume of wear debris Q produced is given by

o

n

p

FkQ

3 (2.22)

Wear materials

Load

Chapter 2

45

where k is the probability of an asperity contact producing a wear particle,

nF is the total normal load and op is the yield pressure. Based on Equation

2.22, three laws of wear have also mentioned that are given as below:

(1) The wear volume is proportional to the distance of sliding.

(2) The wear volume is proportional to the total normal load.

(3) The wear volume is inversely proportional to the yield stress, or the

hardness, of the soft material.

2.4.5.2. Abrasive wear

Wear by abrasion occurs when a rough sliding object is loaded against

a softer surface. Some parts of materials are transferred onto the rough sliding

surface and other parts of materials are trapped between two contacting

surfaces as wear debris. The literature denotes two types of abrasion wear

based on the type of interfaces:

(1) Two-body abrasive wear: A rough hard surface passes over a softer

surface. Only two surfaces are involved in this wear mechanism.

(2) Three-body abrasive wear: Particles of hard surface or foreign

particles are trapped between two sliding surfaces. Dust and grit are the

largest sources of abrasive particles.

The large variability in wear measurements is due to the effect of dust

and foreign particles. The rate of wear can increase or decrease in the abrasion

wear process. Ratner et al. [1964] and Lancaster [1969 and 1973] mentioned

Chapter 2

46

the abrasive wear that was related to the bulk properties and an equation of the

form

eHS

WvKV

(2.23)

where V is the wear volume, K is the proportionality constant, is the

coefficient of friction, W is the normal load, is the sliding velocity, H is the

hardness of the polymer, S is the ultimate tensile stress and e is the %

elongation to break. A linear relation between V and eS

1 has been observed

based on the validation of Equation 2.23. The wear mechanism by adhesion

and abrasion depends on direct contact between two mating surfaces. A

schematic diagram of abrasive wear is shown in Fig. 2.12.

Fig. 2.12. A schematic diagram of abrasive wear.

2.4.5.3. Fatigue wear/rolling wear

Under a repetitive cyclic load, the surface of a material is weakened.

Fatigue wear is characterized by progressive and localized structural damage

Load

Abrasive debris

Chapter 2

47

and occurs when the applied load is higher than the fatigue/fracture strength of

the material. Rolling contact fatigue wear is caused by the contact between

asperities with very high amplitude oscillatory contact stress during sliding.

These stresses lead to the formation and propagation of cracks. In the presence

of fatigue wear, there is a high coefficient of friction and a high degree of

plastic deformation. Factors such as the physical, mechanical and chemical

properties of the solid surface and lubricant, temperature and environment can

have an effect on fatigue wear. A schematic diagram showing the outcomes of

fatigue/rolling wear is shown in Fig. 2.13.

Fig. 2.13. Outcomes of fatigue/ rolling wear.

2.5. Effect of different normal loads

Based on Amontons’first law, the friction force is proportional to the

normal applied load. The friction force, a function of the normal load of the

form, is mentioned as

nWkF (2.24)

Wear particles are

released after N

fatigue cycles

Load

Crack

Chapter 2

48

where k and n are system-dependent constants. Therefore, the coefficient of

friction with load is given as [Stuart 1997]

)1( nkW (2.25)

where 13

2 n . According to Equation 2.25, the decreased coefficient of

friction was occurred while increasing the normal loads. Unal et al. [2003] and

Sirong et al. [2007] observed that the coefficients of friction of the polymers

sliding against metals decrease with increasing normal loads. However,

different observations were found for the coefficients of friction with different

normal loads and the coefficients of friction of the polymers (PTFE, POM and

PEI) increased with increasing the normal loads due to the critical surface

energy of these polymers [Unal et al. 2003]. Stuart [Stuart 1997] observed that

the coefficients of friction for PEEK with chloroform decreased until a critical

load was reached, and thereafter the coefficients of friction dramatically

increased with further increase in normal load. Myshkinet al. [2005]

summarized the effects of normal load on the coefficient of friction as shown

in Table 2.2.

Chapter 2

49

Table 2.2. The effect of normal load on the coefficient of friction [Myshkin et al. 2005]

Reference Materials and load Graphic representation

Bowers, Clinton, and

Zisman

2-15N, steel-polymer

(PTFE, PFCE, PVC,

PVDC, PE)

Shooter and Thomas 10-40N, steel-polymer

(PTFE,PE, PMMA,

PC)

Shooter and Tabor 10-100N, steel-

polymer (PTFE, PE,

PMMA, PVC, nylon)

Rees Steel – polymer (PTFE,

PE, nylon)

Bartenev , Schallamach Theory, steel - rubber

Kragelskii Theory, steel - rubber

2.6. Effect of wettability of surfaces

Surface adhesion plays an important role in the friction and wear of

polymers. A common way toquantify the adhesion of polymers is through the

surface wettability/interfacial energy of tribopair joints.The study of wetting

phenomena on polymer surfaces has been developed since 1980 [Bhatia et al.

1988].Contact angle measurements are performed to check the wettability of

polymer surfaces.

Wettability is classified according to the contact angles. Surfaces can

be highly hydrophilic (contact angle: ~ 0 – 45o) or highly hydrophobic

Chapter 2

50

(contact angle: ~ 90 – 180o) as shown in Fig. 2.14. Lower contact angle

provides higher surface energy and good wetting. Surface wettability is largely

dependent on both chemical composition and surface roughness/texture.

Fig. 2.14. The contact angle between a liquid and a solid surface for (a)

hydrophilic and (b) hydrophobic surfaces.

The wettability of surfaces with rough textures has been widely

measured in order to study their tribological behaviors. A homogeneous

wetting regime is observed when the liquid fills the grooves of rough surfaces

while a heterogeneous wetting regime is observed when the air is trapped in

the grooves of rough surfaces. Cassie-Baxter and Wenzel models are used to

describe the wetting of surfaces.

2.6.1. Wenzel model

The homogeneous wetting regime described by the Wenzel model is shown in

Fig. 2.15 and the measurement of contact angle on surface roughness is given

by the Wenzel equation:

Y cos cos rW (2.26)

where W is the observed contact angle, r is the roughness ratio of the surface

and Y is the Young contact angle on a smooth surface. The roughness ratio, r ,

(a)

(b)

Chapter 2

51

of the surface is the ratio of true area of the solid surface to the apparent area.

Based on the Wenzel’s relation, the contact angle decreases on a hydrophilic

surface due to the effects of surface roughness. Conversely, the contact angle

will increase on a hydrophobic surface.

Fig. 2.15. A schematic illustration of the homogeneous (Wenzel) wetting.

2.6.2. Cassie-Baxter model

The heterogeneous contact angle CB using the Cassie-Baxter equation

(shown in Fig. 2.16) is given as:

airff cos )(1 cos cos WCB (2.27)

The Cassie-Baxter equation can be simplified when the contact angle of water

droplets is 180o:

)1( cos cos WCB ff (2.28)

Hydrophobic contact angles are observed when heterogeneous wetting occurs.

Chapter 2

52

Fig. 2.16. A schematic illustration of the heterogeneous (Cassie-Baxter) wetting.

2.7. Numerical simulations of textured surfaces

Numerical simulation using the finite element method is accepted as a

powerful technique to study frictional processes and is also useful as a

predictive tool to understand the nature of the contact, develop new designs

and asperities and identify critical parameters affecting frictional behaviors. A

lot of research has been carried out through numerical simulations. Tangena

and Wijnhoven [1985] investigated the interactions of two cylindrical

asperities between a hard rigid asperity and a soft asperity. Based on the

results, the adhesion friction between two surfaces increased the shear force

and the coefficient of friction. However, the normal force was not affected.

In 1988, Komvopoulous solved an elastic contact problem between a

rigid surface and a layered medium using the finite element method. It was

shown that the maximum shear stress at the layer substrate interface and the

von Mises stress were strongly dependent on the normalized layer thickness.

Kral et al. [1995 and 1996] examined surface deformation, stresses and strain

fields due to indentation and sliding by performing finite element simulations.

Faulkner and Arnell [2000] developed a statistical model that relates

Chapter 2

53

the sliding interaction of two and three-dimensional elasto-plastic materials

and hemispherical asperities. It was shown that the overall coefficient of

friction was lower when the asperities were spheres as opposed to cylinders.

Moreover, Ramachandra and Ovaert [2000] showed that the normal contact

pressure was affected due to the presence of geometry with the sharp and

rounded edges. Gong and Komvopoulos [2003] presented the contact pressure

distribution, surface tensile stress and surface equivalent plastic stain on the

patterned layered media using the finite element method. It was observed that

the maximum contact pressure is a strong function of pattern geometry and

high pressures occurred at the sharp edges of the patterned surface. Ye and

Komvopoulos [2003] studied the residual stress within the surface layer and

coefficient of friction at the contact region using a finite element analysis of

indentation and sliding contact of elastic-plastic layered media.

Based on the success of previous studies, the finite element analysis

method is well developed in the analysis of tribological processes and can be

used to determine contact characteristics due to surface asperities and patterns.

2.8. Summary of research strategy

From the above literature review on the tribology of polymer thin film,

it is seen that thin polymeric films are widely used act as a protective layer to

enhance the life span of products. Previous studies have shown that SU-8 is a

promising coating layer in BioMEMS applications. Different surface textures

(e.g. lotus leaf, micro-dot and groove) have been developed on SU-8 thin films.

However, to solve tribological problems, a new fabrication technique is still

Chapter 2

54

necessary to obtain rounded corners, which may act as less stress

concentration, in the textured patterns on SU-8 thin films. Patterns with

rounded corners are fabricated on PDMS bulk polymer to study the behavior

of these desired features.

The following is the research strategy adopted in this thesis:

1. Development of surface patterns with both positive asperity

(fingerprint pattern and micro-dot pattern) and negative asperity

(honeycomb pattern) on SU-8 thin films and PDMS bulk polymer

(using different fabrication methods),

2. Understanding of the friction and wear durability of surfaces with

positive asperities such as fingerprint patterns and micro-dot patterns,

3. Understanding of the friction and wear durability of surfaces with

negative asperities such as honeycomb patterns,

4. Understanding of the effect of PFPE thin film layers with textured

surfaces on wear durability,

5. Correlating contact area, spatial texture density and coefficient of

friction using analytical calculations based on Hertzian contact,

6. Correlating coefficient of friction of textured surface and applied

normal loads,

7. Correlating finite element analyses and physical experiments on

tribology of textured surfaces.

Chapter 3

55

Chapter 3

Materials, Test Equipment, and Experimental

Methodologies

In this chapter, a summary of the materials that were used in our

research and various common experimental methods for the formation of

surface textures are described. The fabrication procedures for the textured

polymer surfaces are also discussed.

Chapter 3

56

3.1. Materials

3.1.1. Silicon

Polished n-type bare silicon (Si) wafers (obtained from SYST

Integration Pte Ltd, Singapore) of approximately 525 ± 25 µm in thickness

and 0.4 nm in roughness with an orientation of 100 and a hardness of 12.4

GPa were used as the substrates. An oxygen plasma cleaning process (using a

Harrick Plasma Cleaner, PDC-32G) was conducted for 10 minutes, with the

maximum RF power supply of 18 W, on the surface of bare silicon to remove

any contamination prior to coating and texturing.

3.1.2. SU-8

SU-8 is a high contrast, negative epoxy type, near ultra violet (UV)

photo resist, which is based on EPON SU-8 resin that was originally patented

by IBM-Wastson Research Center (Yorktown Height – USA, US Patent No.

4882245) in 1989 [Pan et al. 2002 and Yu et al. 2006]. However, it is now

primarily supplied by Microchem. SU-8 consists of three fundamental

elements (1) an EPON SU-8 epoxy resin, (2) an organic solvent such as

gamma-butyrolactone (GBL), and (3) a photoinitiator such as triaryl sulfonium

salts. It is capable of patterning high aspect ratio (HAR) structures and it has

been widely used in many new MEMS devices and other applications, due to

the advantages of its high sensitivity, chemical resistance, and

biocompatibility. The chemical structure of an EPON SU-8 epoxy resin

consists of eight epoxy groups; its molecular structure is shown in Fig. 3.1.

Chapter 3

57

Fig. 3.1. The molecular structure of an EPON SU-8 epoxy resin material.

[https://en.wikipedia.org/wiki/SU-8_photoresist]

Table 3.1. The physical properties of Grade 2050 and Grade 2000.5 of SU-8

[Data sheet of SU-8 2000.5 and SU-8 2050, http://www.microchem.com]

Properties Units SU-8 2050 SU-8 2000.5

% Solid - 68.55 14.3

Viscosity cSt 12,900 2.49

Density g/ml 1.219 1.07

Adhesion Strength

Silicon/Glass/Glass &

HDMS

mPa 38/35/35 38/35/35

Glass Transition

Temperature

TgoC 210 210

Thermal Stability oC @5% wt.

loss

315 315

Thermal Conductivity W/mK 0.3 0.3

Coeff. of Thermal

Expansion

CTE ppm 52 52

Tensile Strength MPa 60 60

Elongation at break εb % 6.5 6.5

Young’s Modulus GPa 2.0 2.0

Dielectric Constant @

10 MHz

- 3.2 3.2

Water Absorption % 85oC/85 RH 0.65 0.65

An SU-8 layer was deposited on the surface of a silicon substrate that

is widely used in MEMS devices. Two series, SU-8 2050 and SU-8 2000.5,

were chosen for these experiments. These series of SU-8 materials are

compatible with a wide range of structure thicknesses. The physical properties

CH3 H3C

CH3 H3C

CH3 CH3 H3C H3C

Chapter 3

58

of the Grade 2050 SU-8 and Grade 2000.5 SU-8 that were used in this study

are identified in Table 3.1.

3.1.3. Polydimethylsiloxane (PDMS)

Polydimethylsiloxane (PDMS), supplied by Dow Corning Corp. USA,

is a silicone-based organic polymer. It is an inert, non-toxic, and non-

flammable material that is widely used in many applications, such as contact

lenses, medical devices, and hydraulic fluids. It is also used in cosmetics, food,

and other domestic products. The chemical formula for PDMS is

(H3C)3SiO[Si(CH3)2O)nSi(CH3)3, where n is the number of repeating

[SiO(CH3)2] monomer units. Its brief formula is shown in Fig. 3.2.

Fig. 3.2. The chemical formula of PDMS. [http://en.wikipedia.org/wiki/Polydimethylsiloxane]

Sylgard 184 silicone elastomer base and Sylgard 184 silicone

elastomer curing agent were used in this research project to form PDMS

network samples. Sylgard 184 silicone elastomer base was mixed with the

cross-linking agent of Sylgard 184 silicone in the ratio of 10:1. PDMS was

used as a master piece mould to fabricate positive fingerprint patterns on SU-8.

The physical properties of Sylgard 184 silicone elastomer, which was used in

this study, are described in Table 3.2.

http://en.wikipedia.org/wiki/Polydimethylsiloxane

Chapter 3

59

Table 3.2. The physical properties of Sylgard 184 silicone elastomer [Data

sheet of Sylgard 184 silicone elastomer, http://www.dowcorning.com]

Properties Unit Value/Description

Color - Colorless

Viscosity (Base) cP 5,100

Viscosity (Mixed) cP 3,500

Thermal Conductivity W/moK 0.27

Dielectric Strength kV/mm 19

Volume Resistivity Ohm*cm 2.9E+14

Dielectric Constant at 100 Hz 2.72

Tensile Strength MPa 6.7

3.1.4. Perfluoropolyether (PFPE)

In order to improve the wear lives, PFPE (0.2 wt% solution of PFPE

Zdol 4000 in H-Galden ZV60, purchased from Ausimont INC) was coated on

the SU-8 tested samples. The chemical formulae of Zdol and H-Galden ZV60

are:

PFPE (Zdol 4000): HOCH2CF2O-(CF2CF2O)P-(CF2O)P-CF2CH2OH

H-Galden ZV60: HCF2O-(CF2O)p-(CF2CF2O)q-CF2H

where the ratio of p/q is 2/3. The physical properties of PFPE (Fombin Z dol

4000) are described in detail in Table 3. 3.

3.1.5. SU-8 developer

For the SU-8 negative photo resists, the development step was

conducted to remove non-illuminated regions of SU-8 after the post-baking

process. The SU-8 2000 photoresist can be used for the processes of

immersion, spray, or spray-puddle with an SU-8 developer. The wafer was

immersed in the SU-8 developer to completely remove the unexposed part of

Chapter 3

60

SU-8. The duration of development was dependent on the thickness of the SU-

8 layer. The recommended development times for the immersion process are

provided in Table 3.4 [Data Sheet of SU-8].

Table 3.3. The physical properties of PFPE (Fombin Z dol 4000)

Properties Units Value/Description

Functional Group - Alcohol (-OH)

Appearance Visual Clear liquid

Colour APHA Colourless

MW (NMR) Amu 4000

Difunctional content

(NMR)

% 90

C2/C1 ratio (NMR) - 1

Kinematic Viscosity cSt 100

Density @ 20oC Kg/m

3 1820

Vapour Pressure @ 20oC Torr 1x10

-8

Vapour Pressure @ 100oC Torr 1x10

-4

Refractive Index @ 20oC - 1.296

Surface Tension @ 20oC mN/m 22

Polydispersity @ 20oC Mw/Mn 1.15

Table 3.4. Development times for SU-8 developer [Data sheet of SU-8 2050,

http://www.microchem.com]

Thickness (m) Development time (minutes)

25 – 40 4 – 5

45 – 75 5 – 7

80 – 110 7 – 10

115 – 150 10 – 15

160 – 225 15 – 17

3.1.6. Silicon nitride (Si3N4)

A silicon nitride (Si3N4) ball was used as a counterface ball to

investigate the tribological behaviors of the textured specimens. The used

diameters of counterface balls were 2 mm, 4 mm, and 6 mm. The roughness

and hardness of the silicon nitride ball were 5 nm and 1500 HV, respectively.

Chapter 3

61

The poisson’s ratio and Young’s modulus of the ball were 0.22 and 310 GPa,

respectively.

3.2. Process flow

3.2.1. Process flow chart of the fabrication process

3.2.1.1. Process flow chart of the fabrication process for the SU-8

textured surface

The process flow chart of the fabrication methods for the SU-8

textured surface is shown in Fig 3.3.

3.2.1.2. Process flow chart of the fabrication process for the PDMS

textured surface

The process flow chart of the fabrication methods for the PDMS textured

surface is shown in Fig. 3.4.

Chapter 3

62

Fig. 3.3. Process flow chart of the fabrication methods for the SU-8 textured surfaces.

Preparation of silicon (Si) samples

Cut out 35 mm x 35 mm Si samples

Cleaning process

Clean and oxygen plasma treatment

on silicon substrate

Spin coating of SU-8 (Grade 2050)

Pre-baking of SU-8

Patterns on SU-8

Negative

fingerprint

pattern

Positive

fingerprint

pattern

Honeycomb

pattern

Curing process of SU-8

Post-baking of SU-8

Developing process for

honeycomb pattern

Curing process of SU-8

Post-baking of SU-8

Micro-dot patterns of

SU-8 (Grade 2000.5)

using polymer jet

printer

Pre-baking of SU-8

Chapter 3

63

Fig. 3.4. Process flow chart of the fabrication methods for the PDMS textured

surfaces.

3.2.2. Process flow chart of the experimental tribological characterization

The process flow chart of the experimental tribological

characterization of the SU-8 textured specimens is shown in Fig. 3.5.

3.3. Sample preparation and fabrication methods

3.3.1. Preparing of silicon (Si) substrates

Polished n-type bare silicon (Si) wafers (obtained from SYST

Integration Pte Ltd, Singapore) of approximately 525 ± 25 µm in the thickness

with an orientation of 100 were used as substrates for SU-8 textured and

untextured surfaces. 35 mm x 35 mm Si samples were cut and used as

substrates for the preparation of Si samples; the preparation of a silicon sample

is shown in Fig. 3.6.

Mixing materials

Mixed PDMS prepolymer and the

curing agent by the ratio of 10:1

Degassing process

Degassing PDMS mixture

Pour PDMS mixture onto the

substrate/textured surface

Peel PDMS replica from master piece

Curing process

70oC for 5 hours

Chapter 3

64

Fig. 3.5. Process flow chart of the tribological characterization of the SU-8 textured

specimens.

Fig. 3.6. Preparation of a silicon sample.

3.3.2. Cleaning process for silicon substrates

The surface cleaning process for Si substrates is an essential step for

further coating. In this study, the acetone cleaning was applied to the Si

substrates. The oxygen plasma cleaning was conducted using an oxygen

Lubricated condition

Tribological characterization

- Coefficient of friction

- Wear durability

Dry condition

Surface characterization

- Geometry measurements

- Surface roughness

- Water contact angle and

surface energy

Tribological characterization

- Initial coefficient of friction

- Wear durability

- Different normal loads

- Different counterface balls

Test specimens

35 mm

35 mm

Chapter 3

65

plasma cleaner (Harrick Plasma – PDC32G as shown in Fig. 3.7) for ten

minutes with the maximum RF power supply of 18 W, to remove

contaminants and particles from the Si substrates.

Fig. 3.7. Harrick Plasma – PDC32G machine.

3.3.3. Spin-coating of SU-8 on silicon substrate

Grade 2050 of SU-8 with a viscosity of 4,500 cSt was spin-

coated onto Si substrate before the patterns were made. The process of spin-

coating SU-8 on a silicon substrate is shown in Fig. 3.8. The spin-coating

process was performed at a rotational speed 500 rpm for 10 seconds with an

acceleration of 100 rpm/s, and then at 2000 rpm for 30 seconds with an

acceleration of 100 rpm/s, and finally at 2000 rpm for 30 seconds with an

acceleration of 300 rpm/s using a spin coater (P6700 Specialty Coating

Systems, Indianapolis, IN); this process produced an SU-8 layer that was 80

µm thickness. Although SU-8 thickness range is 40-170 µm, the thickness of

80 µm is a good height to pattern a good fingerprint structure. The relationship

between the thickness of SU-8 Grade 2050 thin film and the spin speed are

shown in Fig 3.9.

Chapter 3

66

Fig. 3.8. The process of spin-coating SU-8 on a silicon substrate.

Fig. 3.9. The relationship between the thickness of SU-8 Grade 2050 thin film and the

spin speed [Data Sheet of SU-8 2050].

3.3.4. The curing process for SU-8 on Si substrate

Fig. 3.10 shows the hot plate for pre-baking and post-baking of SU-8.

3.3.4.1. Pre-baking process

After coating SU-8 on Si substrates, a soft baking (pre-baking) process

was performed using a hot plate. For SU-8 thin film with a thickness of 80 µm,

SU-8 2050

80 µm

Silicon substrate Silicon substrate

SU-8

Chapter 3

67

the soft baking time of five minutes was used at 65oC followed by 15 minutes

at 95oC.

3.3.4.2. Post-baking process

The post exposure baking (PEB) process was performed at 65oC for a

minimum of 30 minutes to create cross linking of SU-8.

Fig. 3.10. Hot plate used for pre-baking and post-baking processes.

3.3.5. Exposing SU-8 to ultraviolet light

The fabricated specimens were exposed to UV light for 10 – 15

seconds using a Mask Aligner machine with a power of 215 mJ/cm2. Fig. 3.11

shows a photo of a mask aligner (Model H94-25C, SVC Technology Co., Ltd)

for UV light exposure treatment to SU-8.

Chapter 3

68

Fig. 3.11. Mask aligner machine for UV light treatment.

3.3.6. Preparation of the flat SU-8 surface

Fig. 3.12 shows the fabrication process for the flat SU-8 surface. A

small silicon specimen (35 mm x 35 mm) that was 525 ± 25 µm in the

thickness was cut from the silicon wafer. The silicon specimen was cleaned

with acetone and then the silicon specimen was conducted with oxygen

plasma for ten minutes using the maximum RF power supply of 18 W. In

order to obtain a thickness of 80 µm of SU-8 layer, the spin coater was then

used to overcoat Grade 2050 polymer on the silicon, as mentioned in Section

3.3.3. Later, a soft baking process was conducted on the Si/SU-8 surface using

a hot plate for 5 minutes at 65oC followed by 15 minutes at 95

oC. The UV

light was exposed to the fabricated specimen for 10 – 15 seconds using a Mask

Aligner machine. The post exposure baking process was then carried out at

65oC for a minimum of 30 minutes. Finally, the resulting specimen for the flat

Chapter 3

69

SU-8 surface was termed Si/SU-8. The fabricated samples were stored in a

clean room for 24 hours before investigating the future process.

Fig. 3.12. The fabrication process for the flat SU-8 surface.

3.3.7. Preparation of SU-8 honeycomb textures on Si/SU-8 surface

The fabrication process for obtaining the honeycomb structure on

Si/SU-8 is shown in Fig. 3.13. Before patterning the honeycomb textures, the

surface of silicon substrate was cleaned with acetone, and the oxygen plasma

cleaning process was performed for ten minutes with the maximum RF power

supply of 18 W. A film of SU-8 (Grade 2050) was overcoated on Si substrates

using a spin coater to obtain a thickness of 80 µm; as mentioned in Section

3.3.3., to remove the solvent from the coated layer, the sample was baked for

five minutes at 65oC, followed by 15 minutes at 95

oC for the soft baking

process. The glass mask with a honeycomb pattern was used to fabricate the

honeycomb pattern on SU-8. An exposure dose of 20 - 25 seconds, at a power

of 210 mJ/cm2, was used to pass through the glass mask and create the

honeycomb patterns on the layer with a thickness of 80 µm. Subsequently, the

SU-8 exposure SU-8 coating Soft baking Post exposure baking

Sample preparation of flat SU-8 surface

Chapter 3

70

post exposure baking (PEB) process was performed at 65oC for five minutes

and at 95oC for 15 minutes. The fabricated sample was then immersed in the

solvent of SU-8 developer for 15 - 20 minutes to remove the unwanted resist

on SU-8.

Fig. 3.13. The fabrication process of the SU-8 honeycomb pattern.

Subsequently, a rinsing process was carried out with acetone to remove

the developer, and then nitrogen gas was used to dry the surface. The final

resultant specimen for the SU-8 honeycomb pattern was termed Si/SU-8/HC.

Before proceeding with the tribological tests, the fabricated samples were

stored in a clean room for 24 hours.

Soft baking SU-8 coating

Sample preparation of SU-8 honeycomb pattern

Post exposure

baking

SU-8 Development Rinse and dry

UV

Light

SU-8 exposure

Chapter 3

71

3.3.8. Preparation of SU-8 micro-dot textures on Si/SU-8 surface

Fig. 3.14 (a) shows the polymer jet printer (JetLab4, MicroFab

Technologies Inc, Plano, TX). The fabrication process of the desired SU-8

micro-dot patterns using a drop-on-demand device is shown in Fig. 3.14 (b).

In order to fabricate the flat SU-8 surface, the cleaning process that included

acetone and oxygen plasma cleaning was carried out on the silicon substrates.

The SU-8 polymer (Grade 2050), with a thickness of 80 µm, was coated on

the silicon substrates. The pre-baking (soft baking) process, exposure process,

and post exposure baking process were conducted on the Si/SU-8 flat surface.

The detailed fabrication procedure for the flat SU-8 surface is described in

Section 3.3.6.

A drop-on-demand PJP method was used to fabricate an SU-8 micro-

dot pattern on Si/SU-8 surfaces. Grade 2000.5 SU-8 with a viscosity of 2.49

cSt was dispensed from the nozzle of PJP machine using the amount of back

pressure (approximately -2.8 KPa) required by the controller to create the

micro-dot pattern. The diameter of the nozzle on the dispenser was 20 µm.

The piezoelectric actuator surrounding the glass capillary required an

adjustment of the parameters of waveform, such as 3.0 µs of rise time, 40 µs

of dwell time at 12 V, 3.0 µs of fall time, 90 µs of echo dwell at 25 V, and 3.0

µs of final rise time, in order to print the desired micro-dot patterns. SU-8

micro-dot patterns were printed as a rectangular array of 40 dots by 40 dots on

Si/SU-8 surfaces with varying pitch lengths (P = 200 µm, 300 µm, 400 µm,

500 µm, and 600 µm).

Chapter 3

72

Fig. 3.14. (a) Polymer jet printer and (b) the fabrication process of the desired SU-8

micro-dot patterns using a drop-on-demand device.

(b)

(a)

SU-8 exposure SU-8 coating Soft baking Post exposure baking

Sample preparation of flat SU-8 surface

Sample preparation of SU-8 micro-dot patterns

Soft baking

SU-8 exposure

Post exposure baking

(a)

Chapter 3

73

Afterwards, the fabrication process of the micro-dot pattern on Si/SU-8

surfaces, the pre-baking process, exposure to UV light, and post-baking

processes were carried out as mentioned in Section 3.3.4 and Section 3.3.5.

The resultant specimen for the SU-8 micro-dot patterns using the PJP method

was labelled Si/SU-8/MD.

3.3.9. Preparation of SU-8 negative fingerprint textures on Si/SU-8 surface

Fig. 3.15 shows the fabrication steps for obtaining the negative

fingerprint textured specimens. The removal of contaminants from the Si

substrates was carried out using acetone and the oxygen plasma cleaner that

was discussed in Section 3.3.2. First, the spin-coating process was carried out

with Grade 2050 SU-8 on the Si substrates with a rotational speed of 500 rpm

for 10 seconds and an acceleration of 100 rpm/s, and then at 2000 rpm for 30

seconds with an acceleration of 300 rpm/s, to fabricate an SU-8 layer that was

80 µm thick.

A soft baking time of 5 minutes at 65oC was followed by 15 minutes at

95oC using the hot plate, after spin-coating an SU-8 layer on the Si surface.

The fingerprint of a subject’s index finger was used to fabricate the negative

fingerprint pattern on the Si/SU-8 surface after the soft baking process. The

index finger was pressed on an SU-8 polymer and the actual fingerprint was

converted as a negative pattern; that is, an inverted 3D fingerprint on the SU-8

polymer. These fabricated specimens were placed under UV light for 10 - 15

seconds at a power of 215 mJ/cm2. A post exposure baking time of at least 30

minutes at 65oC was used for the tested specimens. The same subject’s

Chapter 3

74

fingerprint was used for all of the experiments. Prior to the experiments, the

subject’s hand was washed thoroughly and a tissue was used to remove any

trace of sweat. For this experiment, the skin was assumed to be in the dry

condition. After the post exposure baking process, the fingerprint pattern was

developed on Si/SU-8 surfaces. The final specimen was entitled Si/SU-8/NFP.

Fig. 3.15. The fabrication process for the desired SU-8 negative fingerprint pattern

on Si/SU-8.

3.3.10. Preparation of SU-8 positive fingerprint textures on Si/SU-8 surface

The fabrication process for the SU-8 positive fingerprint pattern on

Si/SU-8 is mentioned in Fig. 3.16. These positive fingerprint patterns on

Si/SU-8 can be formed using the two master piece moulds of the positive

fingerprint and negative fingerprint patterns on PDMS. Polydimethylsiloxane

(PDMS), supplied by Dow Corning Corp. USA, was used in the fabrication of

the positive fingerprint pattern on Si/SU-8. The detailed fabrication steps for

PDMS negative and positive fingerprint patterns used to obtain PDMS moulds

are mentioned in Section 3.3.11 and Section 3.3.12. To prepare the SU-8

SU-8 coating Soft baking Press an index

finger on SU-8

polymer

SU-8

exposure

Post exposure

baking

Sample preparation of SU-8 negative fingerprint pattern

Chapter 3

75

positive fingerprint pattern, the fabricated PDMS negative fingerprint

specimen was used as a master piece mould and it was pressed on the Si/SU-8

surface to obtain a positive fingerprint pattern on Si/SU-8. The pre-baking,

exposure to UV light, and post-baking processes were performed with the

same parameters that were discussed in Section 3.3.4 and Section 3.3.5. The

obtained fabricated sample was a positive fingerprint pattern on Si/SU-8,

which is the same as the pattern of furrows and ridges of a real index finger;

however, the obtained dimensions of the sample were different than a real

index finger. The final resultant specimen of the positive fingerprint pattern on

Si/SU-8 was called Si/SU-8/PFP.

Fig. 3.16. The fabrication process of the desired SU-8 positive fingerprint pattern.

3.3.11. Preparation of the PDMS positive fingerprint pattern

Fig. 3.17 shows the fabrication process of the PDMS positive

fingerprint pattern. Polydimethylsiloxane (PDMS) was supplied by Dow

SU-8 coating Soft baking Press a PDMS negative

fingerprint mould on

SU-8 polymer

SU-8

exposure

Post exposure

baking

Sample preparation of SU-8 positive fingerprint pattern

Chapter 3

76

Corning Corp, USA. To obtain the PDMS moulds, firstly, PDMS prepolymer

was mixed with a cross-linking agent with a ratio of 10:1, in a petri dish. The

mixture of PDMS was then stirred thoroughly for three minutes. To remove

the bubbles from the mixture of PDMS, it was placed in a vacuum chamber

for 20 minutes to degas. When there were no longer bubbles in the mixture,

the degassing process was complete. Once all of the bubbles were removed,

the mixture of PDMS was slowly poured on the master piece of the SU-8

negative fingerprint pattern mould in order to fabricate the positive fingerprint

pattern on PDMS. Then, the sample was placed in a vacuum chamber again to

remove air bubbles from the interface layer. The PDMS specimen was placed

in an oven at 70oC for 5 hours. Once the PDMS casting was fully hardened, it

was removed from the oven and peeled from the master piece mould of the

SU-8 negative fingerprint pattern, in order to fabricate the PDMS positive

fingerprint pattern. The final resulting specimen was termed PDMS/PFP.

Fig. 3.17. The fabrication process for the desired PDMS positive fingerprint pattern.

PDMS mixture (10:1)

Sample preparation of PDMS positive fingerprint pattern

Pour PDMS mixture ono

the SU-8 negative

fingerprint pattern mould

Curing process PDMS positive

fingerprint

pattern

Chapter 3

77

3.3.12. Preparation of the PDMS negative fingerprint pattern

Fig. 3.18 shows the fabrication process of the PDMS negative

fingerprint pattern. In this PDMS negative fingerprint pattern fabrication

process, the PDMS positive fingerprint pattern was used as a mould in order to

fabricate the negative fingerprint pattern on PDMS. A weight ratio of 10:1 of

PDMS and a curing agent was used in this fabrication. Degasing and curing

processes were carried out as explained in Section 3.3.11. The final resulting

specimen for PDMS negative fingerprint pattern was labelled PDMS/NFP.

Fig. 3.18. The fabrication process for the desired PDMS negative fingerprint pattern.

3.3.13. PFPE lubricant

A 0.2 wt% solution of perfluoropolyether (PFPE) Zdol 4000 was

dissolved using H-Galden ZV60 as the solvent. The chemical formulae of

PFPE (Z-dol 4000) and H-Galden ZV60 are:

PDMS mixture (10:1)

Sample preparation of PDMS negative fingerprint pattern

Pour PDMS mixture onto

PDMS positive fingerprint

pattern mould

Curing process PDMS negative

fingerprint

pattern

Chapter 3

78

PFPE (Z-dol 4000): HOCH2CF2–(CF2CF2O)p–(CF2O)q–OCF2CH2OH

H-Galden ZV60: HCF2O–(CF2O)p–(CF2CF2O)q–CF2H

where the ratio of p/q is 2/3. The dip-coating method was used to coat the

tested specimens and the dip-coating machine was shown in Fig.3.19. The

dipping and withdrawal speeds of the specimen during dip-coating were kept

constant at 2.1 mm/s and a dipping duration of 30 seconds was used to obtain

2 – 4 nm thick PFPE films. The elliposmetry was employed to study the

thickness of PFPE. These samples were also kept in a clean room for 24 hours

before carrying out any characterization.

Fig. 3.19. PFPE lubricant on the tested specimen using the dip-coating method.

Chapter 3

79

3.4. Surface analysis equipment and techniques

3.4.1. Optical microscopy

Fig. 3.20 shows an image of an optical microscope. The optical images

of surface textures on SU-8 and PDMS were captured using optical

microscopy. The tested specimen was placed on the specimen holder and LAS

EZ software was used with the microscope to capture the optical images;

measure the dimensions of surface textures; and observe the surface conditions,

behaviors, and wear tracks of the tested specimens. Magnification sizes of the

images from 50 to 500 times could be captured on the tested samples.

Fig. 3.20. The optical microscope.

3.4.2. Optical profiler

Fig. 3.21 shows an image of a Wyko NT1100 optical profiler. The

measurement techniques used in the optical profiler machine are optical phase-

shifting interferometry (PSI) and white light vertical scanning interferometry

Chapter 3

80

(VSI). This profiler is capable of three-dimensional measurements with non-

contact static profile measurements. The phase shifting interferometry (PSI)

mode is used to measure smooth surfaces and small steps of the surfaces, and

the vertical scanning interferometry (VSI) mode is used to measure rough

surfaces and steps up to one millimeter high. The vertical resolution is less

than 1Å and the vertical measurement range is between 0.1 nm and 1 mm.

The investigation of the surface roughness and topography of the

untextured and textured surfaces was carried out using a Wyko NT1100

optical profiler (supplied from Veeco Instruments Inc.). Wyko Vision32

software was used to capture the images and measure the samples. The tested

sample was placed on the stage and the light intensity was adjusted.

Subsequently, fingers appeared by lifting the stage of specimens with the

focus lever. The height and thickness of the film and the textured surface were

also examined by the optical profiler. The result was observed with “2D

analysis” or “3D analysis”.

3.4.3. Field emission scanning electron microscopy (FESEM)

The FESEM (Hitachi S4300) machine is shown in Fig 3.22. The wear

tracks and surface topography of the untextured and textured surfaces of SU-8

and PDMS were observed by a field emission scanning electron microscope

(FESEM) (Hitachi S4300) machine with an energy dispersive spectrometer

(EDS). The observations of the surfaces of SU-8 and PDMS were performed

with FESEM, after coating gold films (coated with JEOL, JFC-1200 Fine

Coater using 20 seconds at 10 mA) on the tested specimens.

Chapter 3

81

Fig. 3.21. A Wyko NT1100 optical profiler.

Fig. 3.22. A field emission scanning electron microscope FESEM (Hitachi S4300).

Chapter 3

82

3.4.4. Contact analysis techniques

The contact angle ( ) is a quantitative measure of wetting of a solid

with a liquid. The contact angle measurement is used to determine surface

energies, and these surface energies facilitate the understanding of wetting of

materials. The wettability of a solid surface by a liquid can be qualified by the

Young equation as:

cos LVSLSV (3.1)

where SV is the surface energy of the solid in contact with vapor, SL is the

surface energy of the solid in contact with liquid, and LV is the surface

energy of the liquid-vapor interface. Fig. 3.23 is a schematic of a liquid drop

showing the quantities in Young’s equation.

Fig. 3.23. A schematic of a liquid drop showing the quantities in Young’s equation.

Based on the profile of a droplet, the wettability of surface can be described as:

Hydrophilic – the water contact angle is lower than 90o.

Hydrophobic – the water contact angle is higher than 90o.

The measurements of water contact angles were obtained for the

untextured and textured surfaces of SU-8 and PDMS, using the VCA Optima

SG

LG

SL

Chapter 3

83

Contact Angle System (AST product Inc., USA) that is shown in Fig. 3.24.

The contact angles of the tested specimens were identified with a 0.5 µL of

deionized (DI) water droplet that was placed on the samples using a syringe.

Subsequently, the recorded water contact angles were obtained using a

microscope. Five measurements of water contact angles were performed on

each tested specimen and an average value was reported. The deviation error

of the contact angle measurements was ±3o. Deionized water, ethylene glycol,

and hexadecane were used to calculate the surface energy.

Fig. 3.24. A VCA optima contact angle system.

3.4.5. Tribological characterization

A custom-built ball-on-disc tribometer (Fig. 3.25) was used to measure

friction and wear life data. Friction force was measured by recording the

deflection of the cantilever that the ball was attached to. The measurements of

the normal and lateral deflections of the cantilever were obtained using laser

Chapter 3

84

displacement sensors (MTI Instruments Inc., New York, USA). These

displacement measurements were converted to normal load and frictional

force using the calibration chart for the cantilever used. The counterface

materials sliding against the unpatterned/patterned polymer surfaces were

silicon nitride (Si3N4) balls with a surface roughness of 5 nm and diameters of

2 mm, 4 mm, and 6 mm. The rotational speeds chosen were 2 rpm (linear

relative speed = 0.21 mm/s) for the friction tests and 500 rpm (linear relative

speed = 52.36 mm/s) for the wear tests. Different sliding speeds were used for

the friction and wear tests. Frictional heat, that is significant effect on the

tribological behaviors, would alter the physical state of polymer surfaces while

high speeds were employed on them. Sampling rates of the friction and wear

data were recorded as 10 Hz and 1 Hz, respectively.

A schematic diagram for a custom-built ball-on-disc tribometer is

shown in Fig. 3.26. Prior to the experiments, the prepared

unpatterned/patterned specimen was mounted onto the sample holder of the

tribometer spindle and Si3N4 counterface balls were cleaned with acetone. The

average coefficient of friction was obtained by conducting five repetitive

sliding tests for every tested sample under ambient conditions: a room

temperature of 25oC and a relative humidity of 58%. All experiments were

conducted in a class 100 clean booth. Experiments were stopped when the

coefficients of friction were greater than 0.3 and/or when wear tracks were

observed with the naked eye (often with fluctuating friction coefficient values)

on the surfaces of the tested specimens. The number of cycles required for

either of these events to occur was recorded as the wear life of that sample.

Chapter 3

85

The test surface conditions of every sample were observed under an optical

microscope to analyze the wear track and possible wear mechanism.

Fig. 3.25. A custom-built ball on disc tribometer.

Fig. 3.26. A schematic diagram of a custom-built ball on disc tribometer.

Ball Holder

Sample

Sample Holder

Air Bearing

Spindle Motor

Laser Devices

Controller Laser

Device

Laser

Device

Data

Logger

Chapter 3

86

3.5. Summary

In this chapter, the materials, fabrication methods, instruments, and

experimental procedures are presented. The results of the experiments are

presented in subsequent chapters.

Chapter 4

87

Chapter 4

Fabrication Method and Tribological Analysis of the

Negative Fingerprint and Positive Fingerprint Patterns

on SU-8 (with and without PFPE nano-lubricant)

This chapter presents an analysis of the tribological characterization

of the SU-8 textured surface, which may help to avoid problems associated

with friction on the flat SU-8 surface. A new method was developed for the

fabrication of a positive asperity (the negative and positive fingerprint

patterns) on surface textures. Surface characterizations were also investigated

on the SU-8 textured surfaces to compare the behaviors of the flat SU-8

surface. Experimental tests with different normal loads were carried out to

determine the load carrying capacity of the SU-8 textured surfaces.

Experimental results from the friction and wear data are presented in this

chapter.

Chapter 4

88

4.1. Introduction and objectives

Based on a review of the literature, surface texture plays an important

role in enhancing the performance of friction and wear behaviors. The positive

asperity (e.g. lotus leaf and pillar structures) of surface patterns decreases the

coefficient of friction, due to the reduction in contact area. Recently, Tay et al.

[2011] developed micro-dot patterns on the Si substrate that showed better

frictional properties of micro-dots. Singh et al. [2011] also demonstrated the

low friction and excellent wear durability of SU-8 surfaces, due to the

presence of bio-inspired polymeric patterns. Alternatively, the negative

asperity (e.g. groove and honeycomb structures) of surface patterns increases

the coefficient of friction. Therefore, the contact area between two surfaces

plays a vital role in the reduction of friction and the improvement in wear

behavior of the polymer.

In this chapter, the investigation of SU-8 polymeric material is

described. Different surface patterns were involved in the investigation to

compare the friction and wear behaviors of the flat SU-8 surfaces. SU-8

polymer was applied to a silicon substrate with a thickness of 80 µm. The

positive asperity of surface textures, such as the negative fingerprint and

positive fingerprint textures, were applied to the SU-8 surface to further

reduce the coefficient of friction and extend the wear durability. A new

development for fabricating a pattern to obtain fine surface patterns is

established. Then, the evaluation of the friction and wear behaviors of the

negative fingerprint and positive fingerprint structures are presented as the

Chapter 4

89

primary part of this chapter. In order to understand the behavior of lubricants

on a textured surface, the flat SU-8 surface and the SU-8 textured surfaces

were overcoated with PFPE using the dip-coating method. The investigations

of the effect of PFPE nano-lubricant on the flat SU-8 surface and the SU-8

textured surfaces are presented in the later part of this chapter.

By developing a positive asperity with a rounded corner on surface

textures, such as the negative fingerprint and positive fingerprint patterns on

SU-8, these protruding parts help to reduce the contact area and friction. Due

to the fabrication of the protruding parts on surface patterns of SU-8, the

contact area has been reduced and the friction drops drastically. The presence

of fingerprint patterns decreases the adhesion and the wear and debris particles

are reduced on the SU-8 textured surfaces. Furthermore, the capability of load

carrying capacity is increased because of the effect of surface patterns with a

positive asperity.

4.2. Sample preparation process for the flat SU-8 surface and the SU-8

textured surfaces

4.2.1. Fabrication of the flat SU-8 surface

Fig. 4.1 shows the fabrication procedures for the flat SU-8 surface,

which is termed Si/SU-8. The fabrication steps used in this study for the flat

SU-8 surface on Si substrate were as follows:

(1) Preparation of Si substrate: Si wafer was used as the substrate for the

SU-8 textured and untextured surfaces. 35 mm x 35 mm Si samples

Chapter 4

90

were cut and used as substrates for the preparation of the flat SU-8

samples.

(2) Cleaning process for Si substrate: The Si substrates were cleaned with

an oxygen plasma cleaner to remove any contamination. The oxygen

plasma cleaning was conducted for 10 minutes with the maximum RF

power supply of 18W.

(3) SU-8 spin-coating process on Si substrate: SU-8 (Grade 2050) was

first spin-coated on Si substrate at a rotational speed of 500 rpm for 10

seconds with an acceleration of 100 rpm/s, and then at 2000 rpm for 30

seconds with an acceleration of 300 rpm/s. The spin-coater (P6700

Specialty Coating Systems, Indianapolis, IN) was used to fabricate an

SU-8 layer that was 80 µm thick.

(4) Pre-baking process on Si/SU-8: After SU-8 was spin-coated, a hot

plate was used to perform a soft baking (pre-baking) process for 5

minutes at 65oC, and again for 15 minutes at 95

oC.

(5) Curing process for Si/SU-8: The curing process was performed under

UV light in a Mask Aligner for 10 – 15 seconds at a power of 215

mJ/cm2.

(6) Post-exposure baking process for Si/SU-8: The post-exposure baking

(PEB) process was carried out for a minimum of 30 minutes at 65oC.

Chapter 4

91

Fig. 4.1. Fabrication procedure for the flat SU-8 surface (Si/SU-8).

4.2.2. Fabrication of 3D SU-8 negative fingerprint textures on Si/SU-8

surface

The steps of 3D negative fingerprint texture fabrication on Si/SU-8

were as follows:

(1) Preparation of Si substrate: 35 mm x 35 mm Si samples were cut

from Si wafers and used as substrates for the preparation of the SU-8

negative fingerprint textured samples.

(2) Cleaning process for Si substrate: The maximum RF power supply of

18W was used with an oxygen plasma cleaning time of 10 minutes.

(3) SU-8 spin-coating process for Si substrate: To create fingerprint

patterns on Si/SU-8, SU-8 (Grade2050) was first spin-coated on Si

substrate at a rotational speed of 500 rpm for 10 seconds with an

SU-8 Si

UV light

(ii) Deposited SU-8 layer onto Si

substrate

(iii) Pre-baking process on Si/SU-

8 layer

(iv) Expose UV light on Si/SU-8

layer

(v) Post-baking process on Si/SU-

8 layer

(i) Clean Si substrate

Chapter 4

92

acceleration of 100 rpm/s, and then at 2000 rpm for 30 seconds with an

acceleration of 300 rpm/s.

(4) Pre-baking process for Si/SU-8: A hot plate was used in the soft

baking (pre-baking) process, which was carried out for 5 minutes at

65oC, and then for 15 minutes at 95

oC.

(5) Patterning process for Si/SU-8: The fingerprint of a subject’s index

finger was used to fabricate a fingerprint pattern on Si/SU-8 after the

soft baking process. A negative pattern that was inverted to the actual

fingerprint was obtained by pressing the index finger on SU-8 polymer.

(6) Curing process for the negative fingerprint textures of Si/SU-8: The

curing process was performed under UV light in a Mask Aligner for 10

– 15 seconds with a power of 215 mJ/cm2.

(7) Post-exposure baking process for the negative fingerprint textures of

Si/SU-8: The post-exposure baking (PEB) process was carried out for

a minimum of 30 minutes at 65oC.

After the patterning and post baking processes, the fabricated samples

were stored in a clean room for 24 hours before conducting any tests. The

steps for the fabrication of the 3D negative fingerprint structure on Si/SU-8

(Si/SU-8/NFP) are shown in Fig. 4.2.

Chapter 4

93

Fig. 4.2. Fabrication procedure for the 3D SU-8 negative fingerprint structure on

Si/SU-8 (Si/SU-8/NFP).

4.2.3. Fabrication of 3D SU-8 positive fingerprint textures on Si/SU-8

surface

In the preparation of 3D positive fingerprint textures on Si/SU-8, the

steps for the fabrication such as preparation, cleaning, spin-coating, and pre-

baking were performed with the same parameters that were discussed in

Section 4.2.2. In the patterning process, a 3D PDMS negative fingerprint

texture was used as a mould to produce a positive fingerprint texture on

Si/SU-8 after the soft baking process.

To obtain a 3D PDMS negative fingerprint pattern, PDMS prepolymer

was mixed with a cross-linking agent, in a ratio of 10:1 in a petri dish, and the

mixture was stirred thoroughly. This mixture was degassed in a vacuum

chamber for 20 minutes to remove bubbles and it was then poured slowly onto

UV light


(ii) Deposit SU-8 layer onto Si

and pre-baking process

(iii) Use an index finger to pattern

Si/SU-8

(iv) Shine UV light on the

negative fingerprint pattern

(v) Fingerprint pattern on Si/SU-8

layer after the post-baking

process

SU-8/NFP Si

Chapter 4

94

the master pieces of Si/SU-8/NFP and PDMS/PFP. Subsequently, it was

placed in a vacuum chamber again for 15 minutes to remove bubbles from the

interface layer.

Fig. 4.3. Fabrication procedures for the 3D SU-8 positive fingerprint pattern on

Si/SU-8 (Si/SU-8/PFP).

The final specimen of PDMS/NFP was obtained after the curing

process was complete, using the oven at 70oC for 5 hours. The detailed

fabrication procedures for PDMS fingerprint textured surfaces are described in

Chapter 7. Then, the curing and post baking process were carried out on

Si/SU-8 surfaces. The fabricated samples were stored in a clean room for 24

Si/SU-8/NFP

SU-8/NFP Si

UV light

Si

SU-8/PFP

UV light

PDMS/NFP

PDMS/PFP PDMS/NFP

PDMS/PFP

Chapter 4

95

hours before conducting any further tests. The fabrication procedures for the

3D positive fingerprint pattern on Si/SU-8 (Si/SU-8/PFP) are shown in Fig.

4.3.

4.2.4. PFPE coating on Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP

In order to improve the wear lives, PFPE (0.2 wt% solution of PFPE Z-

dol 4000 in H-Galden ZV60) was overcoated on the tested samples (Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP) using the dip-coating method. The dipping

and withdrawal speeds of the specimen during dip-coating were kept constant

at 2.1 mm/s, and a dipping duration of 30 seconds was used to obtain 2 – 4 nm

thick PFPE films. These samples were also kept in a clean room for 24 hours

before carrying out any characterization. The final specimens, with a nano-

lubricant layer of PFPE on the flat SU-8 surface, 3D SU-8 negative fingerprint

pattern surface, and 3D SU-8 positive fingerprint pattern surface, were termed

Si/SU-8/PFPE, Si/SU-8/NFP/PFPE, and Si/SU-8/PFP/PFPE, respectively.

Schematic diagrams of the coated layers on these tested specimens are shown

in Fig. 4.4.

4.3. Experimental procedures

After the fabrication process, the optical microscope and optical

profiler (Wyko NT1100) were used, respectively, to capture the wear track of

the tested specimens (Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP) and to

measure the dimensions and roughness of the SU-8 modified surfaces. Water

contact angle measurements were obtained from the fabricated surfaces using

a VCA optima contact angle system (AST product, Inc., USA). The 0.5 µL of

Chapter 4

96

deionized (DI) water was used to measure the water contact angles. The

contact angles of the SU-8 tested specimens are reported as the average of five

independent contact angle measurements. The standard deviations of the data

are within 3o.

Fig. 4.4. The coated layer of PFPE for (a) Si/SU-8/PFPE, (b) Si/SU-8/NFP/PFPE,

and (c) Si/SU-8/PFP/PFPE.

The friction and wear behavior of the SU-8 textured surfaces with the

negative and positive fingerprint patterns were investigated using a

customized ball-on-disc tribometer with a silicon nitride (Si3N4) counterface

ball (diameter = 4 mm, Roughness Ra = 5 nm). The frictional force was

recorded from the deflection of the cantilever that the counterface ball was

attached to. The laser displacement sensors (MTI Instruments Inc., New York,

USA) recorded the measurements of the normal and lateral deflections of the

cantilever, and then the normal load and friction force were obtained by

converting from these displacement measurements with a calibration chart. All

SU-8

PFPE

(a)

Si

SU-8/PFP

PFPE

Si

(c)

SU-8/NFP

PFPE

Si

(b)

Chapter 4

97

of the measurements were performed in air at an ambient temperature and a

relative humidity of 25°C and 58%, respectively. Five repetitive rotational

tests were performed for the friction and wear tests and the averages are

reported here. The wear life was defined as the number of cycles that occurs

before the experiment was stopped when the wear track could be seen on the

tested specimens with a fluctuating coefficient of friction, and/or when the

coefficient of friction was greater than 0.3.

During the friction test, a normal load of 100 mN and a rotational

speed of 2 rpm (linear speed = 0.21 mm/s) were used to obtain friction data

from 20 cycles on the SU-8 untextured and textured surfaces. The coefficients

of friction in the friction tests were recorded using a sampling rate of 10 Hz.

For the wear durability test a nano-lubricant of PFPE was overcoated on the

tested specimens. The long-term wear durability tests were conducted at a

rotational speed of 500 rpm (linear speed = 52.36 mm/s) and a normal load of

100 mN and the number of wear life cycles and the coefficients of friction

were recorded using a sampling rate of 1 Hz. In addition, different normal

loads of 50 mN, 100 mN, 150 mN, 300 mN, and 400 mN were applied to the

tested specimens to investigate the normal load effects. Finally, the recorded

data from 20 cycles for the investigations of different normal load effects on

the tested specimens were observed, with a rotational speed of 2 rpm and a

sampling rate of 10 Hz.

Chapter 4

98

4.4. Results and discussion

4.4.1. Surface morphology

The optical images and dimensions of 3D SU-8 fingerprint textured

surfaces were obtained using an optical microscope and the roughness of the

tested specimens was measured with an optical profiler (Wyko NT1100). The

optical images of the fingerprint textures on Si/SU-8 (Si/SU-8/NFP and Si/SU-

8/PFP) are demonstrated in Fig. 4.5, where the fingerprint ridges and furrows

are shown. Table 4.1 shows the nomenclature of the tested specimens with and

without a PFPE nano-lubricant (Si/SU-8, Si/SU-8/NFP and Si/SU-8/PFP,

Si/SU-8/PFPE, Si/SU-8/NFP/PFPE and Si/SU-8/PFP/PFPE).

Fig. 4.6 shows the schematic diagrams (not to scale) of Si/SU-8/NFP

and Si/SU-8/PFP. The width of the finger ridges of the SU-8 negative

fingerprint and positive fingerprint patterns were 260 µm and 210 µm,

respectively. The dimensions (W), the lengths between the two ridges of the

negative fingerprint and positive fingerprint textures on Si/SU-8, were 226 µm

and 150 µm, respectively. The observed heights of Si/SU-8/NFP and Si/SU-

8/PFP were 1.5 µm and 1.25 µm, respectively, and the measured roughness of

the SU-8 negative fingerprint and positive fingerprint patterns were 814 nm

and 599 nm. The measured dimensions of the tested specimens are shown in

Table 4.2.

Chapter 4

99

Fig. 4.5. An optical image and a 3D image for (a) Si/SU-8/NFP (Ra = 814 nm) and (b)

Si/SU-8/PFP (Ra = 599 nm).

Fig. 4.6. The schematic diagrams (not to scale) of (a) Si/SU-8/NFP and (b) Si/SU-

8/PFP.

(a)

(b)

1.5 µm

Finger ridge

Finger furrow

(b)

d =

1.25

µm

S =

210

µm

L =

360 µm

W =

150 µm

(a)

S = 260

µm

d =

1.5 µm

W=

226 µm

L =

424µm

Chapter 4

100

Table 4.1. Sample nomenclature of the tested surfaces.

Materials Nomenclature

Bare silicon wafer Si

SU-8 polymer spin-coated on silicon wafer Si/SU-8

SU-8 polymer textured with the negative fingerprint

pattern

Si/SU-8/NFP

SU-8 polymer textured with the positive fingerprint

pattern

Si/SU-8/PFP

SU-8 polymer coated with PFPE lubricant Si/SU-8/PFPE

SU-8 polymer textured with the negative fingerprint

pattern and coated with PFPE lubricant

Si/SU-8/NFP/PFPE

SU-8 polymer textured with the positive fingerprint

pattern and coated with PFPE lubricant

Si/SU-8/PFP/PFPE

Table 4.2. The measured dimensions of the tested surfaces (Si, Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP).

Specimens Dimensions

W (µm) L (µm) S (µm) d (µm)

Si - - - -

Si/SU-8 - - - -

Si/SU-8/NFP 226 424 260 1.5

Si/SU-8/PFP 150 360 210 1.25

Table 4.3. The measured contact angles (WCA) of the tested surfaces with and

without PFPE.

Nomenclature WCA (deg)

Si 6

Si/SU-8 82

Si/SU-8/NFP 81

Si/SU-8/PFP 91

Si/SU-8/PFPE 95

Si/SU-8/NFP/PFPE 87

Si/SU-8/PFP/PFPE 97

4.4.2. Water contact angle measurements

Table 4.3 presents the summarized results for the water contact angle

measurements on Si/SU-8, Si/SU-8/NFP, Si/SU-8/PFP, Si/SU-8/PFPE, Si/SU-

8/NFP/PFPE, and Si/SU-8/PFP/PFPE. In addition, the spatial texture densities

for Si/SU-8/NFP and Si/SU-8/PFP are shown in Table 4.4. The water contact

Chapter 4

101

angle measurements were performed with 0.5 µL of deionized water in a VCA

optima contact angle system. The bare silicon (Si) wafer obtained a water

contact angle of 6o. Tay et al. [2011] also observed that bare Si had a water

contact angle of ~10o. The SU-8 coated layer on the Si surface demonstrated a

water contact angle of 82o, which makes its nearly hydrophobic. Moreover, a

near-hydrophobic property with a water contact angle of 81o was obtained for

Si/SU-8/NFP. Based on the results of water contact angle measurements, the

highest water contact angle of 91o was obtained for Si/SU-8/PFP.

Bare Si is known to be highly hydrophilic and it usually attracts water

molecules. SU-8 coated on Si became a hydrophobic surface. Singh et al.

[2011] also observed the near hydrophobic property with a water contact angle

of 83o for SU-8 coated on an Si surface. The negative fingerprint texture on

Si/SU-8 also demonstrated this near-hydrophobic property. Based on these

results, the water contact angle value of Si/SU-8 was not significantly different

from that of Si/SU-8/NFP. A possible reason for the lower water contact angle

for Si/SU-8/NFP is that the water spreads between the ridges and furrows

when the water droplet is placed on the negative fingerprint textured surface.

For Si/SU-8/NFP, the spatial texture density

L

W, which is the ratio of

the width of the recessed cavity W and the pitch length between the centers

of the two projective parts of the textures L , is higher than that of Si/SU-

8/PFP. The ridges are not able to support the water droplet because of the

Chapter 4

102

small height and the wide gap between the ridges. However, the positive

fingerprint pattern on Si/SU-8 observed a highly hydrophobic property (91o).

A possible reason for the higher water contact angle of Si/SU-8/PFP is

that the dimension of the pitch length between two ridges (150 µm) is smaller

than that of Si/SU-8/NFP, and the smaller spatial texture density for Si/SU-

8/PFP supports a high water contact angle. In addition, there is a

heterogeneous wetting regime that occurs on the positive fingerprint pattern.

Based on the Cassie-Baxter model, the water droplet does not wet properly

between the ridges of the positive fingerprint pattern, and the air becomes

trapped between the small grooves. Yoon et al. [2006] observed that the

PMMA patterned surfaces increased the water contact angles and made the

surface more hydrophobic, by comparing with the water contact angle of

PMMA thin film. Singh et al. [2011] also found that surface patterns with a

micro bump on Si/SU-8 (Si/SU-8/MP) increased the observed water contact

angle compared to the unpatterned surfaces.

After PFPE was overcoated on Si/SU-8 (Si/SU-8/PFPE), the value of

the water contact angle significantly increased to 95o, whereas a water contact

angle of 87o was obtained for Si/SU-8/NFP/PFPE. The value of the water

contact angle for the positive fingerprint pattern with PFPE (Si/SU-

8/PFP/PFPE) was 97o and the hydrophobic property was observed for Si/SU-

8/PFPE and Si/SU-8/PFP/PFPE. The water contact angles of Si/SU-8/PFP and

Si/SU-8/PFP/PFPE were 95o and 97

o, respectively; both were in the highly

hydrophobic range. According to these findings on the water contact angles,

Chapter 4

103

the measured water contact angles for the tested specimens are higher when

they are overcoated with PFPE than when they are not. A PFPE modified SU-

8 surface layer provides a higher water contact angle; PFPE lowers surface

tension and renders a surface hydrophobic because of the presence of fluorine

atoms in the PFPE lubricant [Satyanarayana et al. 2005 and 2006, Singh et al.

2007].

Table 4.4. The dimensions and the spatial texture density of Si/SU-8/NFP and Si/SU-

8/PFP.

Specimens Dimensions (µm)

W (µm) L (µm) Density =

L

W

Si/SU-8/NFP 226 424 0.53

Si/SU-8/PFP 150 360 0.42

4.4.3. Friction tests on Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP

Fig. 4.7 shows the representative graphs of the coefficients of friction

for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP from a dry friction test. The

coefficients of friction for the tested specimens were measured with a

customized ball-on-disc tribometer with a normal load of 100 mN and a

rotational speed of 2 rpm. The samples were tested for the first 20 cycles and

the coefficients of friction were recorded.

The observed semi-hydrophobic property of Si/SU-8 demonstrated a

coefficient of friction of 0.2. The modification of the positive fingerprint

pattern on Si/SU-8 (Si/SU-8/PFP) had a coefficient of friction of 0.169 and the

coefficient of friction for Si/SU-8/PFP was slightly lower than that of Si/SU-8.

The lowest coefficient of friction was observed with Si/SU-8/NFP, which had

Chapter 4

104

an average value of 0.08. Most patterned structures with positive asperities

exhibited low coefficients of friction. Based on the current experiments, the

coefficients of friction must have been affected by the development of

different contact areas between the counterface ball and the SU-8 textured

surfaces.

Fig. 4.7. Graph of the coefficients of friction for Si/SU-8, Si/SU-8/NFP, and Si/SU-

8/PFP in dry sliding tests under a normal load of 100 mN and a rotational speed of 2

rpm.

4.4.3.1. Analysis of the effect of contact area versus the spatial density

of the textures on friction

Fig. 4.7 and Table 4.5 show the coefficients of friction and contact

area (contact area is calculated using Hertz’s equation by considering a single

asperity contact and the station condition) for Si/SU-8, Si/SU-8/NFP, and

Si/SU-8/PFP under a normal load of 100 mN. From the frictional test, the flat

SU-8 surface showed a coefficient of friction in the range of 0.2 and the

coefficient of friction for the flat SU-8 surface was higher than that of the

fingerprint textured surfaces. In order to know the effect of surface patterns on

0.2

0.08

0.169

0

0.05

0.1

0.15

0.2

0.25

Si/SU-8 Si/SU-8/NFP Si/SU-8/PFP

Samples

Co

effi

cien

t o

f fr

icti

on


Chapter 4

105

friction, the rate of reduction of the friction coefficient on the patterned

surfaces is determined from,

surfaces patternedfor (COF)

pattern w/oCOF

pattern with COF -pattern w/oCOF friction oft coefficien

of rateReduction

(4.1)

The experimental result showed that the coefficient of friction of Si/SU-8/PFP

was two times higher than that of Si/SU-8/NFP and the reduction rates of 60%

and 16% for the coefficients of friction were observed for Si/SU-8/NFP and

Si/SU-8/PFP, respectively, compared to the flat SU-8 surface.

The contact area between the Si3N4 counterface ball and the SU-8

surface plays a vital role in the change of friction. The calculated contact area

between the counterface ball and the flat SU-8 surface (from Table 4.5) was

large, under a normal load of 100 mN, when compared to the fingerprint

textured specimens. Assuming an elastic deformation between the Si3N4

counterface ball and the SU-8 surfaces, the radius of the elastic contact (a),

according to Hertz’s equation [Hertz 1882], is given in Equation 2.13 as:

3

1

*

*

4

3

E

RFa n

where: 2

2

2

1

2

1

*

111

EEE

The effective Young’s modulus ( 1E ) and the poisson’s ratio ( 1 ) of Si3N4 are

310 GPa and 0.22, respectively, as provided by the supplier. The 2E and 2

Chapter 4

106

of SU-8 were obtained from a reference as 5.25 GPa and 0.22, respectively

[Halhouli et al. 2008]. The frictional force ( F ) is directly proportional to the

real contact area, 2aAr based on the frictional law by Bowden and Tabor

[Bowden and Tabor 1986], which is described in Equation 2.10:

rfriction AF .

The investigations of the calculated contact area showed that the ridges

of the fingerprint do not permit a large contact area because this requires

greater elastic deformation of the ridges. Si/SU-8/NFP and Si/SU-8/PFP,

which have rounded ridges, registered lower frictional forces, due to lower

contact area, compared to the flat SU-8 surface. Rounded projected textures

have less contact area because the elastic deformation of the ridges

counterbalances the normal load and the adhesive forces; this is impossible for

the flat top untextured SU-8 surfaces. However, the calculation of contact area

on a single asperity for Si/SU-8/PFP was less than that of Si/SU-8/NFP.

Although the contact area of a single asperity for Si/SU-8/PFP is small, the

overall real contact area may be large due to the small width and length (~150

µm) between the two ridges of Si/SU-8/PFP. The calculations used to

determine the contact area, using Hertz’s equation, on a single asperity are

provided in detail in Chapter 8.

Therefore, the coefficient of friction for the textured surface

depends on the contact area, which is related to the spatial texture density. The

observed spatial texture densities for Si/SU-8/NFP and Si/SU-8/PFP were 0.53

and 0.42, respectively (see Table 4.4). Based on the calculated data of the

Chapter 4

107

spatial density, the spatial texture density for Si/SU-8/PFP is smaller than that

of Si/SU-8/NFP. When the spatial texture density for Si/SU-8/PFP is small,

the real contact area may be large on multiple asperities; therefore, the

adhesion force between the counterface ball and Si/SU-8/PFP is high, which

increases the coefficient of friction. As the spatial texture density for Si/SU-

8/NFP is increased to 0.53, the real contact area may be small on multiple

asperities, decreasing the coefficient of friction. Therefore, a small coefficient

of friction for Si/SU-8/NFP was observed with increasing spatial texture

density. In conclusion, spatial texture density of the pattern geometry is

required to change the real contact area and improve friction and wear

behavior. In addition, the roughness on Si/SU-8/NFP (814 nm) reduces the

contact area and the friction.

Table 4.5. The coefficients of friction and contact area (contact area is calculated

using Hertz’s equation by considering a single asperity contact and the static

condition) for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP under a normal load of 100

mN.

Specimens Contact area

[x10-9

m2]

Contact

pressure

(MPa)

Coefficient of

friction ( )

Si/SU-8 2.87 6.96 0.2

Si/SU-8/NFP 2.16 46.3 0.08

Si/SU-8/PFP 2.11 47.39 0.169

4.4.3.2. Analysis of wear mechanisms for the friction tests

Fig. 4.8 shows the optical images of wear track from the samples and

debris on the counterface ball for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP,

from dry friction tests with a normal load of 100 mN and a rotational speed of

2 rpm. The wear track on Si/SU-8 demonstrates that wear took place in less

Chapter 4

108

than 20 sliding cycles and a visible amount of wear debris was transferred to

the counterface ball after the friction test. Adhesive wear behaviors were

occurred and plastic deformation led to the transfer materials from the worn

parts of Si/SU-8.

Fig. 4.8. The optical images of debris on the counterface ball and wear track on the

samples for (a) Si/SU-8, (b) Si/SU-8/NFP, and (c) Si/SU-8/PFP.

However, no wear track was visible on the negative and positive

fingerprint patterns of Si/SU-8 surfaces and there was no wear debris attached

to the counterface ball. Improved wear behavior was observed on the negative

fingerprint and positive fingerprint patterns in dry sliding tests, because of the

presence of protrusions from the fingerprint patterns. Due to the reduction of

100 µm 100 µm

Wear

track

100 µm 100 µm

(b)

(a)

(c)

Chapter 4

109

real contact area on the negative and positive fingerprint patterns on Si/SU-8,

the adhesion force was reduced leading to improved wear behavior.

4.4.4. Wear tests on Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP (with and

without PFPE)

Fig. 4.9 shows the coefficients of friction for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP, with and without PFPE nano-lubricant, in the investigation

of wear tests. In the wear studies, a normal load of 100 mN and a high

rotational speed of 500 rpm were applied to the tested specimens. The

sampling rate of 1 Hz was selected to record data for the coefficients of

friction. Further, a 0.2 wt% solution of PFPE Z-dol 4000 in H-Galden ZV60

was overcoated on the tested specimens in order to improve the wear

resistance as SU-8 alone was poor at wear resistance. The dip-coating method

was used with a dipping and withdrawal speed of 2.1 mm/s and a dipping

duration of 30 seconds to obtain 2 – 4 nm thick PFPE films.

In the wear tests, the flat SU-8 surface without PFPE nano-lubricant

(Si/SU-8) provided high value of coefficient of friction (~0.18) and

demonstrated adhesive wear mechanisms against the Si3N4 counterface ball. It

was observed that Si/SU-8 was worn out after approximately 300 cycles with

large fluctuations in the coefficient of friction that reached to 0.18. After

patterning the negative fingerprint pattern on Si/SU-8, the coefficient of

friction was reduced to 0.07 when tested without PFPE nano-lubricant. The

high speed (500 rpm) experiments showed the highest coefficient of friction of

0.19 on Si/SU-8/PFP without PFPE nano-lubricant. There was no significant

Chapter 4

110

difference in the coefficients of friction between Si/SU-8 and Si/SU-8/PFP. At

relatively high rotational speeds, the coefficient of friction for Si/SU-8/PFP

was high due to the small spatial texture density which increased the real

contact area and high shear stress on multiple asperities. However, in the wear

studies, the reduction rate of 61% in the coefficient of friction was observed

for Si/SU-8/NFP compared to the flat SU-8 surface.

The SU-8 spin-coated specimen with PFPE nano-lubricant showed the

coefficient of friction of 0.08 (Fig. 4.9). It was obvious that the lowest

coefficient of friction (~0.05) was obtained at the Si/SU-8/NFP/PFPE system.

However, Si/SU-8/PFP/PFPE had a coefficient of friction greater than that of

Si/SU-8/NFP/PFPE and Si/SU-8/PFPE. It is known that the contact area

influences the coefficient of friction and there is greater contact area between

the Si3N4 counterface ball the SU-8 textured surfaces when the spatial texture

density is small. It was also observed that the PFPE lubricated specimens had

low coefficients of friction, compared to the unlubricated specimens.

The wear life cycles of the tested specimens (Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP), with and without PFPE nano-lubricant, under a normal load

of 100 mN and a rotational speed of 500 rpm are shown in Fig. 4.10. The

highest wear lives of 2,000 cycles were registered for Si/SU-8/NFP without

PFPE nano-lubricant and the wear lives for Si/SU-8/PFP were 1,800 cycles.

The experimental results showed that Si/SU-8/NFP had lower coefficient of

friction and better wear durability than Si/SU-8 and Si/SU-8/PFP when no

lubricant was applied to the tested specimens.

Chapter 4

111

When PFPE nano-lubricant was applied to Si/SU-8, Si/SU-8/NFP, and

Si/SU-8/PFP, the friction was reduced and the wear lives were significantly

improved. Based on the experimental results, the wear durability of Si/SU-

8/PFPE increased to 5,000 cycles and the coefficient of friction reduced to

0.08. The coefficient of friction for Si/SU-8 decreased from 0.18 to 0.08 after

coating the PFPE nano-lubricant on Si/SU-8. Investigating friction and wear

life cycles for Si/SU-8/PFP/PFPE demonstrated a coefficient of friction of

0.16 and the wear lives significantly improved to 30,000 cycles. The highest

wear lives of 60,000 cycles were registered for Si/SU-8/NFP/PFPE and the

coefficient of friction of Si/SU-8/NFP/PFPE was lower than that of Si/SU-

8/NFP. The wear lives of Si/SU-8/NFP/PFPE were nearly thirty times higher

than Si/SU-8/NFP.

Fig. 4.9. The coefficients of friction for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP with

and without PFPE nano-lubricant. under a normal load of 100 mN and a rotational

speed of 500 rpm.

Tay et al. [2011] found that the wear durability of PFPE coated SU-8

micro-dots on Si substrate increased to more than 100,000 cycles at a pitch

0.18

0.07

0.19

0.08

0.05

0.16

0

0.05

0.1

0.15

0.2

0.25

0.3


Samples

Coef

fici

ent

of

fric

tion

Without PFPE

With PFPE

Chapter 4

112

length of 450 µm. According to the experimental results, the PFPE lubricated

specimens had enhanced wear durability and low coefficients of friction when

compared to the unlubricated specimens; due to their lubricity, the lubricant

worked considerably better with surface texturing from fingerprints. The nano-

lubricant of PFPE can reduce the shear stress, leading to a lower coefficient of

friction [Lui et al. 2003 and Sayanarayana et al. 2006]. The presence of

fluorine atoms in PFPE and the strong bonding of molecules can provide

better wear performance on the SU-8 textured surface, particularly with the

negative fingerprint patterns because of the optimized contact area and the

texture geometry.

Fig. 4.10. The wear life cycles of different specimens (Si/SU-8, Si/SU-8/NFP, and

Si/SU-8/PFP) with and without PFPE nano-lubricant under a normal load of 100 mN

and a rotational speed of 500 rpm.

4.4.4.1. Analysis of wear mechanisms for the wear tests

Fig. 4.11 shows the optical images of wear debris on the Si3N4

counterface ball and wear tracks on the tested specimens with and without

PFPE nano-lubricant, such as Si/SU-8, Si/SU-8/PFPE, Si/SU-8/NFP, Si/SU-

300

2,000 1,800

5,000

60,00030,000

1

10

100

1000

10000

100000


Samples

Wea

r L

ife

Cy

cles

(L

og S

cale

)

Without PFPE

With PFPE

Chapter 4

113

8/NFP/PFPE, Si/SU-8/PFP, and Si/SU-8/PFP/PFPE, under a normal load of

100 mN and a high rotational speed of 500 rpm. The wear durability of Si/SU-

8 was short and the wear track was clearly observed on Si/SU-8 under the

optical microscope; the wear debris was transferred to the counterface ball.

However, after overcoating PFPE nano-lubricant on Si/SU-8, a small wear

trace occurred. Fig. 4.11 shows that less wear debris was transferred to the

counterface ball for Si/SU-8/PFPE than for Si/SU-8. Moreover, the coefficient

of friction remained in the range of 0.1; this value is low for a wear test, but a

trace of the wear track was observed on the negative fingerprint textured

surface without the PFPE nano-lubricant layer.

A small amount of wear debris accumulated on the counterface ball

and the wear track showed deformation marks on Si/SU-8/NFP at 2,000 cycles.

The width of the wear track was markedly reduced and the wear durability

was significantly increased to 60,000 cycles after coating the PFPE nano-

lubricant on Si/SU-8/NFP. Experiments showed that in dry conditions, the

wear durability on Si/SU-8/PFP was increased to 1,800 cycles and the wear

track was clearly observed on Si/SU-8/PFP. Coating of PFPE on Si/SU-8/PFP

led to a reduction in the wear track and less wear debris was accumulated on

the counterface ball. Significant improvements in wear durability and

reduction in wear debris occurred after overcoating PFPE nano-lubricant on

the textured surfaces.

Chapter 4

114

Fig. 4.11. The optical images of wear debris on the Si3N4 counterface ball and wear

track on the samples of (a) Si/SU-8 without PFPE (300 cycles), (b) Si/SU-8 with

PFPE (5,000 cycles), (c) Si/SU-8/NFP without PFPE (2,000 cycles), (d) Si/SU-8/NFP

with PFPE (60,000 cycles), (e) Si/SU-8/PFP without PFPE (1,800 cycles), and (f)

Si/SU-8/PFP with PFPE (30,000 cycles).

(a) Si/SU-8 (300 cycles)

(b) Si/SU-8/PFPE (5,000 cycles)

100 µm 100 µm

Wear track

100 µm 100 µm

Wear track

(c) Si/SU-8/NFP (2,000 cycles)

100 µm Wear track

(d) Si/SU-8/NFP/PFPE (60,000 cycles)

100 µm 100 µm

Wear track

Wear track

(e) Si/SU-8/PFP (1,800 cycles)

(f) Si/SU-8/PFP/PFPE (30,000 cycles)

100 µm

Wear track

Chapter 4

115

4.4.5. Effects of different normal loads on friction for Si/SU-8, Si/SU-

8/NFP, and Si/SU-8/PFP

The resultant data for the coefficients of friction of the tested

specimens (Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP) are shown in Fig. 4.12.

Different normal loads of 50 mN, 100 mN, 150 mN, 300 mN, and 400 mN

were applied to the tested specimens at a fixed disc rotational speed of 2 rpm.

For Si/SU-8, the coefficient of friction was slightly reduced from 0.24

to 0.19 when the normal load was increased from 50 mN to 100 mN.

Thereafter, the coefficients of friction stayed nearly constant for the remaining

normal loads (i.e. for 150 mN, 300 mN, and 400 mN). Si/SU-8/PFP showed a

coefficient of friction of 0.2 at a normal load of 50 mN. Moreover, the

coefficient of friction for Si/SU-8/PFP rapidly decreased from 0.2 at 50 mN to

0.09 at 150 mN, and then it sharply increased again after a normal load of 150

mN.

The lowest coefficient of friction of 0.09 was observed at 150 mN for

Si/SU-8/PFP; this specimen showed the highest coefficient of friction of 0.28

at 400 mN. The coefficients of friction for Si/SU-8/PFP were higher than that

of Si/SU-8/NFP for all normal loads. Mixed behavior (decreasing and

increasing trends) was observed for Si/SU-8/NFP; the coefficient of friction

first showed a decreasing trend between the normal loads of 50 mN to 150 mN,

remained constant until 300 mN, and then increased again between 300 mN

and 400 mN. Si/SU-8/NFP had a coefficient of friction of 0.11 at 50 mN,

which decreased to a value of 0.05 when the normal load was increased to 150

Chapter 4

116

mN, and then remained at this value until a normal load of 300 mN was

applied. Subsequently, it increased to 0.12 when the normal load was

increased to 400 mN.

According to Amonton’s law, frictional force is proportional to applied

normal load. Based on the experiments, the coefficients of friction for Si/SU-8,

Si/SU-8/NFP, and Si/SU-8/PFP decreased when the normal loads increased, in

the range from 50 mN – 150 mN. However, the proportionality between the

frictional force and normal force for Si/SU-8/PFP and Si/SU-8/NFP dissolved

and Amonton’s First Law was not supported before the normal loads of 150

mN and 300 mN. Based on experimental results, Myshkin et al. [Myshkin et

al. 2005] stated that the friction forces of different polymeric materials, such

as PTFE, PE, and nylon, generally increased when the normal loads increased

(from 10 N to 100 N); however, frictional forces were observed to be nearly

constant at normal loads between 10 N and 40 N. Experimental results from

Shooter and Tabor [Shooter and Tabor et al. 1952] showed that the contact

area was proportional to the applied normal load for a steel slider against

polymers (PTFE, PE, PMMA, PVC, and nylon), and their frictional forces

(and coefficients of friction) significantly increased with increasing normal

loads.

Table 4.4 shows the summarized results of the contact area that was

calculated for a single asperity using the Hertz equation and the corresponding

coefficients of friction for different normal loads for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP. The detailed calculation is provided in Chapter 8. Based on

Chapter 4

117

the Hertz contact equation, the contact area of the negative and positive

fingerprint patterns were less than that of the flat surface for all normal loads

used. When increasing the applied normal load within the range of 50 – 400

mN, the contact area of the textured surfaces increased, as the textured

surfaces deformed elastically or plastically. The coefficients of friction for

Si/SU-8 were significantly greater than that of the negative fingerprint and

positive fingerprint patterns, due to the larger contact area of the flat SU-8

surface.

The adhesion forces between the counterface ball and the flat SU-8

surface increased because of the large contact area that increased the

coefficient of friction. The calculated contact area for Si/SU-8 increased with

increasing normal loads. Due to the small contact area between the Si3N4

counterface ball and Si/SU-8/NFP, low coefficients of friction were registered

for all normal loads. For the positive fingerprint textured surface, the

coefficients of friction increased after 150 mN and these coefficients of

friction were greater than that of a flat surface behind a normal load of 250

mN. Moreover, the occurrence of fracture and plastic deformation at the

contact region and small wear particles were found on the surface of Si/SU-

8/NFP and Si/SU-8/PFP. The fingerprint textured surfaces can reduce the

friction and increase the load carrying capacity compared to the flat surface. In

conclusion, the coefficients of friction are greatly affected by applied normal

loads and the geometry of the controlled contact area.

Chapter 4

118

Fig. 4.12. The resulting coefficients of friction versus the different normal loads for

Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP.



condition) versus the applied normal load for Si/SU-8, Si/SU-8/NFP, and Si/SU-

8/PFP.

Load

(mN)


Contact

area

[x10-9

m2]

Coefficient

of friction

( )

Contact

area

[x10-9

m2]

Coefficient

of friction

( )

Contact

area

[x10-9

m2]

Coefficient

of friction

( )

50 1.36 0.2 1.36 0.11 1.33 0.2

100 2.16 0.198 2.16 0.08 2.11 0.16

150 2.83 0.175 2.83 0.05 2.78 0.09

300 4.5 0.186 4.5 0.05 4.39 0.26

400 5.45 0.169 5.45 0.12 5.32 0.28

4.5. Summary

In Chapter 4, the friction and wear behaviors of the SU-8 untextured

and textured (3D negative fingerprint and 3D positive fingerprint) surfaces

were explored. Based on the fabrication process, the protruding parts of the

textures with rounded corners, such as the negative fingerprint and positive

fingerprint patterns, were well developed using an index finger and a PDMS

negative fingerprint mould. All experimental tests were conducted using a

0

0.1

0.2

0.3

0 100 200 300 400 500

Load (mN)

Coef

fici

ent

of

fric

tion

Si/SU-8Si/SU-8/NFPSi/SU-8/PFP

Chapter 4

119

customized ball-on-disc tribometer with a Si3N4 counterface ball of 4 mm in

diameter as the counterface. The frictional tests were performed with a normal

load of 100 mN and a slow rotational speed of 2 rpm. The wear tests were then

carried out with a normal load of 100 mN and a high rotational speed of 500

rpm, and the tested specimens were investigated with and without PFPE nano-

lubricant.

In the dry rotational tests, Si/SU-8/NFP showed the lowest coefficient

of friction (~0.08), compared to that of Si/SU-8 (~0.2) and Si/SU-8/PFP

(~0.169). The coefficient of friction for Si/SU-8/NFP was nearly 2.5 times

lower than that of Si/SU-8. Although the contact area of a single asperity

(based on Hertz’s contact theory) for Si/SU-8/NFP was larger than that of

Si/SU-8/PFP, the large spatial texture density of Si/SU-8/NFP decreased the

real contact area and the friction. The small spatial texture density for Si/SU-

8/PFP provided a high coefficient of friction. After applying the textures on

Si/SU-8, a coefficient of friction reduction rate of 60 % was observed for

Si/SU-8/NFP and the reduction rate for Si/SU-8/PFP was 15%. Less wear

particles accumulated on the SU-8 textured surface compared to the flat SU-8

surface. With an increase in the normal loads, the texture surfaces supported

the increased load carrying capacity. The coefficients of friction for Si/SU-

8/NFP decreased within the range of 50 mN to 300 mN; however, the

coefficients of friction of Si/SU-8/PFP increased after the normal load of 150

mN. Mixed behavior (a decreasing and increasing trend) was observed for

both Si/SU-8/NFP and Si/SU-8/PFP.

Chapter 4

120

In addition, the coefficient of friction for Si/SU-8/NFP was low for all

normal loads compared to that of Si/SU-8 and Si/SU-8/PFP. In the

investigations of wear behavior, the SU-8 textured surfaces demonstrated

better friction and wear behavior than the flat SU-8 surface and the wear

behavior of the SU-8 textured surfaces was significantly improved. Moreover,

the SU-8 textured surface with the nano-lubricant PFPE layer increased the

wear life cycles markedly. Among the wear durability tests, the highest wear

life cycles (~60,000 cycles) were observed for Si/SU-8/NFP/PFPE.

In conclusion, the negative fingerprint textured surface on SU-8 shows

promise for excellent tribological performances. Moreover, the negative

fingerprint textured surface, which is a cost-effective way to produce a

rounder corner pattern on a surface such as SU-8, can serve to reduce friction

and increase wear life cycles for future MEMS/Bio-MEMS devices such as

portable or implantable biomedical mircofluidic devices in biomedical

application.

Chapter 5

121

Chapter 5

Fabrication Methods and Tribological Analysis of the

Micro-dot Pattern and the Honeycomb Pattern on

Si/SU-8 (with and without PFPE nano-lubricant)

Different types of surface textures were investigated to determine their

friction and wear behaviors. This chapter presents the analysis of the

tribological characterization of negative asperity (honeycomb patterns) and

positive asperity (micro-dot patterns) on SU-8 surfaces. Their friction and

wear behaviors are compared to ensure that the positive asperity of surface

textures improves tribological behavior. A polymer jet printer was used to

fabricate the micro-dot patterns on Si/SU-8. Investigations of surface

characterization were carried out; PFPE nano-lubricant was used in these

investigations to enhance the wear durability of the surfaces. In addition, the

positive asperity of surface textures was largely attributed to their load

carrying capacities.

Chapter 5

122


The improvements and advantages of the tribological characterizations

of the SU-8 positive asperity textured surfaces (negative fingerprint and

positive fingerprint) on Si/SU-8 were presented in Chapter 4. The effects of

PFPE nano-lubricant on the friction and wear performances of the SU-8

negative fingerprint and positive fingerprint surfaces were also mentioned in

Chapter 4. Though superior tribological properties on the positive asperity of

the SU-8 textured surfaces that are related to the spatial texture density have

been observed, an exact relationship between the spatial texture density and

the coefficient of friction on the SU-8 textured surfaces is required to improve

friction. In this Chapter, the investigations of the negative asperities on the

SU-8 textured surface (honeycomb patterns) are examined. In order to analyze

the tribological behavior of the positive asperity of different patterns, the

micro-dot patterns were developed on Si/SU-8 using a polymer jet printing

method.

In the investigations described in this chapter, Si material was used as

a substrate and SU-8 material was used as a protective coated layer on the Si

substrate. Two different surface textures with honeycomb patterns and micro-

dot patterns were attached to Si/SU-8 surfaces. During the preparation of

samples, the honeycomb patterns were transferred to Si/SU-8 through a glass

mask with the honeycomb patterns, and the micro-dot patterns were obtained

using a polymer jet printer. The fabricated pitch lengths of the micro-dot

patterns were 200 µm, 300 µm, 400 µm, 500 µm, and 600 µm. Furthermore,

Chapter 5

123

the frictional investigations of the honeycomb and micro-dot patterns on

Si/SU-8 were conducted using a fixed normal load of 100 mN and a low

rotational speed of 2 rpm. Finally, the frictional tests were conducted using

different normal loads, to understand the effects of these loads and determine

the critical loads for the failure of the SU-8 textured surfaces.

The correlation between the frictional behavior and the different

normal loads was investigated using a ball-on-disc configuration. Finally, the

honeycomb patterns and the micro-dot patterns on the Si/SU-8 surfaces were

investigated with a rotational speed of 500 rpm, in order to determine the wear

durability with a high rotational speed on the negative asperity (honeycomb

pattern) and positive asperity (micro-dot pattern) of the textured surfaces.

Perfluoropolyether (PFPE) was overcoated on the SU-8 textured surfaces to

reduce the friction and shear stress and to enhance the wear durability of the

SU-8 textured surfaces.

5.2. Sample preparations

5.2.1. Fabrication of 3D SU-8 micro-dot textures on Si/SU-8 surfaces

In the fabrication process for the SU-8 micro-dot patterns, two grades

of SU-8 (Grade 2050 and Grade 2000.5) were used. The steps for fabricating

the SU-8 micro-dot textured surfaces were as follows:

(1) Preparation of Si substrate: For the preparation of SU-8 micro-dot

patterned samples, 35 mm x 35 mm Si samples were used as the

substrates.

Chapter 5

124

(2) Cleaning process for Si substrate: An oxygen plasma cleaner (Harrick

Plasma, PDC-32G) was used for 10 minutes with the maximum RF

power supply of 18 W to remove any contaminants.

(3) SU-8 spin-coating process for Si substrate: To fabricate micro-dot

patterns on Si/SU-8 substrate, Grade 2050 SU-8 polymer was first

spin-coated onto Si substrate at a rotational speed 500 rpm for 10

seconds with an acceleration of 100 rpm/s, and then at 2000 rpm for 30

seconds with an acceleration of 300 rpm/s. An SU-8 layer with a

thickness of 80 µm was obtained using a spin-coater (P6700 Specialty

Coating Systems, Indianapolis, IN).

(4) Pre-baking process for Si/SU-8: A soft baking (pre-baking) process

was performed with a hot plate for 5 minutes at 65oC, followed by 15

minutes at 95oC.

(5) Curing process for Si/SU-8: The curing process was performed on the

flat Si/SU-8 substrate, using UV light in a Mask Aligner for 10 – 15

seconds with a power of 215 mJ/cm2.

(6) Post-exposure baking process for Si/SU-8: After the curing process,

the post-exposure baking (PEB) process was performed for at least 30

minutes at 65oC.

(7) Micro-dot patterning process for Si/SU-8: Grade 2000.5 SU-8 with

2.49 cSt of viscosity was used in the fabrication of the micro-dot

patterns on Si/SU-8. A polymer jet printing (PJP) method (Jetlab 4

Chapter 5

125

polymer jet printer machine, from MicroFab Technologies Inc., Plano,

TX) was used to fabricate and fine tune micro-dot patterns, and SU-8

polymer was dispensed from the nozzle of PJP with a diameter of 20

µm. The desired micro-dot patterns (a rectangular array of 40 dots by

40 dots) were printed out using the parameters of waveform, such as

3.0 µs of rise time, 40 µs of dwell time at 12.0 V, 3.0 µs of fall time,

90 µs of echo dwell at 25 V, and 3.0 µs of final rise time. The desired

pitch lengths of micro-dot patterns on Si/SU-8 were 200 µm, 300 µm,

400 µm, 500 µm, and 600 µm.

(8) Pre-baking, curing, and post-baking process for the micro-dot

patterns on Si/SU-8: Once the patterning process for the micro-dot

patterns was complete, the pre-baking, exposure to UV light, and post-

baking processes were carried out as described in the steps 4, 5, and 6.

The final specimens for different pitch lengths of micro-dot patterns

are referred to as: Si/SU-8/MD (200 µm), Si/SU-8/MD (300 µm), Si/SU-

8/MD (400 µm), Si/SU-8/MD (500 µm), and Si/SU-8/MD (600 µm) and the

fabrication steps for 3D micro-dot patterns on Si/SU-8 are shown in Fig. 5.1.

5.2.2. Fabrication of 3D honeycomb patterns on Si/SU-8 surfaces

The steps for fabricating 3D honeycomb textures on Si/SU-8 surfaces

were as follows:

(1) Preparation of Si substrate: A four inch Si wafer was used as the

substrate to fabricate the honeycomb patterns.

Chapter 5

126

(2) Cleaning process for Si substrate: A cleaning process was carried

out on the Si substrates using an oxygen plasma cleaner (Harrick

Plasma, PDC-32G) to remove any contaminants. The oxygen plasma

cleaning was conducted for 10 minutes with the maximum RF power

supply of 18W.

(3) SU-8 spin-coating process for Si substrate: Grade 2050 SU-8 with

4,500 cSt of viscosity was coated onto the Si substrate to obtain a

thickness of 80 µm. To fabricate the SU-8 layer, the spin-coater

(P6700 Specialty Coating Systems, Indianapolis, IN) was used at a

rotational speed of 500 rpm for 10 seconds with an acceleration of 100

rpm/s, and then at 2000 rpm for 30 seconds with an acceleration of 300

rpm/s.

(4) Pre-baking process for Si/SU-8: Soft-baking was performed for 5

minutes at 65oC, and then again for 15 minutes at 95

oC to remove the

solvent from the coated layer.

(5) Honeycomb patterning process on Si/SU-8: UV light was used with

an exposure dose of 20 – 25 seconds and a power of 210 mJ/cm2. The

honeycomb patterns on Si/SU-8 were then created by passing UV light

through a glass mask with the honeycomb patterns.

Chapter 5

127

Fig. 5.1. Fabrication procedures for 3D micro-dot patterns on Si/SU-8 surfaces

(using PJP machine).

UV light


(ii) Deposited SU-8 (Grade 2050)

layer on Si and pre-baking

process

(iii) Shine UV light on the flat

SU-8 layer

(iv) Post-baking process

(v) Patterning micro-dots

(with Grade 2000.5 of

SU-8) on Si/SU-8 using

PJP machine

(vi) Pre-baking process

(vii) Shine UV light on the

SU-8 micro-dot patterns

UV light

(viii) Post-baking process on

the SU-8 micro-dot

patterns

Pitch length

dimension, P

Chapter 5

128

(6) Post-exposure baking process for the honeycomb textures of Si/SU-8:

Post-baking was performed for 30 minutes at 65oC.

(7) Developing process for the honeycomb textures of Si/SU-8: The

fabricated samples were then immersed in a solvent of SU-8 developer

for 15 - 20 minutes to remove the unwanted SU-8 polymer from the

honeycomb patterns.

(8) Rinse and dry process for the honeycomb textures of Si/SU-8: After

using an SU-8 developer, a rinsing process was performed with

acetone and then nitrogen gas was used to dry the surface.

The procedure for fabricating the 3D honeycomb pattern on Si/SU-8

(Si/SU-8/HC) is shown in Fig. 5.2.

Fig. 5.2. Fabrication procedure for 3D honeycomb patterns on Si/SU- surface.


(ii) Deposit SU-8 layer onto Si and

pre-baking process

(iii) Expose to UV light using the chrome mask

with the honeycomb structure to pattern on

Si/SU-8

UV light Chrome mask with the honeycomb structure

(iv) Post-baking process on the SU-8

honeycomb patterns

(v) Honeycomb pattern after developing with

SU-8 developer, cleaning with acetone, and

drying with nitrogen gas

Chapter 5

129

5.2.3. PFPE coating on 3D SU-8 micro-dot and honeycomb patterns

The fabricated specimens were overcoated with a nano lubricant of

PFPE (0.2 wt% solution of PFPE Z-dol 4000 in H-Galden ZV60). A thin film

of a PFPE layer (thickness of 2 – 4 nm) was obtained using the dip-coating

method with a dipping and withdrawal speed of 2.1 mm/s and a dipping

duration of 30 seconds. The final specimens were kept in a clean room for 24

hours for further investigation of the tested specimens, Si/SU-8/HC/PFPE and

Si/SU-8/MD/PFPE. A schematic diagram of the coated layers on the

fabricated specimens is shown in Fig. 5.3.

Fig. 5.3. The coated layer of PFPE for (a) Si/SU-8/MD/PFPE and (b) Si/SU-

8/HC/PFPE.


The optical images were captured with an optical microscope and

measurements for the dimensions and roughness of the modified textured

surfaces were obtained with a Wyko NT1100 optical profiler. To investigate

the surface characterization of Si/SU-8/MD and Si/SU-8/HC, water contact

angle measurements were performed using 0.5 µL of deionized (DI) water in a

VCA optima contact angle system (AST product, Inc., USA). The averages of

five independent contact angle measurements for Si/SU-8/MD and Si/SU-

8/HC are reported in this Chapter and the standard deviations were within 3o.

PFPE

SU-8/MD

Si

(a) (b)

PFPE

SU-8/HC

Si

Chapter 5

130

The frictional tests for 3D micro-dot patterns and 3D honeycomb

patterns were carried out with a normal load of 100 mN and a rotational speed

of 2 rpm (linear speed = 0.21 mm/s), in a customized ball-on-disc tribometer

that was attached to a 4 mm diameter silicon nitride (Si3N4) counterface ball.

A sampling rate of 10 Hz was used to record the friction data of 20 cycles. The

experiments were stopped when the coefficient of friction was greater than 0.3

and a wear track was observed on the tested specimen. Moreover, the wear

durability was investigated on Si/SU-8/MD and Si/SU-8/HC with and without

the nano-lubricant of PFPE layer. A rotational speed of 500 rpm (linear speed

= 52.36 mm/s) and a normal load of 100 mN were applied to investigate the

wear life cycles and the coefficients of friction. In the wear tests, a sampling

rate of 1 Hz was used to record the coefficient of friction and to verify the

wear life cycles of the tested specimens. The wear durability was recorded as

the number of cycles that occurred before failure of wear was observed on the

samples. Moreover, PFPE nano-lubricant was used as a cover layer on the

textured specimens to enhance the long-term wear life cycles.

To investigate the effects of different normal loads on Si/SU-8/MD and

Si/SU-8/HC, normal loads of 50 mN, 100 mN, 150 mN, 300 mN, and 400 mN

were used. A rotational speed of 2 rpm and a sample rate of 10 Hz were used

in these experiments. For all of the friction and wear tests, the average of five

repetitive sliding tests is reported here. These measurements were carried out

in air at an ambient temperature of 25oC and a relative humidity of 58%. The

tribological tests were conducted in a class-100 clean booth room.

Chapter 5

131



The optical image and 3D images of Si/SU-8/MD (pitch length of 400

µm) and Si/SU-8/HC that were captured by the optical microscope and optical

profiler are shown in Fig. 5.4. The schematic diagrams (not to scale) of Si/SU-

8/HC and Si/SU-8/MD are shown in Fig. 5.5. The pore of the honeycomb

structure (a) was 230 µm. The widths between the two honeycomb structures

(b) and (c) were 240 µm and 500 µm, respectively.

Fig. 5.4. An optical image and a 3D image for (a) Si/SU-8/HC (Ra = 28 µm) and (b)

Si/SU-8/MD (Ra = 92 nm).

(a)

(b)

W

L

W = 200 µm

L = 440 µm

a = 440 µm

Chapter 5

132

Fig. 5.5. Schematic diagrams (not to scale) of (a) Si/SU-8/HC and (b) Si/SU-8/MD.

The width (W) of Si/SU-8-HC was 200 µm and the pitch length

between two honeycomb patterns was 440 µm. The depth (d) of the

honeycomb textured surface on Si/SU-8 was 80 µm. For the SU-8 micro-dot

patterns, the average diameter of the micro-dots obtained was 100 µm and the

average height of the dots was 0.8 µm. The SU-8 micro-dot pitch lengths were

set at 200 µm, 300 µm, 400 µm, 500 µm, and 600 µm. The values of

roughness for Si/SU-8/HC and Si/SU-8/MD were 28 µm and 92 nm,

respectively, when measured with a Wyko NT1100 optical profiler. Table 5.1

shows the measured dimensions of Si/SU-8/MD and Si/SU-8/HC.

Table 5.1. The measured dimensions of the tested surfaces (Si/SU-8/MD and Si/SU-

8/HC).

Specimens Dimensions

W (µm) L (µm) S (µm) d (µm)

Si/SU-8/MD(200 µm) 100 200 100

0.8

Si/SU-8/MD(300 µm) 200 300 100

Si/SU-8/MD(400 µm) 300 400 100

Si/SU-8/MD(500 µm) 400 500 100

Si/SU-8/MD(600 µm) 500 600 100

Si/SU-8/HC 200 440 - 80

d =

0.8 µm

S =100 µm

L

d = 80 µm

c = 500 µm b = 240 µm

a = 230 µm

Chapter 5

133

5.4.2. Water contact angle measurements

The investigations of water contact angle measurements were carried

out with a VCA optima contact angle system (AST product, Inc., USA). 0.5

µL of deionized water was dropped on the tested specimens to investigate the

surface wettabilites of Si/SU-8/HC and Si/SU-8/MD (different pitch lengths).

The nomenclature and the findings from the water contact angle measurements

on the SU-8 honeycomb pattern and the SU-8 micro-dot pattern with different

pitch lengths are summarized in Table 5.2. Five independent measurements

were carried out to measure the water contact angles, which is a vital

parameter of adhesion on the SU-8 textured surfaces.

The honeycomb textured surface demonstrated a water contact angle of

95o and without PFPE nano-lubricant, Si/SU-8/HC showed a hydrophobic

property. Based on the results from the different micro-dot pitch lengths, the

smallest pitch length (200 µm) of the micro-dot patterns provided a water

contact angle of 87o. When the pitch length increased to 300 µm, the water

contact angle for Si/SU-8/MD (300 µm) increased to 90o. The highest water

contact angle (90o) was observed at the pitch length of 300 µm, among

different pitch lengths of micro-dot patterns. However, after increasing the

pitch lengths of the micro-dot patterns to 400 µm and 500 µm, the water

contact angles dropped to 85o. The largest pitch length (600 µm) of the micro-

dot patterns provided a water contact angle of 87o and the water contact angle

for Si/SU-8/MD (600 µm) was higher than that of 400 µm and 500 µm.

Chapter 5

134

In current research studies, the nano-lubricated layer on the

honeycomb pattern showed a hydrophobic property and it provided a water

contact angle of 99o. The highest water contact angle of 101

o was obtained for

Si/SU-8/MD (400 µm)/PFPE and Si/SU-8/MD (600 µm)/PFPE. The smallest

pitch length (200 µm) of the micro-dot pattern with PFPE nano-lubricant had a

water contact angle of 98o. The water contact angles of 96

o and 98

o were

observed for Si/SU-8/MD (300 µm)/PFPE and Si/SU-8/MD (500 µm)/PFPE,

respectively.

Based on the experimental results, the surface features of the

honeycomb structure increased the water contact angle of the hydrophobic

surface on Si/SU-8. This may be attributed to the heterogeneous wetting

regimes, because air is trapped in between the honeycomb features, which is

similar to the behavior observed of lotus leaves and SU-8 micro-patterns

[Yoon et al. 2006 and Singh et al. 2007 and 2011]. The near hydrophobic

property of SU-8 (82o) changed to a high hydrophobic property, in the

presence of the honeycomb patterns on SU-8 that were implemented using a

negative asperity with pores. Moreover, among different pitch lengths, Si/SU-

8/MD observed the greatest water contact angle of 90o at a pitch length of 300

µm. Other pitch lengths of micro-dot patterns showed water contact angles in

the range of 85 o - 87

o.

Tay et al. [2011] observed a decreasing trend in water contact angles

with increasing pitch lengths of micro-dots on Si. In the literature, researchers

have found that patterned surfaces increased water contact angles and made

Chapter 5

135

Table 5.2. Sample nomenclature and the measured water contact angles (WCA) for

the SU-8 honeycomb patterns and the SU-8 micro-dot patterns with different pitch

lengths.

Materials Nomenclature WCA

(deg)

Textured surfaces without PFPE nano-lubricant

SU-8 polymer textured with

the honeycomb pattern

Si/SU-8/HC 95


the micro-dots pattern (200 µm)

Si/SU-8/MD (200 µm) 87



Si/SU-8/MD (300 µm) 90



Si/SU-8/MD (400 µm) 85



Si/SU-8/MD (500 µm) 85



Si/SU-8/MD (600 µm) 87

Textured surfaces with PFPE nano-lubricant


the honeycomb pattern

Si/SU-8/HC/PFPE 99



Si/SU-8/

MD (200 µm)/PFPE

98



Si/SU-8/

MD (300 µm)/PFPE

96



Si/SU-8/

MD (400 µm)/PFPE

101



Si/SU-8/

MD (500 µm)/PFPE

98



Si/SU-8/

MD (600 µm)/PFPE

101

surfaces more hydrophobic [Yoon et al. 2006 and 2007 and Singh et al. 2011].

Furthermore, the results from the experiments demonstrate that SU-8 patterned

surfaces increased the water contact angles, and the highest contact angle was

observed at the optimum pitch length of 300 µm. Regarding the spatial density

(in Table 5.3) of the micro-dot patterns, the water contact angles decreased as

the spatial texture densities increased, except at the optimized pitch lengths of

Chapter 5

136

300 µm and 600 µm. It can be concluded from the textured surfaces, that the

water contact angles can change according to the width of the recessed cavity.

In this study, there was a considerable increase in the water contact

angles on the SU-8 textured surfaces, due to the presence of the PFPE nano-

lubricant layer. In addition, the increased water contact angle affected the

tension of the surfaces. The presence of the fluorine atoms in PFPE enhanced

the hydrophobic property for contact angles greater than 90o. Moreover, lower

surface energy reduced the adhesion between the counterface ball and the

textured surfaces; this occurred because of the high contact angles.

5.4.3. Friction tests on Si/SU-8/HC and Si/SU-8/MD with different pitch

lengths

Fig. 5.6 shows the representative graphs of the coefficients of friction

for Si/SU-8/HC and Si/SU-8/MD (different pitch lengths) in dry frictional

tests. The fabricated pitch lengths of micro-dot patterns were 200 µm, 300 µm,

400 µm, 500 µm, and 600 µm. In dry frictional tests, a customized ball-on-

disc tribometer was used, and a normal load of 100 mN and a rotational speed

of 2 rpm were applied to the tested specimens. The coefficients of friction

were recorded for 20 rotational cycles and the average value of five different

tests is reported in here. The observed semi-hydrophobic property of the flat

SU-8 surface had a friction coefficient of 0.2. During the slow sliding tests, the

coefficients of friction were obtained in the range of 0.19 ~ 0.28 for the micro-

dot patterns with different pitch lengths.

Chapter 5

137

Among the different pitch lengths of micro-dot patterns, the coefficient

of friction (~0.198) at the optimized micro-dot pitch length of 400 µm was

slightly lower than that of the SU-8 untextured surfaces (~0.2). The

modification of the micro-dot patterns on the Si/SU-8 polymer had friction

coefficients of 0.23 and 0.28 at pitch lengths of 200 µm and 300 µm,

respectively. However, the coefficient of friction at the micro-dot pitch length

of 400 µm was lower than that which was observed at 200 µm and 300 µm.

The highest coefficient of friction was observed at the micro pitch length of

300 µm. The friction coefficients of the SU-8 micro-dot patterns were often

higher at low pitch lengths (200 – 300 µm), due to the large contact area on

multiple asperities. The coefficients of friction of the SU-8 micro-dot patterns

increased beyond the pitch length of 400 µm. In addition, Tay et al. [2011]

discussed the results of different micro-dot pattern pitch lengths on an Si

surface, with high coefficients of friction at both small pitch length of 50 µm

and a large pitch length of 450 µm.

The experimental results demonstrated that, after modifying the Si/SU-

8 surface with the honeycomb texture, the coefficient of friction of the

honeycomb textured surface was almost two times higher than that of the

untextured surfaces. Although most patterned structures with positive

asperities exhibited low coefficients of friction, the honeycomb structure with

pores (the negative asperity) had higher friction coefficient values than the flat

surface. In the literature, researchers have also observed higher frictional

values for textured surfaces with negative asperities. For example, Vilhena et

al. [2011] have observed that the dimple-textured steel disc (distance = 200

Chapter 5

138

µm, depth = 5.3 µm) showed more friction than the untextured disc, at a low

sliding speed of 0.015 m/s. Based on the current experiments, the coefficients

of friction must have been affected by the development of different contact

areas between the counterface ball and the SU-8 textured surfaces.

Despite the high water contact angle (lower apparent surface energy)

for Si/SU-8/MD (300 µm), the coefficient of friction for this texture was much

larger than that of other micro-dot patterns. Si/SU-8/MD (300 µm) had the

highest friction coefficient of 0.28. This observation indicates that the role of

texture geometry in deciding the coefficient of friction is important, in

addition to lower apparent surface energy. Moreover, the optimized micro-dot

pattern pitch length of 400 µm on Si/SU-8 provides a low coefficient of

friction at a spatial density of 0.75. Therefore, the spatial texture density plays

a primary role in the coefficient of friction and wear behavior.

5.4.3.1. Analysis of contact area versus the spatial density of the

textures on friction

Table 5.3 shows the summarized data of the dimensions and the spatial

texture density for Si/SU-8/HC, Si/SU-8/MD (200 µm), Si/SU-8/MD (300

µm), Si/SU-8/MD (400 µm), Si/SU-8/MD (500 µm), and Si/SU-8/MD (600

µm) and the contact area for a single asperity (based on the Hertz contact

equation) for Si/SU-8/HC and Si/SU-8/MD (400 µm). The negative asperity of

the honeycomb pattern showed the highest coefficient of friction (~0.41) at a

spatial texture density of 0.46. For the positive asperity of the textures, such as

the micro-dot patterns with different pitch lengths, the highest coefficient of

friction (~0.28) was observed at a pitch length of 300 µm. When the spatial

Chapter 5

139

texture density was small, at the pitch lengths of 200 µm and 300 µm, the real

contact area on multiple asperities was quite large and the coefficients of

friction increased due to the contribution of adhesion components between the

textures and the counterface ball.

Among different micro-dot pattern pitch lengths, a minimum

coefficient of friction occurred at a spatial texture density of 0.75 that was

formed at 300 µm. As the spatial texture density increased (within the range of

0.75 – 0.83), the real contact area may have decreased on multiple asperities,

however; the coefficients of friction increased due to the presence of plastic

deformation. Moreover, under this condition of high spatial densities, the

counterface ball could pass through the micro-dot patterns, potentially

touching the bottom of the flat SU-8 surface. This may increase the adhesion

forces and the coefficient of friction. Due to the size of the textures and the

spatial texture density, the contact area and the overall coefficients of friction

were varied. Moreover, a point of increase for the coefficient of friction may

be the roughness of the surface. In addition, the high surface roughness of the

honeycomb pattern can increase the contact area and increase the frictional

force. Fig. 5.7 represents the coefficients of friction versus the spatial texture

densities for the SU-8 honeycomb patterns and the SU-8 micro-dot patterns

with different pitch lengths.

Chapter 5

140

Table 5.3. The dimensions, spatial texture densities, coefficients of friction (COF)

and calculated contact area based on the Hertz contact equation for the tested

specimens under a normal load of 100 mN.

Samples

Dimensions (µm)

COF

Contact

area

[x10-9

m2]

(µm)

(µm)

Density =

Negative asperity

Si/SU-8/HC 200 440 0.46 0.41 2.87

Positive asperity

Si/SU-8/MD

(200 µm)

100 200 0.5 0.23 -

Si/SU-8/MD

(300 µm)

200 300 0.67 0.28 -

Si/SU-8/MD

(400 µm)

300 400 0.75 0.198 1.05

Si/SU-8/MD

(500 µm)

400 500 0.8 0.26 -

Si/SU-8/MD

(600 µm)

500 600 0.83 0.25 -

Moreover, in order to analyze the contact area on a single asperity, the

micro-dot pattern with a pitch length of 400 µm and the honeycomb pattern

were included in this calculation that was based on Hertz’s contact equation.

The detailed calculation method for a single asperity of the surface texture is

provided in Chapter 8. Based on the calculation results of Hertz’s contact

equation on a single asperity, the contact area for Si/SU-8/MD (400 µm) was

lower than that of the flat SU-8 surface (Si/SU-8) and the SU-8 honeycomb

patterned surface (Si/SU-8/HC). Experiments showed that the contact area on

a single asperity for the positive asperity of the texture was less than that of

the negative asperity of the texture. Thus, the protruding part of the surface

texture reduced the contact area and the coefficient of friction. In addition,

contact areas on multiple asperities are also required to change the friction.

W L

L

W

Chapter 5

141

Therefore, the spatial density is of great significance to the optimization of

friction behavior.

Fig. 5.6. The summarized results of the coefficients of friction for the SU-8 micro-dot

patterns with different pitch lengths (200 µm, 300 µm, 400 µm, 500 µm, and 600 µm)

and the SU-8 honeycomb pattern in dry sliding tests under a normal load of 100 mN

and a rotational speed of 2 rpm.

Fig. 5.7. The coefficients of friction and the spatial texture densities for the

honeycomb patterns and the micro-dot patterns with different pitch lengths.

0.23

0.28

0.198

0.26 0.25

0.41

0

0.2

0.4

0.6

Si/SU-8

/MD

(200

µm)

Si/SU-8

/MD

(300

µm)

Si/SU-8

/MD

(400

µm)

Si/SU-8

/MD

(500

µm)

Si/SU-8

/MD

(600

µm)

Si/SU-8

/HC

Samples

Coef

fici

ent

of

fric

tion


0

0.2

0.4

0.6

0.8

1

Si/SU-8/MD(200 um)

Si/SU-8/MD(300 um)

Si/SU-8/MD(400 um)

Si/SU-8/MD(500 um)

Si/SU-8/MD(600 um)

Samples

Co

eff

icie

nt

of

fric

tio

n

Coefficient of friciton Spatial Texture Density (W/L)

Si/SU-8


Spatial Texture Density

Negative asperity of honeycomb pattern (Spatial Density = ~0.46)

Chapter 5

142

5.4.3.2. Analysis of wear mechanisms for the friction tests

The optical images of wear track on the tested specimens and wear

debris on the Si3N4 counterface ball under a normal load of 100 mN and a

rotational speed of 2 rpm are shown in Fig. 5.8. The wear behaviors were

examined under an optical microscope in dry friction tests. The wear track was

clearly observed and the wear particles were generated at the corners of the

micro-dot patterns at the small pitch lengths of 200 µm and 300 µm, and

visible small wear debris was attached to the counterface ball. However, less

wear particles were seen on the micro-dot specimens with pitch lengths of 400

µm, 500 µm, and 600 µm. A reduction in wear debris was observed when the

micro-dot pattern pitch lengths increased.

However, a wear track was observed on the ground of the flat SU-8

surface within the micro-dots. A wear track was also seen on the Si/SU-8/HC

sample, and a small amount of debris had accumulated on the counterface ball.

The sharp corners on the honeycomb pattern could act as stress concentration

points and lead to early wear during sliding; thus, these sharp corners can lead

to high friction. Better tribological behaviors were observed on the positive

asperity of the textures. Thus, the wear particles can be reduced on the positive

asperity of the textures. In conclusion, contact conditions and pattern geometry

play vital roles in reducing friction and increasing wear life cycles.

Chapter 5

143

Wear track

(a) Micro-dot patterns with 200 µm (Si/SU-8/MD (200 µm))

(b) Micro-dot patterns with 300 µm (Si/SU-8/MD (300 µm))

Wear track

Wear track

Wear track

(d) Micro-dot patterns with 500 µm (Si/SU-8/MD (500 µm))

(c) Micro-dot patterns with 400 µm (Si/SU-8/MD (400 µm))

Chapter 5

144

Fig. 5.8. The optical images of debris on the counterface ball and wear track on

Si/SU-8/MD(200 µm), Si/SU-8/MD(300 µm), Si/SU-8/MD(400 µm), Si/SU-8/MD(500

µm), Si/SU-8/MD(600 µm), and Si/SU-8/HC (from FESEM) in dry sliding tests under

a normal load of 100 mN and a rotational speed of 2 rpm.

5.4.4. Wear tests on Si/SU-8/HC and Si/SU-8/MD with different pitch

lengths

Fig. 5.9 shows the values of the coefficeints of friction for the tested

specimens that were overcoated with and without PFPE lubricant in the wear

tests, such as Si/SU-8/MD (200 µm), Si/SU-8/MD (300 µm), Si/SU-8/MD

(400 µm), Si/SU-8/MD (500 µm), Si/SU-8/MD (600 µm), and Si/SU-8/HC. A

normal load of 100 mN was applied to the tested specimens with a high

rotational speed of 500 rpm, and the experimental measurements were

recorded with a sampling rate of 1 Hz. These samples were overcoated with

nano-lubricant of PFPE (0.2 wt% solution of PFPE Z-dol 4000 in H-Galden

(f) Honeycomb pattern (Si/SU-8/HC)

Wear track

Wear track

(e) Micro-dot patterns with 600 µm (Si/SU-8/MD (600 µm))

Chapter 5

145

ZV60) to enhance their wear durability. The coated layer of PFPE was

obtained as 2 – 4 nm, using the dip-coating method, with a dipping and

withdrawal speed of 2.1 mm/s and a dipping duration of 30 seconds.

In wear tests that used a high speed of 500 rpm on the honeycomb

patterns, the coefficient of friction was 0.44 for the uncoated layer of the

honeycomb surface and the coefficient of friction for Si/SU-8/HC was higher

than that of the flat SU-8 surface. However, according to the wear tests, the

wear life cycles of Si/SU-8/HC were better than that of Si/SU-8 and the wear

durability increased to 1,200 cycles without PFPE nano-lubricant. A

remarkable wear track was observed on this honeycomb pattern and wear

debris was also found on the Si3N4 counterface ball. The high speed (500 rpm)

experiments showed the highest friction coefficient of 0.29 on the SU-8 micro-

dot pattern at a pitch length of 400 µm without PFPE nano-lubricant; the wear

durability was 650 cycles. The investigation of friction and wear behaviors of

different micro-dot pitch lengths demonstrated the smallest coefficient of

friction (~0.23) for a pitch length of 300 µm under a dry condition. A low

wear life (650 cycles) was observed at 300 µm and 400 µm pitch lengths.

Si/SU-8/MD had friction coefficients of 0.25, 0.27 and 0.26 for the pitch

lengths of 200 µm, 500 µm, and 600 µm, respectively. Based on the results

from the low sliding speed (2 rpm) in friction tests and the high sliding speed

(500 rpm) in wear tests, the coefficients of friction for all tested specimens

(except Si/SU-8/MD (300 µm)) were significantly increased with high speeds.

This was due to the generation of high temperatures at the contact between the

counterface ball and the textured surfaces.

Chapter 5

146

The representative graph for the wear life cycles of the micro-dot

patterns and honeycomb patterns on Si/SU-8 with and without a PFPE nano-

coated layer is shown in Fig. 5.10. The wear life cycles were high, at 1,700

cycles, for Si/SU-8/MD (600 µm) without the PFPE nano-lubricant, although

the coefficient of friction for this pitch length was high among the different

micro-dot pitch lengths of SU-8. The wear life cycle for Si/SU-8/MD (600 µm)

was more than six times that of the SU-8 untextured surface. A wear track was

seen on the SU-8 micro-dots at a small pitch length (200 µm) and at large

pitch lengths (500 µm and 600 µm) without any lubricant coating. A small

wear track was observed on the surface of Si/SU-8/MD, at the pitch lengths of

300 µm and 400 µm. To determine the friction and wear behaviours of the SU-

8 textured surface during high speed wear tests, the optical images of wear

debris on the Si3N4 counterface ball and the wear tracks on these tested

samples were observed under the microscope; these are shown in Fig. 5.11.

Significant improvements in wear durability and a reduction in the

coefficient of friction were observed after overcoating PFPE nano-lubricant on

the textured surfaces. After coating a layer of PFPE nano-lubricant on the

tested samples, the wear life cycles of the tested samples increased

significantly and the coefficients of friction decreased further. The presence of

PFPE coating on the SU-8 micro-dot pitch lengths of 300 µm and 400 µm

showed the lowest coefficients of friction among the SU-8 micro-dot textured

surfaces. The wear life cycles of the SU-8 micro-dot textured surfaces, with

300 µm and 400 µm pitch lengths, significantly increased from 650 to 20,000

cycles and 21,000 cycles, respectively, due to the presence of the PFPE coated

Chapter 5

147

layer. Wear lives of 25,000 and 14,500 cycles were registered for Si/SU-8/MD

(500 µm)/PFPE and Si/SU-8/MD (600 µm)/PFPE, respectively and the

coefficient of friction was reduced by two times for Si/SU-8/MD (500 µm)

and Si/SU-8/MD (600 µm) after coating the PFPE nano-lubricant. The SU-8

micro-dot patterns without the PFPE nano-lubricant were easily worn out for

all pitch lengths. A large amount of wear particles were generated at the corner

of the micro-dot patterns when the lubricant layer was absent from these

specimens. The presence of the PFPE coated layer can cover the textured

surfaces if they are not deteriorated, and it can increase the wear durability of

the specimens.

Fig. 5.9. The coefficients of friction for Si/SU-8/MD (200 µm), Si/SU-8/MD (300 µm),

Si/SU-8/MD (400 µm), Si/SU-8/MD (500 µm), Si/SU-8/MD (600 µm), and Si/SU-

8/HC with and without PFPE nano-lubricant under a normal load of 100 mN and a

rotational speed of 500 rpm.

0.25 0.23

0.290.27 0.26

0.44

0.15

0.07 0.070.13 0.10 0.12

0.00

0.10

0.20

0.30

0.40

0.50

0.60

Si/SU-8/MD

(200 um)

Si/SU-8/MD

(300 um)

Si/SU-8/MD

(400 um)

Si/SU-8/MD

(500 um)

Si/SU-8/MD

(600 um)

Si/SU-8/HC

Samples

Co

eff

icie

nt

of

fric

tio

n

Without PFPE

With PFPE

Chapter 5

148

Fig. 5.10. The wear life cycles for different specimens (Si/SU-8/MD (200 µm), Si/SU-

8/MD (300 µm), Si/SU-8/MD (400 µm), Si/SU-8/MD (500 µm), Si/SU-8/MD (600

µm), and Si/SU-8/HC) with and without PFPE nano-lubricant under a normal load of

100 mN and a rotational speed of 500 rpm.

650 650 8001700

3000

20000

12001500

21000 22,5001450025000

1

10

100

1000

10000

100000

Si/SU

-8/M

D (2

00 u

m)

Si/SU

-8/M

D (3

00 u

m)

Si/SU

-8/M

D (4

00 u

m)

Si/SU

-8/M

D (5

00 u

m)

Si/SU

-8/M

D (6

00 u

m)

Si/SU

-8/H

C

Samples

Wea

r L

ife

Cy

cles

(C

ycl

es)

Without PFPE

With PFPE

Si/SU-8

Wear track

1,500 cycles

(c) Si/SU-8/MD (200 µm)

Wear track

(d) Si/SU-8/MD (200 µm)/PFPE

3,000 cycles

650 cycles

(e) Si/SU-8/MD (300 µm)

Wear track

20,000 cycles

Wear track

(f) Si/SU-8/MD (300 µm)/PFPE

650 cycles 21,000 cycles

Wear track Wear track

(g) Si/SU-8/MD (400 µm) (h) Si/SU-8/MD (400 µm)/PFPE

Wear track

(a) Si/SU-8/HC

1,200 cycles

Wear track

5,000 cycles

(b) Si/SU-8/HC/PFPE

Chapter 5

149

Fig. 5.11. The optical images of wear debris on the counterface ball and wear track

on the tested samples with and without PFPE nano-lubricant under a normal load of

100 mN and a high rotational speed of 500 rpm.

5.4.5. Effects of different normal loads on friction for Si/SU-8/HC and

Si/SU-8/MD (400 µm)

Different normal loads of 50 mN, 100 mN, 150 mN, 300 mN, and 400

mN were used and a rotational speed of 2 rpm was applied to the tested

specimens. The honeycomb pattern and the micro-dot patterns of 400 µm were

used in the investigations of the effect of different normal loads. Fig. 5.12

indicates the resulting coefficients of friction versus the different normal loads

for the SU-8 honeycomb patterns and the SU-8 micro-dot patterns, with a 400

µm pitch length. The coefficient of friction for Si/SU-8/HC decreased from a

value of 0.43 at 50 mN to 0.32 at 400 mN. The coefficient of friction for

Si/SU-8/HC was higher than that of Si/SU-8/MD (400 µm) at all normal loads.

Experiments showed that a friction coefficient of 0.2 was obtained for Si/SU-

8/MD (400 µm) at 50 mN. The coefficient of friction steadily decreased after

the normal load of 50 mN until the normal load of 400 mN, which had the

lowest coefficient of friction of 0.169. Thus, the coefficient of friction for

Wear track (i) Si/SU-8/MD (500 µm)

800 cycles 25,000 cycles

(j) Si/SU-8/MD (500 µm)/PFPE

Wear track

1,700 cycles 14,500 cycles

(k) Si/SU-8/MD (600 µm)

Wear track Wear track

(l) Si/SU-8/MD (600 µm)/PFPE

Chapter 5

150

Si/SU-8/MD (400 µm) decreased with increasing normal loads in the range of

50 mN to 400 mN. Likewise, Unal et al. [2004] investigated the influence of

test speed and load values on the friction and wear behaviors of pure

polytetrafluoroethylene (PTFE), glass fibre reinforced (GFR), and bronze and

carbon (C) filled PTFE polymers. They found that the coefficient of friction

for pure PTFE and its composite decreased with increasing load, and the

improved load carrying capability was observed on the reinforcement PTFE

with glass fibres.

According to the Amonton’s Law, the friction force is directly

proportional to the applied normal load. Table 5.4 shows that the calculated

contact area on a single asperity for the SU-8 tested specimens significantly

increased as the normal loads increased from 50 mN to 400 mN. Based on the

Hertz contact calculation for a single asperity, the contact area of Si/SU-8/MD

(400 µm) was less than that of Si/SU-8/HC for all of the normal loads used.

With a low load of 50 mN, the coefficient of friction for the tested samples

was high, as the contact area was small and the contact pressure was high; no

plastic deformation of the SU-8 textures was involved.

It is also possible that at low normal load the effect of surface force is a

dominant contributor to the coefficient of friction. A reduction in the

coefficient of friction has been observed as normal loads increase [Shooter and

Tabor 1952, Zou et al. 2006, Unal et al. 2004 and He et al. 2008]. These

experiments also prove that the coefficients of friction for Si/SU-8/HC

decrease with increasing normal loads (in the range of 50 mN – 400 mN), and

Chapter 5

151

Si/SU-8/MD (400 µm) demonstrated a decreasing trend with increasing

normal loads. Based on Equation 2.25:)1( nkW , by Stuart, the

coefficients of friction decrease with increasing normal loads for visco-elastic

properties.

Fig. 5.12. The resulting coefficient of friction versus the different normal loads for

Si/SU-8/HC and Si/SU-8/MD (400 µm).

The findings clearly show that the effect of texturing provides different

contact areas for different textures under the same load. Within the elastic

range, the real contact area can change with different normal loads and

different surface geometries. The real contact area coupled with the actual

geometry of the pattern plays an important role in determining the coefficient

of friction. Hence, the friction may be mainly controlled by the nominal

contact area. Moreover, the height of the texture is also important as the lower

height may allow the counterface to touch the bottom surface of the valleys

because of the large elastic deformation on the polymeric surface. Thus, the

positive asperity of the micro-dot patterns improved coefficients of friction

and enhanced load carrying capacity for loads between 50 mN and 400 mN.

0

0.1

0.2

0.3

0.4

0.5

0 100 200 300 400 500Load (mN)

Coeff

icie

nt

of

fric

tion

Si/SU-8/MD (400 µm) Si/SU-8/HC

Chapter 5

152



condition) versus the applied normal loads for Si/SU-8/HC and Si/SU-8/MD (400

µm).

Load

(mN)

Si/SU-8/HC Si/SU-8/MD(400 µm)

Contact

area

[x10-9

m2]

Coefficient of

friction ( )

(from

experiment)

Contact

area

[x10-9

m2]

Coefficient of

friction ( )

(from

experiment)

50 1.81 0.43 0.65 0.2

100 2.87 0.41 1.05 0.198

150 3.76 0.38 1.37 0.175

300 5.97 0.35 2.17 0.186

400 7.24 0.31 2.63 0.169

5.5. Summary

The current chapter presents the advantages of the positive asperity on

surface textures, regarding the tribological performance of Si/SU-8. The

spatial texture densities for the positive asperities on the micro-dot patterns

were between 0.5 and 0.83. After patterning the micro-dot pattern on the

Si/SU-8 surface, the coefficient of friction for the micro-dot with a 400 µm

pitch length was slightly less than that of the flat SU-8 surface, under dry

sliding tests with a slow sliding speed of 2 rpm. However, the coefficeint of

friction was two times greater for the negative asperity of the honeycomb

pattern when comparing it to the flat SU-8 surface. The coefficients of friction

increased, after the optimized pitch length of 400 µm, as the spatial texture

density increased. The minimum coefficient of friction was observed at the

optimized spatial texture density of 0.75 for the micro-dot patterns.

The contact area between the counterface ball and the honeycomb

pattern was larger than that of the micro-dot pattern. At the small pitch lengths

Chapter 5

153

of 200 µm and 300 µm, the real contact area was quite large when the spatial

texture density was small. The coefficients of friction increased due to the

contribution of the adhesion components and the large contact area between

the textures and the counterface ball. The coefficients of friction for the

honeycomb pattern and the micro-dot pattern on Si/SU-8 decreased with

increasing normal loads. As a consequence, the surface textures provided a

higher load carrying capacity. Therefore, the spatial texture density and the

real contact area play a vital role in influencing the coefficient of friction. An

investigation of the effect of contact area on the frictional behaviors is

presented using different size counterface balls in Chapter 6.

Patterning with surface textures, such as the honeycomb and micro-dot

patterns on Si/SU-8, provided higher wear life cycles than the flat SU-8

surface. The wear durability of the observed honeycomb pattern was 1,200

cycles. The wear particles on the tested surfaces were significantly reduced,

due to the presence of surface textures on Si/SU-8. Overcoating of a PFPE top

layer on the honeycomb pattern increased the wear durability to 22,500 cycles;

the wear lifecycle was improved by eighteen times for Si/SU-8/HC/PFPE.

Among different micro-dot pitch lengths, the highest wear life cycles were

observed for micro-dots with 500 µm pitch lengths. Moreover, the coefficients

of friction markedly decreased when a PFPE coated layer was applied to the

SU-8 textured surfaces.

In conclusion, on the SU-8 layer surface, both the positive asperity and

the negative asperity of surface textures can be successfully developed and

Chapter 5

154

excellent tribological performances are observed on the positive asperity of

surface patterns. The protruding parts of surface textures also support high

wear durability and low friction; they also provided load carrying capacity.

PFPE is an important nano-layer, which covers textured surfaces that are not

deteriorated and improves their wear life cycles.

Chapter 6

155

Chapter 6

The Effect of Different Size Counterface Balls on SU-8

Textured Surfaces

The protuberances of surface patterns have enhanced the friction and

wear behavior of MEMS devices. The contact area on the surface texture

plays a vital role in the friction and adhesion forces. This chapter presents

tribological studies on SU-8 textured surfaces using different counterface

balls. Surface textures with positive asperities (negative fingerprint and

positive fingerprint) were involved, to investigate tribological behavior using

different size counterface balls of 2 mm, 4 mm, and 6 mm that were attached

to a customized ball-on-disc tribometer. PFPE nano-lubricant was used in the

wear investigations to determine the wear behavior and the resistance of

positive asperity on the SU-8 surface texture.

Chapter 6

156


Surface patterns offer a challenge to SU-8 polymer surfaces, as the

frictional characteristics are affected by differences in surface properties and

surface topography, such as width, length, and height [Tay et al. 2011 and

Singh et al. 2011]. Furthermore, surface texture, including the spatial texture

density, plays an important role in the influence of friction and wear. In the

previous chapters, different fabrication techniques were used to pattern the

negative asperity (e.g. groove/honeycomb) and positive asperity (e.g.

pillar/fingerprint) on surface textures. SU-8 polymer was used as a thin film

layer on silicon; this material has been widely used in MEMS devices. It was

observed that the frictional property of the SU-8 thin layer was better than that

of the silicon wafer. In order to obtain better friction and wear lives for SU-8

samples, surface textures were fabricated on the SU-8 layer. In addition to the

friction and wear behavior tests described in the previous two chapters,

different surface patterns with positive and negative asperities were

investigated, using a customized ball-on-disc tribometer with a 4 mm

counterface ball.

By changing the surface patterns on SU-8 polymer, the contact areas

on multiple asperities between two sliding surfaces were significantly varied

and changes in the contact area on multiple asperities had a great influence on

the friction and wear properties of the SU-8 textured surfaces. It was also

observed that the coefficients of friction were primarily affected by changes in

spatial texture density. Understanding changes in surface texture, is therefore

Chapter 6

157

important to the exploration of friction. Moreover, different types of sliding or

rolling surfaces may change the contact area and friction. Thus, in this Chapter

the effects of different sliding surfaces (different counterface balls) on the

friction and wear behaviors of the SU-8 textured surfaces are discussed. An

investigation of the relationship between the contact area of the different

counterface balls and the friction is also discussed in this Chapter.

In these investigations, the flat SU-8 surface and the negative and

positive fingerprint patterns on Si/SU-8 were used as the tested specimens. An

index finger and a PDMS negative fingerprint pattern were used as master

piece moulds to fabricate and fine tune patterns with positive asperities on

Si/SU-8. The detailed fabrication methods of the SU-8 patterned surfaces are

described in Chapter 4. Both dry (unlubricated) and lubricated (with PFPE

nano-lubricant) tests were conducted and the concentration of the PFPE

lubricant solution used was 0.2 wt%.

The main purpose of this Chapter is to investigate the changes in

contact area and the effects of friction and wear when different counterface

balls are applied to the SU-8 textured surfaces. Different dimensions of

counterface balls, such as 2 mm, 4 mm, and 6 mm (hardness and roughness of

the balls are 1500 HV and 5 nm), were used in these experiments. First, the

frictional properties of Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP are

investigated with different counterface balls (2 mm, 4 mm, and 6 mm) at a

fixed applied normal load of 100 mN. The frictional data were recorded under

dry conditions with a low rotational speed of 2 rpm and a sampling rate of 10

Chapter 6

158

Hz. Second, the effects of different counterface balls on wear durability were

examined, in order to understand the resistance of the SU-8 textured surfaces.

In the wear durability tests, a high rotational speed of 500 rpm was used with a

fixed normal load of 100 mN.

6.2. Materials and sample preparation

Polished n-type bare Si (100) wafers of approximately 525 ± 25 µm in

thickness, with a hardness of 12.4 GPa (obtained from SYST Integration Pte

Ltd, Singapore), were used as substrates. The cleaning process was carried out

on an Si substrate with an acetone and oxygen plasma cleaner. The detailed

procedures for cleaning Si substrates are described in Section 3.3.2. Grade

2050 SU-8 was deposited on the clean Si substrate and the spin-coating

process (described in detail in Section 3.3.3) was used to obtain an SU-8 film

with a thickness of 80 µm (measured by Wyko NT1100 Optical Profiler). The

pre-baking process was implemented with a soft baking time of 5 minutes at

65oC, and was repeated for 15 minutes at 95

oC. These fabricated specimens

were exposed to UV light for 10 – 15 seconds at a power at 215 mJ/cm2. The

final fabrication step to obtain the flat SU-8 specimens was the post-baking

process. The detailed process for the fabrication of the flat SU-8 surface is

described in Section 3.3.6.

In order to obtain the negative fingerprint pattern on Si/SU-8, an index

finger was used as the master piece mould to attach the fabricated Si/SU-8

samples. After forming the negative fingerprint pattern on Si/SU-8, exposure

to UV light and the curing process for the SU-8 textured surfaces were carried

Chapter 6

159

out. The detailed steps for fabricating Si/SU-8/NFP are presented in Section

3.3.9. The PDMS negative fingerprint pattern was used as the master piece

mould in order to fabricate the positive fingerprint pattern on Si/SU-8 (Si/SU-

8/PFP). In this fabrication of Si/SU-8/PFP, polydimethylsiloxane (PDMS)

(supplied by Dow Corning Corp., USA) was used in a ratio of 10:1. The final

specimen of Si/SU-8/PFP was obtained after finishing the pre-baking process,

exposure to UV light, and the post-baking process. The detailed fabrication

process for Si/SU-8/PFP is presented in Section 3.3.10. The tested specimens,

such as Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP were used in the

investigations of tribological behaviors with different counterface balls.

PFPE (0.2 wt% solution of PFPE Zdol 4000 in H-Galden ZV60) was

dip-coated onto the tested specimens at dipping and withdraw speeds of 2.1

mm/s with a fixed dipping duration of 30 seconds. The thickness of the PFPE

nano-lubricant on the samples was 2 - 4 nm (measured by ellipsometer). The

samples were kept in a clean room for 24 hours for further tribological tests.

6.3. Experimental procedure

After the fabrication process was complete, surface characterizations

(such as surface dimension and surface roughness) for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP were obtained with an optical microscope and an optical

profiler (Wyko NT1100). Subsequently, a surface analyzing technique, such as

contact angle measurement, was carried out to investigate the surface

properties of the textured surfaces, with and without a PFPE layer. A

deionized water droplet of 0.5 µL was ejected onto the tested samples to

Chapter 6

160

measure surface wettability. The results of surface characterization and the

discussions are explained in detail in Chapter 4.

The friction and wear tests were carried out using a custom-built ball-

on-disc tribometer (Fig. 3.25). During the rotational tests, the friction force

was recorded from the deflection of the cantilever to which the ball was

attached. The laser displacement sensors (MTI Instruments Inc., New York,

USA) recorded the vertical and lateral displacement of the forces, and these

forces were then converted to normal load and frictional force using a

calibration chart. The friction data was recorded with an applied normal load

of 100 mN and a slow rotational speed of 2 rpm (linear relative speed at the

sliding contact = 0.21 mm/s).

For the wear tests, a high rotational speed of 500 rpm (linear relative

speed = 52.36 mm/s) was used, with a fixed normal load of 100 mN. Different

sampling rates of 10 Hz and 1 Hz were used to record friction and wear data,

respectively. In addition, PFPE lubricant was involved in the investigation of

wear tests. The wear durability of the tested samples was defined as the

number of cycles. The experiments were stopped when the coefficient of

friction was greater than 0.3, fluctuations of friction were observed, and/or

wear tracks were observed on the specimens. The average coefficient of

friction from five repetitive sliding tests on different samples was reported and

the experiments were carried out under ambient conditions: a room

temperature of 25oC and a relative humidity of 58%. Different sizes of silicon

nitride balls, such as 2 mm, 4 mm, and 6 mm, were used to conducted tests on

Chapter 6

161

the specimens to investigate their friction and wear behaviors. Before carrying

out the experiments, the silicon nitride counterface balls were cleaned with

acetone.


6.4.1. Effects of different size silicon nitride counterface balls on friction

for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP

Fig. 6.1 shows the coefficients of friction versus different counterface

balls (2 mm, 4 mm, and 6 mm) for the flat SU-8 surface (Si/SU-8), using a

normal load of 100 mN and a rotational speed of 2 rpm. The investigations of

the friction data were carried out on Si/SU-8 using a customized ball-on-disc

tribometer with different diameters of silicon nitride balls (2 mm, 4 mm, and 6

mm), a normal load of 100 mN, and a rotational speed of 2 rpm (linear speed =

0.21 mm/s). The average from five repetitive rotational tests that used a

sampling rate of 10 Hz is reported here.

Based on the experimental results, the lowest coefficient of friction

was observed when the Si3N4 counterface ball with 4 mm diameter was used

to test the Si/SU-8. The smallest diameter of 2 mm provided a coefficient of

friction of 0.28. The coefficient of friction was increased to 1.01 when a 6 mm

diameter counterface ball on Si/SU-8 was used. The coefficient of friction for

a 6 mm silicon nitride ball found by investigations was markedly higher than

that of the 2 mm counterface ball. The experimental results showed that the

friction on Si/SU-8 increased above 0.3, which was the coefficient of friction

limit in these experiments, when testing with 6 mm counterface ball.

Chapter 6

162

Fig. 6.2 shows the coefficients of friction for different counterface

balls (2 mm, 4 mm, and 6 mm) on Si/SU-8/NFP, under a normal load of 100

mN and a rotational speed of 2 rpm. The highest coefficient of friction (~0.241)

was observed on Si/SU-8/NFP while testing with the 2 mm counterface ball.

The coefficient of friction for Si/SU-8/NFP with a 4 mm counterface ball was

less than that of 2 mm. For 6 mm counterface balls, the observed coefficient of

friction was 0.12. The lowest coefficient of friction (~0.08) was obtained when

the tested counterface ball diameter increased to 4 mm. Based on the

experimental results of different counterface balls on Si/SU-8/NFP, the

smallest counterface ball of 2 mm provided the highest coefficient of friction

(~0.241). Although a 2 mm diameter counterface ball was small for sliding on

Si/SU-8/NFP, the contact area between the counterface ball and Si/SU-8/NFP

was large on multiple asperities. The best coefficient of friction was observed

on Si/SU-8/NFP with the optimized counterface ball of 4 mm in diameter.

Fig. 6.1. The coefficients of friction versus different counterface balls (2 mm, 4 mm,

and 6 mm) on the flat SU-8 surface (Si/SU-8) under a normal load of 100 mN and a


0.28 0.2

1.01

0

0.2

0.4

0.6

0.8

1

1.2

1.4

2 mm 4 mm 6 mm

Coef

fici

ent

of

fric

tion

Different size counterface balls

Si/SU-8

Chapter 6

163

Fig. 6.2. The coefficients of friction versus different counterface balls on the SU-8

negative fingerprint pattern (Si/SU-8/NFP) under a normal load of 100 mN and a


Fig. 6.3. The coefficients of friction versus different counterface balls on the SU-8

positive fingerprint pattern (Si/SU-8/PFP) under a normal load of 100 mN and a


Fig. 6.3 shows the values of the coefficient of friction for the different

counterface balls on Si/SU-8/PFP, under a normal load of 100 mN and a

rotational speed of 2 rpm. The smallest (2 mm) counterface ball provided a

coefficient of friction of 0.2188. When testing with the larger 4 mm

0.241

0.08

0.12

0

0.2

0.4

2 mm 4 mm 6 mm

Co

effi

cien

t o

f fr

icti

on

Different counterface balls

Si/SU-8/NFP

0.2188

0.164

0.35

0

0.4

2 mm 4 mm 6 mm

Coef

fici

ent

of

fric

tion


Si/SU-8/PFP

Chapter 6

164

counterface ball, the coefficient of friction significantly decreased to 0.164;

yet, the observed coefficient of friction was 0.35 with a 6 mm counterface ball.

According to the experimental results, the minimum coefficient of friction was

observed with a 4 mm diameter counterface ball.

6.4.1.1. Analysis of the dependence of coefficients of friction on contact

area with spatial texture density

Table 6.1 shows the summarized results for the coefficients of friction

and the spatial texture density of Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP.

The representative summarized graph for the coefficients of friction with

different counterface balls for all tested specimens is shown in Fig. 6.4. From

the experimental results of the 2 mm counterface ball on Si/SU-8, Si/SU-

8/NFP, and Si/SU-8/PFP, the highest coefficient of friction was observed on

Si/SU-8. On the textured surfaces with a 2 mm counterface ball, Si/SU-8/NFP

had a coefficient of friction of 0.241, and the lowest coefficient of friction

(0.2188) was observed with Si/SU-8/PFP. In the investigations with the small

(2 mm) counterface ball on the textured surfaces, the coefficient of friction

was high for Si/SU-8/NFP at a large spatial texture density of 0.53; the small

spatial texture density of 0.42 provided a low coefficient of friction for Si/SU-

8/PFP. Based on the calculation of the contact area of a 2 mm counterface ball,

the contact area under a static condition for Si/SU-8/PFP was smaller than that

of Si/SU-8 and Si/SU-8/NFP.

The calculated contact areas and contact pressure on Si/SU-8, Si/SU-

8/NFP, and Si/SU-8/PFP are shown in Table 6.2. The reduction in contact area

Chapter 6

165

on Si/SU-8/PFP and Si/SU-8/NFP decreased the adhesion forces and the

coefficient of friction for the small dimension of 2 mm. However, it was

observed that the calculated mean contact pressures were high after patterning

the fingerprint texture on Si/SU-8. The mean contact pressure on Si/SU-8/NFP

and Si/SU-8/PFP were 64.52 MPa and 65.36 MPa, respectively, under the

normal load of 100 mN.

When the diameter of the counterface ball was increased to 4 mm, the

coefficients of friction slightly decreased for all tested specimens under a

normal load of 100 mN. Based on the experimental results from the 4 mm

counterface ball, Si/SU-8 obtained the highest coefficient of friction of 0.2.

The minimum coefficient of friction was observed with Si/SU-8/NFP, at a

spatial texture density of 0.53. Moreover, Si/SU-8/PFP had a higher

coefficient of friction than Si/SU-8/NFP in the investigations using a large

counterface ball of 4 mm. As shown in Table 6.2, the calculated contact area

on a single asperity for Si/SU-8/NFP was higher than that of Si/SU-8/PFP.

However, Si/SU-8/NFP showed lower contact mean pressure than Si/SU-

8/PFP. The mean contact pressures for Si/SU-8, Si/SU-8/NFP and Si/SU-

8/PFP were 34.84 MPa, 45.87 MPa, and 47.39 MPa, respectively.

When investigations were carried out on different test specimens with

the 6 mm counterface ball, the results demonstrated that the highest coefficient

of friction (~1.01) was observed for Si/SU-8, and Si/SU-8/PFP had a

coefficient of friction of 0.35. According to the experimental results, the

coefficients of friction for Si/SU-8 and Si/SU-8/PFP were higher than 0.3,

Chapter 6

166

which was the coefficient of friction limit in our experiments. Based on the

calculated contact area of the 6 mm counterface ball on a single asperity, using

Hertzian’s contact method, a high contact area was obtained for Si/SU-8.

Based on investigations with different counterface balls, the contact

areas on multiple asperities of surface textures played a vital role in changing

the frictional force. On the flat SU-8 surface, the coefficients of friction were

high for all counterface balls due to the large contact area obtained using

Hertzian’s contact calculation. According to the experimental results of the

SU-8 fingerprint textures, the minimum coefficients of friction were for

Si/SU-8/NFP, using 4 mm and 6 mm counterface balls. It was observed that

although the contact area on a single asperity for Si/SU-8/PFP was lower than

that of Si/SU-8/NFP, the coefficient of friction for Si/SU-8/PFP was high due

to the large overall contact area on multiple asperities.

A small spatial texture density (0.42) of Si/SU-8/PFP (in Table 6.1)

created a large real contact area on multiple asperities. Due to the large contact

area on multiple asperities, high adhesion forces occurred between the large

counterface balls (4 mm and 6 mm) and Si/SU-8/PFP, thus supporting the

increase of the coefficient of friction for Si/SU-8/PFP. Moreover, a large

spatial texture density for Si/SU-8/NFP contributed to a reduction in contact

area and improved friction and wear behaviors. The smallest coefficient of

friction (~0.08) for Si/SU-8/NFP was observed with the optimized counterface

ball of 4 mm, rather than the 2 mm and 6 mm counterface balls.

Chapter 6

167

Although the large counterface balls of 4 mm and 6 mm covered a

large contact area on the tested specimens, the lowest coefficients of friction

were observed for all tested specimens when a 4 mm counterface ball was

used. One factor affecting the contact area was the spatial texture density. Due

to the small spatial texture density of Si/SU-8/PFP, the contact area on

multiple asperities for Si/SU-8/PFP was large and the coefficient of friction

significantly increased compared to Si/SU-8/NFP. Though the calculated

contact area of a single asperity on the 2 mm counterface ball was small, the

coefficients of friction for the 2 mm counterface ball on Si/SU-8/NFP were

higher than that of the 4 mm and 6 mm counterface balls (except the 6 mm

counterface ball on Si/SU-8 and Si/SU-8/PFP). This may be because the small

2 mm counterface ball slides on the bottom surface of the furrows, enlarging

the contact area and increasing the adhesion forces and the coefficient of

friction. As a result of different coutnerface balls (2 mm, 4mm and 6 mm), the

contact area increases gradually with a significant increase in the contact

pressure. The protruding part of the fingerprint pattern can reduce the contact

area and the coefficient of friction.

Different sizes of sliding counterface balls can change the behavior of

friction. The real contact area between two sliding surfaces plays a primary

role in changing the tribological behavior and the real contact area can be

changed due to the spatial texture density. In conclusion, with the small 2 mm

counterface ball on the textured surfaces, reducing the spatial texture density

decreases the coefficient of friction. However, for the large 4 mm and 6 mm

counterface balls on the textured surfaces, increasing the spatial texture

Chapter 6

168

density decreases the coefficient of friction. Based on the experimental results

from the 4 mm and 6 mm counterfaces, when the spatial texture density

increases, the real contact area decreases on multiple asperities and the

adhesion component of friction decreases as well. Therefore, surface patterns

are a primary concern when improving friction and the sliding counter

surfaces play a major role in providing better friction behavior. Both surface

patterns and sliding counterface balls are major influences and both of them

are necessary to optimize tribological behavior.

Fig. 6.4. The summarized results of coefficients of friction versus different size

counterface balls on Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP under a normal load of


Table 6.1. The coefficients of friction versus the spatial texture density of the tested

specimens (Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP).

Samples

Dimensions (µm) Coefficient of friction

W

(µm)

L

(µm)

Density =

L

W

2 mm 4 mm 6 mm

Si/SU-8 - - - 0.28 0.2 1.01

Si/SU-8/NFP 226 424 0.53 0.241 0.08 0.12

Si/SU-8/PFP 150 360 0.42 0.2188 0.164 0.35

0.28 0.2

1.01

0.241 0.08 0.12

0.2188 0.164

0.35

0

0.4

0.8

1.2

1.6

2 mm 4 mm 6 mm Coef

fici

ent

of

fric

tion



Chapter 6

169

Table 6.2. The calculated contact area and contact pressure on the tested specimens

(Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP).

Samples

Contact area

[x10-9

m2]

Contact pressure

[MPa]

2 mm 4 mm 6 mm 2 mm 4 mm 6 mm

Si/SU-8 1.81 2.87 3.76 55.24 34.84 26.56

Si/SU-8/NFP 1.55 2.18 2.55 64.52 45.87 39.22

Si/SU-8/PFP 1.53 2.11 2.47 65.36 47.39 40.46

6.4.2. Wear tests with different size counterface balls on Si/SU-8, Si/SU-

8/NFP, and Si/SU-8/PFP under dry conditions (without PFPE nano-

lubricant)

Fig. 6.5 shows the coefficients of friction versus different counterface

balls for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP under dry conditions

(without PFPE nano-lubricant), a normal load of 100 mN, and a high

rotational speed of 500 rpm. In these investigations, different sizes of

counterface balls, such as 2 mm, 4 mm, and 6 mm, were used to investigate

friction and wear behavior. A normal load of 100 mN and a rotational speed of

500 rpm were conducted on the tested specimens. The wear data was recorded

with a sampling rate of 1 Hz on Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP

under dry sliding conditions.

With dry sliding conditions and a 2 mm counterface ball, the highest

coefficient of friction (~0.52) was observed on Si/SU-8. The smallest

coefficient of friction occurred on Si/SU-8/NFP while testing with a high

rotational speed of 500 rpm. However, Si/SU-8/PFP had a coefficient of

friction of 0.35. When using a 4 mm counterface ball, improved frictional

behaviors were observed for all tested specimens. During the 4 mm

Chapter 6

170

counterface ball tests, the coefficients of friction for Si/SU-8, Si/SU-8/NFP,

and Si/SU-8/PFP were lower than that of the 2 mm and 6 mm counterface

balls. Si/SU-8/NFP observed the smallest coefficient of friction of 0.07 when a

4 mm counterface ball was used with a high rotational speed of 500 rpm.

However, with the 4 mm counterface ball, the coefficients of friction between

Si/SU-8 and Si/SU-8/PFP were similar. The coefficients of friction for Si/SU-

8 and Si/SU-8/PFP were 0.18 and 0.19, respectively.

When the diameter of the counterface ball was increased to 6 mm and

the specimens were tested with a high rotational speed of 500 rpm, Si/SU-8

demonstrated a coefficient of friction of 0.75 and Si/SU-8/PFP had a

coefficient of friction of 0.34. A coefficient of friction of 0.21 was indicated

for Si/SU-8/NFP. When the diameter of the counterface ball was increased to

6 mm, the contact area between the two sliding surfaces also increased. Due to

the increased contact area, the adhesion forces between the 6 mm counterface

ball and the sliding surface increased.

Based on the experimental results for both low (2 rpm in the friction

tests) and high (500 rpm in the wear tests) rotational speeds, the coefficients of

friction significantly increased with increasing speed, during tests with 2 mm

and 6 mm counterface balls. Si/SU-8 provided higher coefficients of friction

for all counterface balls compared to the SU-8 fingerprint textured surfaces

and the flat SU-8 surface provided a larger contact area for all of the

counterface balls. When comparing the experimental results between Si/SU-

8/NFP and Si/SU-8/PFP, Si/SU-8/NFP had the lowest coefficient of friction

Chapter 6

171

among the tested specimens at high rotational speeds. Moreover, due to the

presence of low spatial texture density for Si/SU-8/PFP, its coefficient of

friction was higher than that of Si/SU-8/NFP at high sliding speeds. It was

observed at a low sliding speed (friction tests) and a high sliding speed (wear

tests), that there was a strong relationship between the coefficient of friction

and the spatial texture density. The larger the spatial texture density, the lower

the contact area and coefficient of friction.

Fig. 6.6 shows the wear durability versus different counterface balls,

such as 2 mm, 4 mm, and 6 mm for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP

under dry conditions (without PFPE nano-lubricant), a normal load of 100 mN,

and a high rotational speed of 500 rpm. With the 2 mm counterface ball,

Si/SU-8 obtained a wear life time of 220 cycles. When the wear tests were

conducted on Si/SU-8/NFP and Si/SU-8/PFP, the wear lives of 1,500 cycles

and 1,050 cycles were observed, respectively. The wear life cycles of Si/SU-

8/NFP were higher than that of Si/SU-8/PFP and the smallest coefficient of

friction was observed for Si/SU-8/NFP, during investigations with the 2 mm

counterface ball.

Chapter 6

172

Fig. 6.5. The coefficients of friction versus different size counterface balls such as 2

mm, 4 mm, and 6 mm for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP under dry

conditions (without PFPE nano-lubricant) with a normal load of 100 mN and a


After increasing the diameter of the counterface ball to 4 mm, the wear

life cycles were increased to 300 cycles for Si/SU-8. The investigations with

the 4 mm counterface ball on Si/SU-8/NFP and Si/SU-8/PFP also showed that

the wear life cycles of Si/SU-8/NFP were higher than that of Si/SU-8/PFP.

The resultant values of wear durability for Si/SU-8/NFP and Si/SU-8/PFP

were 2,000 cycles and 1,800 cycles, respectively. For the tests on the 6 mm

counterface ball, the wear life cycles increased for all tested specimens. The

wear lives for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP increased to 600

cycles, 2,300 cycles, and 2,100 cycles, respectively. Among the results of

different counterface balls, the best wear lives occurred when a 6 mm

counterface ball was used on the tested specimens.

0.52

0.18

0.75

0.3

0.07

0.21

0.35

0.19

0.34

0

0.2

0.4

0.6

0.8

1

2 mm 4 mm 6 mm

Coef

fici

ent

of

fric

iton



Chapter 6

173

Fig. 6.6. The wear durability versus different size counterface balls such as 2 mm, 4

mm, and 6 mm for Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP under a dry condition

with a normal load of 100 mN and a rotational speed of 500 rpm.

Based on the experimental results under dry conditions, the SU-8

textured surfaces provided better wear life cycles for all of the counterface

balls. These observations showed that the positive asperities of the textures

(the fingerprint patterns) reduced the contact area, and these matters

influenced the coefficients of friction and the wear life cycles. It was also

observed for all tested specimens, that larger counterface balls had higher wear

durability. Although the wear life cycles with the 4 mm counterface ball were

lower than that of the 6 mm counterface ball, the minimum coefficients of

friction were observed with the 4 mm counterface ball for all tested specimens.

6.4.2.1. Analysis of wear mechanisms under dry conditions with

different size counterface balls

Fig. 6.7 shows the optical images of debris on the counterface balls

and wear tracks on the tested samples, such as Si/SU-8, Si/SU-8/NFP, and

Si/SU-8/PFP under a normal load of 100 mN and a high rotational speed of

220 300

600

1500

2,000

2300

1050

1,800

2100

0

1000

2000

3000

2 mm 4 mm 6 mm

Wea

r L

ife

Cycl

es

(No. of

Cycl

es)



Chapter 6

174

500 rpm. The wear tests were carried out under dry conditions with 2 mm, 4

mm, and 6 mm silicon nitride counterface balls. The investigations showed

that the wear tracks were seen on Si/SU-8 for all counterface balls and a wear

track was clearly observed under the 4 mm counterface ball. The wear debris

that accumulated along the wear track was generated on Si/SU-8 after ~220

cycles, 300 cycles, and 600 cycles under 2 mm, 4 mm and 6 mm counterfaces,

respectively.

There was a large wear trace on Si/SU-8 after conducting a test with

the 4 mm counterface ball. In addition, small wear traces occurred on Si/SU-8

under the wear tests of 2 mm and 6 mm. As the sliding advanced, SU-8

polymer film deteriorated more easily and wear particles were scattered

around the sharp wear trace and an increase in the material transferred to the

counterface ball surface was observed. It was clear from the counterface ball

that SU-8 polymer was transferred from the thin film to the counterface ball

for all wear tests with 2 mm, 4 mm, and 6 mm counterface balls. The

transferred SU-8 polymer debris greatly influenced the coefficient of friction

and the adhesion between the film and the counterface ball. Although there

was a large amount of polymer transferred from the SU-8 thin film to the 4

mm counterface ball, the coefficient of friction was as low as 0.18 with a high

rotational speed.

After patterning the fingerprint texture on Si/SU-8 and conducting tests

with different counterface balls, the optical images showed that a reduction in

wear debris occurred on Si/SU-8/NFP with 2 mm, 4 mm, and 6 mm

counterface balls. A small wear trace was observed on Si/SU-8/NFP when the

Chapter 6

175

6 mm counterface ball was used, and the widths of wear tracks from the tests

with 2 mm and 4 mm counterface balls were slightly larger than those caused

by the 6 mm one. On Si/SU-8/PFP, the wear debris was significantly

decreased for all counterface balls. Less wear particles accumulated on Si/SU-

8/PFP with the 2 mm and 6 mm counterfaces.

Fig. 6.7. The optical images of debris on the counterface balls and wear track on the

samples for (a) Si/SU-8, (b) Si/SU-8/NFP, and (c) Si/SU-8/PFP under 2 mm, 4 mm,

and 6 mm counterface balls.

The wear particles accumulated at the edges of the fingerprint and the

amount of wear debris was reduced on the fingerprint patterns. It was

observed that the width of the wear track was directly related to the amount of

polymer transferred from the wear track to the samples. Si/SU-8/NFP and

Si/SU-8/PFP provided the most optimal tribological behavior during all of the

counterface ball tests, compared to Si/SU-8. There are many factors of the

fingerprint texture that can provide a reduction in wear debris. First, the

rounded corners of the fingerprint pattern reduce the contact area between the

textured surfaces and the counterfaces. Second, the real contact area on

100 µm 100 µm

100 µm 100 µm

(a) Si/SU-8


(b) Si/SU-8/NFP

(c) Si/SU-8/PFP



Chapter 6

176

multiple asperities may decrease due to the increase in spatial texture density,

and the adhesion component of friction may decrease as well. Third, surface

textures help to reduce shear stress, and surface energy has a strong influence

on the reduction of material transfer.

6.4.3. Wear tests with different size counterface balls on Si/SU-8, Si/SU-

8/NFP, and Si/SU-8/PFP in lubricated conditions

Fig. 6.8 shows the coefficients of friction versus different size

counterface balls on Si/SU-8, Si/SU-8/NFP, and Si/SU-8/PFP under PFPE

nano-lubricated conditions. The wear durability of the tested specimens with

PFPE nano-lubricant, under a normal load of 100 mN and a high rotational

speed of 500 rpm is shown in Fig 6.9. The wear data were collected on the

tested samples with a sampling rate of 1 Hz, in order to determine the effects

of the PFPE nano-lubricant (0.2 wt% solution of PFPE Z-dol 4000 in H-

Galden ZV60) using different counterface balls.

Fig. 6.8. The coefficients of friction versus different counterface balls such as 2 mm,

4 mm, and 6 mm for Si/SU-8/PFPE, Si/SU-8/NFP/PFPE, and Si/SU-8/PFP/PFPE

under lubricated conditions (with PFPE nano-lubricant) with a normal load of 100

mN and a rotational speed of 500 rpm.

0.47

0.08

0.5

0.17

0.05 0.11

0.21 0.16 0.14

0

0.2

0.4

0.6

0.8

2 mm 4 mm 6 mm

Coef

fici

ent

of

fric

tion


Si/SU-8/PFPE Si/SU-8/NFP/PFPE

Chapter 6

177

For the wear tests with a thin lubricated layer, the observed coefficients

of friction were 0.47, 0.17, and 0.21 for Si/SU-8/PFPE, Si/SU-8/NFP/PFPE,

and Si/SU-8/PFP/PFPE, respectively, with the 2 mm counterface ball. A

reduction in the coefficient of friction was observed after coating PFPE nano-

lubricant on the tested specimens. However, the coefficient of friction for

Si/SU-8 was high with the 2 mm counterface ball; a coefficient of friction of

0.47 was obtained, which was greater than the coefficient of friction limit of

0.3. The wear life cycle for Si/SU-8/PFPE was 500 cycles. With the 2 mm

counterface ball, the lowest coefficient of friction was observed on Si/SU-

8/NFP/PFPE and the wear life cycles for Si/SU-8/NFP/PFPE were increased

by nearly fifteen times that of Si/SU-8/PFPE. A wear life of 5,800 cycles was

registered for Si/SU-8/PFP/PFPE. Moreover, improvements in the coefficients

of friction and wear life cycles were observed on Si/SU-8/PFP/PFPE, due to

the presence of the PFPE nano-lubricant.

While conducting tests with the 4 mm counterface ball, the smallest

coefficient of friction was observed on Si/SU-8/NFP/PFPE and the wear life

cycles were significantly increased to 60,000 cycles. The wear durability for

the negative fingerprint pattern with PFPE (Si/SU-8/NFP/PFPE) was twelve

times higher than that of the negative fingerprint pattern without PFPE (Si/SU-

8/PFPE). In the high speed test on Si/SU-8/PFP/PFPE, the highest coefficient

of friction (~0.16) was observed with the 4 mm counterface ball. The wear

durability of Si/SU-8/PFP/PFPE was half that of Si/SU-8/NFP/PFPE. With the

6 mm counterface ball, Si/SU-8/PFPE had a coefficient of friction of 0.5, the

highest coefficient of friction that was obtained. Si/SU-8/NFP/PFPE and

Chapter 6

178

Si/SU-8/PFP/PFPE demonstrated coefficients of friction of 0.11 and 0.14,

respectively. Although the coefficient of friction for Si/SU-8/PFP/PFPE was

higher than that of Si/SU-8/NFP/PFPE, the wear life for Si/SU-8/PFP/PFPE

was higher than that of Si/SU-8/NFP/PFPE.

Fig. 6.9. The wear durability versus different size counterface balls such as 2 mm, 4

mm, and 6 mm for Si/SU-8/PFPE, Si/SU-8/NFP/PFPE, and Si/SU-8/PFP/PFPE in

lubricated conditions with a normal load of 100 mN and a rotational speed of 500

rpm.

6.4.3.1. Analysis of wear mechanisms in lubricated conditions under

different size counterface balls

Fig. 6.10 shows optical images of wear tracks on the tested samples

and debris on the counterface balls for the wear tests of Si/SU-8/PFPE, Si/SU-

8/NFP/PFPE, and Si/SU-8/PFP/PFPE in PFPE nano-lubricated conditions with

counterface balls of 2 mm, 4 mm and 6 mm (conducted with a normal load of

100 mN and a high rotational speed of 500 rpm). The optical images showed

that the wear debris was significantly reduced with the presence of a PFPE

nano-coated layer. The SU-8 thin film layer without any protective coating

had a high coefficient of friction (0.75) and a low wear life (600 cycles) with

500

5,000

1,200

7,500

60,000 22,000

5,800

30,000 38,750

1

10

100

1000

10000

100000

2 mm 4 mm 6 mm

Wea

r L

ife

Cycl

es

(No. of

Cycl

es)

Different counterface balls

Si/SU-8/PFPE Si/SU-8/NFP/PFPE Si/SU-8/PFP/PFPE

Chapter 6

179

the 6 mm counterface. However, the wear life cycles were increased by two

times and the width of the wear track was significantly reduced with the 6 mm

counterface ball, after overcoating with PFPE nano-lubricant. In addition,

small wear tracks were formed on Si/SU-8/PFPE with the 2 mm and 4 mm

counterfaces.

While comparing two tested specimens of Si/SU-8/NFP/PFPE and

Si/SU-8/PFP/PFPE, better wear behaviors were observed on Si/SU-

8/NFP/PFPE for all counterface balls. Nevertheless, a wear track was formed

on Si/SU-8/NFP/PFPE with the 6 mm counterface and wear debris was

attached to the 6 mm counterface ball. Fig 6.10 (c) shows that a wear track

was formed on Si/SU-8/PFP/PFPE with the small counterface ball (2 mm).

The optical images of Si/SU-8/PFP/PFPE showed that no wear track was seen

and less wear debris was attached to the 4 mm and 6 mm counterface balls. A

maximum wear durability of 60,000 cycles was observed on Si/SU-

8/NFP/PFPE and then the wear track was small after overcoating with PFPE

nano-lubricant on the negative fingerprint pattern. Satyanarayana et al. [2005

and 2007] observed that a PFPE overcoating on SAMs coated Si enhanced the

wear durability. Then, Julthongpiput [2003] proposed the concept of PFPE

coating to improve the wear resistance of epoxy nanocomposite polymer

bilayers.

Fig 6.10 also demonstrates that the protective coating of PFPE

lubricant on the flat SU-8 surface and the SU-8 fingerprint textures provides

the best tribological behavior. The PFPE nano-lubricant did not easily

Chapter 6

180

generate wear debris, and the presence of a thin layer of PFPE lubricant on the

fingerprint pattern provided low friction, which was coupled with a lower

contact area. Based on the experiments, PFPE molecules contribute greatly to

improvements in wear durability. Satyanarayana et al. [2006] proved that a

PFPE layer on UHMWPE film reduces the coefficient of friction.

Perfluoropolyether serves as liquid lubricant and it can reduce shear stress and

friction and increase the wear life cycles by several times, to a few orders of

magnitude. Moreover, PFPE nano-lubricant can also provide as a lubricated

layer on the SU-8 fingerprint textures, due to its properties such as low surface

tension, low vapor pressure, lubricity, and chemical and thermal stability [Liu

and Bhushan 2003]. It is possible that the role of surface forces on the

fingerprint textured surfaces were changed, due to the presence of a nano-

lubricant layer on them.

Fig. 6.10. The optical images of debris on the counterface ball and wear track on the

samples for (a) Si/SU-8/PFPE, (b) Si/SU-8/NFP/PFPE, and (c) Si/SU-8/PFP/PFPE

under 2 mm, 4 mm, and 6 mm counterface balls.

100 µm 100 µm

4 mm

(a) Si/SU-8/PFPE

2 mm 2 mm 4 mm 6 mm 6 mm

100 µm 100 µm

(b) Si/SU-8/NFP/PFPE


(c) Si/SU-8/PFP/PFPE


Chapter 6

181

6.5. Summary

In Chapter 6, the contributions of Si/SU-8, Si/SU-8/NFP, and Si/SU-

8/PFP to the analysis of friction and wear properties are discussed.

Tribological properties were measured using a customized ball-on-disc

tribometer. These tested specimens were investigated with a normal load of

100 mN and a slow rotational speed of 2 rpm in the frictional test. In order to

analyze the wear behaviors, a high rotational speed of 500 rpm was conducted

on the tested specimens under dry and lubricated conditions. PFPE (0.2wt%

solution of PFPE Z-dol 4000 in H-Galden ZV60) was dip-coated on the tested

specimens at dipping and withdraw speeds of 2.1 mm/s, with a fixed dipping

duration of 30 seconds. The primary aim of this chapter is to analyze the effect

of friction and wear behaviors using different sliding counterfaces, such as 2

mm, 4 mm, and 6 mm Si3N4 counterface balls.

6.5.1. Frictional properties of the SU-8 tested specimens with different

counterface balls

The following conclusions are drawn from this research study:

1. In tests conducted with the 2 mm counterface, the smallest coefficient

of friction (~0.2188) was observed on Si/SU-8/PFP, and a small spatial

texture density of 0.42 provided the lowest coefficient of friction on

Si/SU-8/PFP. The fingerprint patterns (Si/SU-8/NFP and Si/SU-8/PFP)

slightly reduced the coefficients of friction with the 2 mm counterface.

The calculated contact area on a single asperity of Si/SU-8/PFP (using

Chapter 6

182

Hertzian’s contact method) was less than that of Si/SU-8 and Si/SU-

8/NFP.

2. In tests conducted with the 4 mm counterface, Si/SU-8/NFP obtained

the smallest coefficient of friction of 0.08. The highest coefficient of

friction was observed on Si/SU-8, due to the large contact area on a

single asperity, based on Hertzian’s contact calculation. A large spatial

texture density of Si/SU-8/NFP provided the lowest coefficient of

friction for a large counterface ball of 4 mm, when tests were

conducted on the textured surfaces.

3. When tests were conducted with the 6 mm counterface, Si/SU-8 had

the highest coefficient of friction (~1.01), which was greater than the

coefficient of friction limit in this experiment. When testing on the

textured surfaces, a large spatial texture density of Si/SU-8/NFP

showed a low coefficient of friction. Si/SU-8/PFP demonstrated a

coefficient of friction of 0.35 with a spatial texture density of 0.42.

Based on the experimental results from the different counterface balls,

the fingerprint patterns reduced the contact area and the friction. With the 2

mm counterface, decreasing the spatial texture density reduced the coefficient

of friction. However, with the large contact areas of the 4 mm and 6 mm

counterfaces, reducing the spatial texture density increased the coefficient of

friction. The real contact area decreases on multiple asperities when the spatial

texture density increases and the adhesion component of friction decreases.

Therefore, the spatial texture density is a primary concern when optimizing the

Chapter 6

183

tribological behavior of surface textures. Both surface patterns and sliding

counterfaces need to match for optimal tribological behavior.

6.5.2. Wear properties of the SU-8 tested specimens with different

counterface balls under dry and lubricated conditions

The following conclusions are drawn from this research study:

1. Under the dry condition with contact from the 2 mm counterface,

Si/SU-8 obtained the highest coefficient of friction (~0.52) and the

lowest wear durability (220 cycles). The lowest coefficient of friction

(~0.3) and the highest wear life cycles (1,500 cycles) were observed on

Si/SU-8/NFP, with a spatial texture density of 0.53. Wear tracks were

clearly observed on Si/SU-8 and Si/SU-8/NFP.

Under the PFPE lubricated condition with contact from the 2 mm

counterface, the coefficient of friction for Si/SU-8 was reduced to 0.47

and the wear life was increased to 5,000 cycles. Si/SU-8/NFP showed

wear durability up to 7,500 cycles, with the lowest coefficient of

friction value of 0.17. The analysis of worn surfaces showed that less

wear particles were observed after coating with PFPE nano-lubricant.

2. Under the dry condition with contact form the 4 mm counterface, the

lowest coefficient of friction (~0.07) was observed for Si/SU-8NFP.

The coefficients of friction for Si/SU-8 and Si/SU-8/PFP were similar.

The highest wear life cycles (2,000 cycles) were observed on Si/SU-

8/NFP and a reduction in wear particles was observed on Si/SU-8/NFP.

Chapter 6

184

Under the PFPE lubricated condition with contact from the 4 mm

counterface, Si/SU-8/NFP/PFPE provided the lowest coefficient of

friction (~0.05), and the wear durability of Si/SU-8/NFP/PFPE was

thirty times higher than that of Si/SU-8/NFP. However, the coefficient

of friction for Si/SU-8/PFP/PFPE was two times higher than that of

Si/SU-8/PFPE.

3. Under the dry condition with contact from the 6 mm counterface, the

presence of a large contact area on Si/SU-8 showed the highest

coefficient of friction and the lowest wear durability of 600 cycles.

The large spatial texture density of Si/SU-8/NFP helped to reduce the

coefficient of friction to 0.21 and increase the wear durability to 2,300

cycles. The friction and wear behaviors were significantly improved

with the 6 mm counterface ball after coating PFPE nano-lubricant on

the tested specimens, due to the presence of PFPE molecule layers on

the samples. These results demonstrate that the positive asperity of

surface texture (fingerprint pattern) can reduce the contact area and

friction, at high rotational speeds, under both dry and PFPE lubricated

conditions.

In conclusion, the negative fingerprint patterns with a spatial texture density

of 0.53 can provide excellent tribological behavior, with contact from the 4

mm counterface ball. Therefore, the coefficient of friction may occur at a

minimum spatial texture density.

Chapter 7

185

Chapter 7

Fabrication Methods and Tribological Analysis of

Positive and Negative Fingerprint Patterns on PDMS

(with and without PFPE nano-lubricant)

Based on the previous experimental results, the positive asperity of

surface texture provided better friction and wear behaviors of SU-8. The

protruding part of the surface texture reduced the contact area and the

coefficients of friction. This chapter presents the study of the tribological

behaviors of surface textures on PDMS. The negative and positive fingerprint

patterns were fabricated on PDMS. Friction and wear tests were conducted to

determine the behavior of surface textures on an elastomeric polymer. A

normal load of 100 mN was applied to the elastomeric specimens, with a

rotational speed of 2 rpm for the friction tests and a rotational speed of 500

rpm for the wear tests. The wear tests were also investigated under dry and

lubricated conditions.

Chapter 7

186


The advantages of the tribological behavior of the positive asperity on

Si/SU-8 textures are presented in Chapter 4 and Chapter 5. Currently,

modifying surface textures on a polymer is an effective method used in

tribology to reduce contact area, thus achieving better tribological behavior

and increasing the load carrying capacity. In this Chapter, the effect of positive

asperities of surface textures on the friction of another type of material

(polydimethylsiloxane), developed in MEMS and Bio-MEMS applications,

will be presented.

In the development of realistic applications, polydimethylsiloxane

(PDMS) is widely used, especially for biomedical applications. PDMS is a

silicone elastomeric and nontoxic material. It is also biocompatible and

composed of optically transparent and non-fluorescent material. One

elastomeric polymer, polydimethylsiloxane (PDMS), can be used as an

effective material, because it is easy to modify different surface patterns using

soft lithography and replica techniques. Moreover, a PDMS elastomer has a

low surface energy (surface energy ~ 22 – 25 mJ/m2) and non-wetting

(hydrophobicity) properties [Chaudhury and Whitesides 1991]. It is mainly

useful for the development of MEMS devices and bio-medical applications,

due to its properties such as biocompatibility, flexibility, ease of fabrication,

and low cost [Whitesides 2006 and Schneider et al. 2009].

In recent years, the friction on polymeric surfaces has been reduced by

the modification of surface patterns (nano/micro-patterns) [Yoon et al. 2006,

Chapter 7

187

He et al. 2008 and Singh et al. 2007, 2009 and 2011]. He et al. [2008]

mentioned that the reduction in the coefficient of friction (59% at macro-scale

tests and 38% at micro-scale tests) occurred on the PDMS elastomeric

surfaces, due to the presence of the pillar and groove patterns. Wakuda et al.

[2003] found that the tribological behaviors were sensitive, due to the size and

surface density of the micro-dimples on the ceramic plate.

Based on the research studies, surface morphologies with positive

(protruding) and negative (recessed) asperities had a great influence on the

contact area and friction forces. Investigations of friction coefficients for the

positive and negative asperities of different structures (hexagonal, circular,

square, diamond, and triangular cross-sections) on nickel indicated that the

coefficients of friction for the textured surfaces changed as the size of cross-

sections changed [Siripuram et al. 2004]. This research study showed that the

protruding patterns (fingerprint patterns) on Si/SU-8 provided better

tribological behaviors than the recessed patterns (honeycomb patterns) [Sandar

et al. 2013]. The protuberances on the polymeric surfaces reduced the contact

area and exhibited low friction. Furthermore, heterogeneous wetting occurred

due to the presence of protruding patterns on the surface; this effect caused a

reduction in surface energy.

In the experiments, PDMS that was supplied by Dow Corning Corp.,

USA was used to fabricate 3D PDMS negative and positive fingerprint

textures. The positive fingerprint and negative fingerprint patterns were

fabricated on PDMS using a 3D SU-8 negative fingerprint pattern (the detailed

Chapter 7

188

fabrication steps are outlined in Chapter 4) and a 3D PDMS positive

fingerprint pattern as the master piece moulds. The positive asperity of the

surface texture was attached to PDMS to further reduce the coefficient of

friction and extend the wear durability. In this chapter, the investigations of

friction and wear behaviors of the textured surfaces of PDMS will be

presented under unlubricated sliding conditions and lubricated sliding

conditions. In the unlubricated frictional tests, the friction behavior was

investigated on the PDMS tested specimens, using a customized ball-on-disc

tribometer with a normal load of 100 mN and a slow rotational speed of 2 rpm.

The wear tests were carried out with a high sliding speed of 500 rpm.

Moreover, perfluoropolyether (PFPE) was overcoated on the patterned surface

of PDMS in order to analyze the effect of the nano-lubricant layer on the

hydrophobic patterned surfaces. The analysis of surface wettability on the

PDMS textured surfaces were carried out. The friction behavior of the PDMS

textured specimens was determined using different normal loads, such as 50

mN, 100 mN, 150 mN, 300 mN, and 400 mN.

7.2. Sample preparation of the PDMS flat surface and the PDMS

textured surfaces

7.2.1. Fabrication process for the 3D PDMS positive fingerprint textures

To fabricate a 3D PDMS positive fingerprint pattern, the negative

fingerprint pattern on Si/SU-8 (Si/SU-8/NFP) was used as a master piece

mould; the fabrication process for Si/SU-8/NFP was presented in detail in

Chapter 4. Fig. 7.1(a) shows the fabrication steps of the 3D PDMS positive

Chapter 7

189

fingerprint textures. The fabrication procedures for 3D PDMS positive

fingerprint textures were as follows:

(1) PDMS prepolymer was mixed with a curing agent in a ratio of 10:1 in

a petri dish, and then the mixture of PDMS was stirred thoroughly for a

few minutes. In order to remove the bubbles, this mixture was placed

in a vacuum chamber for 20 minutes for degassing.

(2) This mixture was poured slowly onto the master piece mould of the

SU-8 negative fingerprint pattern (Si/SU-8/NFP).

(3) Bubbles were removed from the interface layer by placing the sample

in a vacuum chamber for 15 minutes.

(4) The curing process was carried out by placing the tested samples in an

oven at 70oC for 5 hours.

(5) The PDMS thick film layer was peeled from the master piece mould of

the SU-8 negative fingerprint pattern, in order to fabricate the PDMS

positive fingerprint pattern. The specimen for the PDMS positive

fingerprint pattern is referred to as PDMS/PFP.

7.2.2. Fabrication process for 3D PDMS negative fingerprint textures

In the fabrication of 3D PDMS negative fingerprint textures, the

positive fingerprint pattern on PDMS (PDMS/PFP) was used as a master piece

mould and the rest of the fabrication procedures for the 3D PDMS negative

fingerprint pattern were the same as those described in Section 7.2.1. The

PDMS/PFP mould was obtained as described in Section 7.2.1. The fabrication

Chapter 7

190

steps for the negative fingerprint pattern on PDMS are shown in Fig 7.1(b).

The specimen for the negative fingerprint pattern on PDMS is referred to as

PDMS/NFP.

7.2.3. Fabrication process for the PDMS flat surface

To obtain the PDMS flat surface, a mixture of PDMS prepolymer and

a curing agent (ratio of 10:1) was poured onto the cleaned silicon substrate.

The rest of the fabrication procedures for the PDMS flat surface were the same

as those described in steps (1), (3), and (4) of Section 7.2.1. The specimen for

the PDMS flat surface is referred to as PDMS.

7.2.4. PFPE coating on the PDMS flat surface, 3D PDMS positive

fingerprint patterns, and 3D PDMS negative fingerprint patterns

A 0.2 wt% solution of PFPE Z-dol 4000 in H-Galden ZV60 was

overcoated on the PDMS flat and textured surfaces. The dip-coating method

was used with a dipping and withdrawal speed of 2.1 mm/s and a dipping

duration of 30 seconds. The coated layer of PFPE on the negative fingerprint

and positive fingerprint patterns on PDMS are referred to as

PDMS/NFP/PFPE and PDMS/PFP/PFPE, and the PFPE coated layer of

PDMS/NFP and PDMS/PFP (not to scale) are shown in Fig. 7.2.

Chapter 7

191

Fig. 7.1. Fabrication procedures for (a) 3D PDMS positive fingerprint structure and

(b) 3D PDMS negative fingerprint structure.

(b)

Fabrication of PDMS negative

fingerprint pattern

(i) Pour the mixture of

PDMS on the master

piece of PDMS positive

fingerprint specimen and

remove the bubbles in a

vacuum chamber

(iii) Peel away from the master

piece of PDMS positive

fingerprint specimen

(ii) Cure at 70oC for 5 hours

PDMS

Si

SU-8/NFP

Si

Si

SU-8

UV

light

(a) Fabrication of SU-8 negative

fingerprint pattern

Fabrication of PDMS positive

fingerprint pattern

(i) Pour the mixture of

PDMS on the master

piece of SU-8

negative fingerprint

specimen and

remove the bubbles

in a vacuum chamber

(iii) Peel away from the

master piece of

SU-8 negative

fingerprint

specimen

(ii) Cure at 70oC

for 5 hours

PDMS

Chapter 7

192

Fig. 7.2. The PFPE coating layer (not to scale) of (a) PDMS/PFPE, (b)

PDMS/NFP/PFPE, and (c) PDMS/PFP/PFPE.


7.3.1. Surface characterizations

First, the optical images and surface dimensions of the 3D PDMS

negative fingerprint and positive fingerprint patterns were measured with a

Wyko NT1100 optical profiler and an optical microscope. Then, surface

wettability of the textures (negative fingerprint and positive fingerprint) on

PDMS were determined by measuring the contact angles with 0.5 µL of

deionized (DI) water in a VCA optima contact angle system (AST product,

Inc., USA). The standard deviations for the tested specimens were within 3o.

7.3.2. Tribological characterizations

The friction and wear tests were conducted on 3D PDMS negative and

positive fingerprint patterns with a customized ball-on-disc tribometer. The

measurements of the normal and lateral deflections of the cantilevers were

(a)

PDMS

PFPE

(b) PDMS/NFP

PFPE

PDMS/PFP

Chapter 7

193

recorded using laser displacement sensors, and then the recorded

measurements were converted to the normal load and friction force using a

calibration chart. The Si3N4 counterface ball (with a roughness of 5 nm) was

cleaned with acetone before the experimental tests, and the tested specimen

was placed onto the sample holder of the tribometer spindle. A 4 mm diameter

silicon nitride ball slid against the tested surfaces. The normal load and

rotational speed used for the friction tests were 100 mN and 2 rpm (linear

speed = 0.21 mm/s), respectively. The coefficient of friction was taken as an

average of five repeated experimental rotational tests. The rotation cycles (20

cycles) were recorded with a sampling rate of 10 Hz. In the experiments, when

the coefficient of friction exceeded 0.3, and fluctuation of the friction data

occurred, and/or a wear track was observed on the tested surface, the

experiments were stopped.

For the wear durability tests, a normal load of 100 mN and a rotational

speed of 500 rpm (linear speed = 52.36 mm/s) were applied to the tested

specimens and a 4 mm diameter silicon nitride ball was used as a counterface

ball. In these wear tests; a PFPE nano-lubricant layer was coated on the

untextured and textured surfaces to enhance the wear life cycles and to

observe the wear behaviors of the hydrophobic PDMS surfaces. The wear

experimental tests were carried out with rotational cycles of 50,000 cycles,

100,000 cycles, and 200, 000 cycles. The dimensions of wear track on the

tested specimens and wear mechanisms on the counterface ball were

investigated under an optical microscope.

Chapter 7

194

7.3.3. Different normal load tests on the PDMS textured surfaces

The effects of different normal loads on PDMS, PDMS/NFP, and

PDMS/PFP were studied with an Si3N4 counterface ball using loads of 50 mN,

100 mN, 150 mN, 300 mN, and 400 mN. A low sliding speed of 2 rpm (linear

speed = 0.21 mm/s) was used to record the friction data from the visco-

elastomeric polymer.



Fig. 7.3 shows the full dimensions (not to scale) and the optical images

of the negative fingerprint and positive fingerprint patterns on PDMS. The

dimensions of the 3D PDMS negative and positive fingerprint patterns were

measured under an optical microscope. The lengths between the two ridges of

PDMS/PFP and PDMS/NFP were 210 µm and 170 µm, respectively.

Moreover, the widths (ap and an) of the finger ridges for PDMS/PFP and

PDMS/NFP were 225 µm and 240 µm, and the depths (dp and dn) for

PDMS/PFP and PDMS/NFP were 40 µm and 35 µm, respectively.

7.4.2. Surface analysis

Table 7.1 presents the water contact angles of the tested specimens

with and without PFPE (PDMS, PDMS/PFP, PDMS/NFP, PDMS/PFPE,

PDMS/PFP/PFPE, and PDMS/NFP/PFPE). A small amount (0.5 µL) of

deionized (DI) water was ejected onto the tested surfaces using a syringe; the

contact angle was measured at the edge of water droplet on a surface, which

Chapter 7

195

was recorded with a microscope. The standard deviations for the tested

specimens were within 3o. The PDMS flat surface (PDMS) obtained a water

contact angle of 102o and the PDMS flat surface is known to be highly

hydrophobic. It was found that the water contact angle decreased by a small

amount, after patterning the negative and positive fingerprint patterns on

PDMS. Patterning of negative and positive fingerprints on PDMS

(PDMS/NFP and PDMS/PFP) demonstrated water contact angles of 99o and

100o, respectively. Fig. 7.4 shows the optical images of water contact angles

for PDMS, PDMS/NFP, and PDMS/PFP, which are captured under the

microscope of a VCA optima contact angle system.

Fig. 7.3. An optical image and a schematic diagram with full dimensions (not to

scale) of (a) the positive fingerprint pattern on PDMS (PDMS/PFP) and (b) the

negative fingerprint pattern on PDMS (PDMS/NFP).

PDMS/PFP

ap = 225 µm

L = 435 µm

W = 210 µm d

p = 40 µm

(a)

(b)

dn = 35 µm

PDMS/NFP

W = 170 µm

an = 240 µm

L = 410 µm

Chapter 7

196

Table 7.1. Sample nomenclature and the measured water contact angles (WCA) for

the PDMS flat surface and the PDMS textured surfaces with and without PFPE.

Materials Nomenclature WCA

(deg)

Textured surfaces without the PFPE nano-lubricant

PDMS polymer flat surface PDMS 102

PDMS polymer textured with negative

fingerprint pattern

PDMS/NFP 99

PDMS polymer textured with positive

fingerprint pattern

PDMS/PFP 100

Textured surfaces with the PFPE nano-lubricant

PDMS polymer flat surface with PFPE PDMS/PFPE 111

PDMS polymer textured with negative

fingerprint pattern with PFPE

PDMS/NFP/PFPE 109

PDMS polymer textured with positive

fingerprint pattern with PFPE

PDMS/PFP/PFPE 105

Fig. 7.4. Optical images of the water contact angles for PDMS, PDMS/PFP, and

PDMS/NFP.

Based on the previous research, lower water contact angles for the

fingerprint texture on Si/SU-8 surface were also observed compared to the flat

SU-8 surface [Sandar et al. 2013]. The hydrophobic properties of the PDMS

fingerprint textured surfaces were described by Cassie-Baxter [Cassie and

Baxter 1944] and Wenzel [Wenzel, 1936], who stated that the water drop

partially wet between the grooves of fingerprint patterns. Thus, the water

contact angles of the PDMS negative and positive fingerprint specimens were

slightly lower than that of the PDMS flat specimen. There were no

significantly different values of water contact angles between the flat surface

PDMS PDMS/PFP PDMS/NFP

Chapter 7

197

and the fingerprint patterns on PDMS. However, both the PDMS flat surface

and the PDMS fingerprint textured surfaces provided effective surface wetting

properties.

Surface energy was calculated using the contact angles of deionized

water, ethylene glycol, and hexadecane. Based on Young’s equation (from

Equation 3.1), the total apparent surface free energy is written in the

following form:

cosLVSLSV

where is the contact angle, is the surface energy, and the subscripts SLSV,

and LV are surface-vapor, surface-liquid, and liquid-vapor interfaces,

respectively. The graph of Fig. 7.5 represents the water contact angles and

surface energies of the tested specimens (PDMS, PDMS/PFP, and

PDMS/NFP). The surface energy of the PDMS flat specimen was 25.56 mJ/m2.

PDMS/PFP had a lower surface energy and water contact angle than PDMS

and PDMS/NFP. However, the apparent surface energies of PDMS/NFP and

PDMS/PFP were slightly reduced, as they were 22.74 mJ/m2 and 23.6 mJ/m

2,

respectively. With a contact angle greater than 90o, a low surface energy

occurred due to the hydrophobic properties of the PDMS unpatterned and

patterned surfaces. Compared to the PDMS flat specimen, which has shown a

contact angle of 102o, all PDMS fingerprint textured specimens with different

pitches have shown lower contact angles.

Chapter 7

198

Fig. 7.5. The water contact angles and surface energies of the PDMS flat surface and

the PDMS negative fingerprint and positive fingerprint patterns.

7.4.3. Friction behaviors of the PDMS flat surface and the PDMS textured

surfaces

The coefficients of friction versus the friction data of 20 rotational

cycles for PDMS, PDMS/NFP, and PDMS/PFP are shown in Fig. 7.6. A

normal load of 100 mN was applied to the PDMS tested specimens and the

normal and friction forces were recorded at a rotational speed of 2 rpm. The

friction data was recorded at a sampling rate of 10 Hz. The highest coefficient

of friction (~0.59) occurred on the PDMS flat surfaces while testing with a 4

mm Si3N4 counterface ball. Based on the experimental results, the coefficients

of friction for the PDMS flat, positive and negative fingerprint specimens were

0.59, 0.44, and 0.48, respectively. The results in Fig 7.6 showed that the

coefficients of friction for the PDMS negative and positive fingerprint surfaces

0

20

40

60

80

100

120


102 99 100

25.56 22.74 23.6

Samples

WCA(deg) Surface Energy(mJ/m2)

Chapter 7

199

were lower than that of the PDMS unpatterned surfaces. PDMS/NFP had a

slightly higher coefficient of friction than PDMS/PFP.

Due to the effect of fingerprint patterns on PDMS, the coefficients of

friction were significantly reduced and a reduction in the rate change of the

friction coefficient occurred. A coefficient of friction reduction rate of 25%

was obtained, after patterning the positive fingerprint texture on PDMS

(PDMS/PFP). Although the coefficient of friction was reduced, due to the

formation of the negative fingerprint pattern on PDMS, the reduction rate for

PDMS/NFP was 19%, less than that of PDMS/PFP.

The protuberances (fingerprint pattern) on PDMS surfaces can reduce

the contact area, which is a primary method for decreasing the friction force.

The contact area was calculated on a single asperity of surface texture based

on the Hertzian contact model. Table 7.2 and Table 7.3 show the dimensions

(width, length, and depth), spatial texture density, and the calculated contact

area that was based on a single asperity, using the Hertzian model for PDMS,

PDMS/PFP, and PDMS/NFP under a normal load of 100 mN. Based on the

calculation of contact area on a single asperity, the contact area for the PDMS

flat surface was large and the coefficient of friction was high at 0.59. However,

it was found that the protruding part of the fingerprint pattern on PDMS

reduced the contact area and the adhesion forces. The rounded corner of the

fingerprint patterns on the soft PDMS elastomeric polymer also decreased

contact area and friction force. The calculation of the contact area on a single

asperity of PDMS/NFP was low under a normal load of 100 mN compared to

Chapter 7

200

PDMS/PFP. However, the results showed that the coefficient of friction for

PDMS/NFP was slightly higher than that of PDMS/PFP. This may have

created a large contact area on multiple asperities because the width of the

negative fingerprint pattern was smaller than that of the positive fingerprint

pattern.

Table 7.2. The measured dimensions (width, length, and depth), spatial density, and

calculated contact area that were based on a single asperity using the Hertzian model

for PDMS, PDMS/PFP, and PDMS/NFP.

Materials/

Samples

Surface morphology Spatial

Density

(W/L)

Width

(W-µm)

Length

(L-µm)

Depth

(d-µm)

PDMS - - - -

PDMS/NFP 170 410 35 0.414

PDMS/PFP 210 435 40 0.483

Table 7.3. The calculated contact area that were based on a single asperity using the

Hertzian model for PDMS, PDMS/PFP, and PDMS/NFP.

Materials/

Samples

Contact

area

(10-9

m2)

Coefficient of

friction

Reduction rate in

coefficient of

friction

PDMS 733.35 0.59 -

PDMS/NFP 94.78 0.48 19 %

PDMS/PFP 139.47 0.44 25 %

The spatial texture density is also a good tool for analyzing the contact

areas on multiple asperities of polymer. Table 7.2 demonstrates that the

spatial texture density of PDMS/NFP (0.414) was smaller than that of

PDMS/PFP (0.483). Therefore, the contact areas on multiple asperities were

large for PDMS/NFP. These large contact areas may have increased the

adhesion forces between the two surfaces. Moreover, the coefficient of

friction for PDMS/PFP was low at slow sliding speeds because the high

Chapter 7

201

spatial texture density reduced the contact areas on multiple asperities. As the

spatial texture density increased, the contact area was reduced.

Fig. 7.6. The resulting coefficients of friction for the tested specimens (PDMS,

PDMS/NFP, and PDMS/PFP).

In 2008, He et al. showed that the coefficient of friction for the pillar

structure on PDMS was lower than that of the PDMS flat surface, and the

coefficient of friction for the pillar structure was reduced by about 38% during

the micro-scale tests. Li et al. [2011] investigated the coefficient of friction of

the dimple patterns on the PDMS under water lubrication and found that a

reduction in the coefficient of friction was observed at low speeds, due to the

presence of the surface texture (with dimple patterns); however, the coefficient

of friction increased at high speeds.

0.480.44

0.59

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7


Samples

Coef

fici

ent

of

fric

tion


Chapter 7

202

Fig. 7.7. The coefficients of friction versus the number of cycles for PDMS,

PDMS/PFP, and PDMS/NFP under a normal load of 100 mN and a rotational speed

of 2 rpm. Distinct stick-slip behaviors were observed for PDMS and PDMS/NFP.

As reported in Chapter 5, the recessed asperity of the honeycomb

pattern on Si/SU-8 provided a large contact area, and the sharp corners of the

honeycomb pattern created stress, causing the coefficients of frictions to be

higher. In this experiment, fluctuations in the friction curves occurred on the

soft elastomeric PDMS because the cantilever was over-swinging at the slow

sliding velocity, due to stick-slip motion [Overney et al. 1994 and Rozman et

al. 1996]; these fluctuations were reduced on the textured surfaces. In Fig. 7.7,

a great amount of fluctuation is seen on the PDMS flat surface and the PDMS

negative pattern surface. On the other hand, less fluctuation was observed on

the PDMS positive fingerprint pattern and high spatial texture density reduced

the contact area and demonstrated improved stick-slip behavior. The wear

debris formation on PDMS was reduced, due to the presence of the fingerprint

pattern. Thus, surface geometry is critical for improving tribological behavior,

and surface spatial density improves friction force. The rounded corner of the

0

0.2

0.4

0.6

0.8

0 2 4 6 8 10 12 14 16 18 20

Number of cycles

Co

effi

cien

t o

f fr

icti

on

(µ

)

PDMS/PFP PDMS PDMS/NFP

Chapter 7

203

protruding part provides better tribological behavior and reduces stick-slip

behavior on the elastomeric PDMS surfaces.

7.4.4. Wear behavior on the PDMS flat surface and the PDMS textured

surfaces under dry and lubricated conditions

7.4.4.1. Wear behavior under dry conditions

Fig. 7.8 shows the analysis of wear track dimensions on PDMS,

PDMS/NFP, and PDMS/PFP for 50,000 cycles, 100,000 cycles, and 200,000

cycles. In these wear tests, a normal load of 100 mN and a high rotational

speed of 500 rpm were applied to the PDMS tested specimens to record the

wear and friction data. The experiments were conducted with rotational cycles

of 50,000 cycles, 100,000 cycles, and 200,000 cycles. The dimensions of the

wear tracks were examined under the optical microscope for the rotation of

50,000 cycles, 100,000 cycles, and 200,000 cycles. In the wear tests of 50,000

cycles, the contact interface between the PDMS flat surface and the

counterface ball was noticeably worn out and a wear track was clearly seen on

the PDMS flat surface.

The average dimension of the wear track on the PDMS flat surface was

655 µm. Wear debris particles were attached to the counterface after the wear

tests on the PDMS flat surface; however, no wear track was observed on the

PDMS negative and positive fingerprint patterns in the wear tests of 50,000

cycles; thus, damage did not occur on the fingerprint patterns. Significant

improvements to wear durability occurred on the surface patterns of the

negative and positive fingerprints. The optical images of wear debris on the

Chapter 7

204

counterface ball and the wear track on the PDMS flat surface and the PDMS

negative and positive fingerprint patterns, under a normal load of 100 mN and

a rotational speed of 500 rpm, are shown in Fig. 7.9.

After the wear durability of 100,000 cycles, the dimension of the wear

track for the PDMS flat surface slightly increased. The average dimension of

the wear track (~670 µm) was observed on the flat surface after 100,000

cycles. The contact interfaces between the counterface ball and the PDMS

fingerprint patterns did not show any significant wear and no wear track was

observed on them. However, a small amount of wear particles were attached to

the counterface ball when investigating PDMS/NFP at 100,000 cycles. It was

observed that the protuberances (fingerprint pattern) improved wear

performance under dry conditions for high speed tests

The experimental results also demonstrated a large wear track on the

PDMS flat surface, after the wear durability of 200,000 cycles. The contacting

PDMS surface was highly damaged and the large wear track (~813 µm) on the

PDMS flat surface was investigated under the optical microscope. The PDMS

surfaces with the negative and positive fingerprint patterns were worn out by

increasing the wear test to 200,000 cycles, causing a noticeable wear track on

PDMS/PFP and PDMS/NFP. The wear debris was transferred to the

counterface ball while investigating the wear durability of 200,000 cycles. The

dimensions of the wear track for PDMS/PFP and PDMS/NFP were 320 µm

and 247 µm, respectively.

Chapter 7

205

Fig. 7.8. The dimensions (µm) of wear tracks for the tested specimens (PDMS,

PDMS/PFP, and PDMS/NFP) for rotational tests of 50,000 cycles, 100,000 cycles,

and 200,000 cycles.The experimental tests were carried out under a normal load of


In this experiment, the fingerprint textured surface on PDMS offered

better wear durability performance, and there was no failure of wear durability

until 100,000 cycles. The fingerprint ridges reduced the contact area and

improved the wear life and wear durability performance. However, for the

long test duration of 200,000 cycles on PDMS tested specimens, failures of the

fingerprint textured surfaces were found. The interfacial energy increased

between the counterface ball and the surfaces for the long duration tests,

resulting in potential failure of wear durability performance. The PDMS flat

surface had a higher interfacial energy value than the fingerprint textured

surfaces, and the fingerprint patterns improved the wear behaviors of PDMS.

Thus, surface texture is a primary component of tribology, and wear durability

performance can be improved for long duration tests by changing surface

patterns.

50,000100,000

200,000

PDMS/NFP

PDMS/PFP

PDMS

655

588

813

00

320

00

247

0

300

600

900

Wea

r tr

ack

dim

ensi

on

(um

)

No. of cycles (Cycles)

Sam

ple

s

PDMS/NFP PDMS/PFP PDMS

Chapter 7

206

Number of

cycles

Wear track

dimension

(µm)

Debris on the

counterface ball

Wear track on

the samples

50, 000 cycles

PDMS

(655 µm)

PDMS/PFP

(0 µm)

PDMS/NFP

(0 µm)

100, 000 cycles

PDMS

(670 µm)

PDMS/PFP

(0 µm)

PDMS/NFP

(0 µm)

200,000 cycles

PDMS

(813 µm)

PDMS/PFP

(320 µm)

Chapter 7

207

PDMS/NFP

(247 µm)

Fig. 7.9. Optical images of wear tracks on the tested specimens (PDMS, PDMS/PFP,

and PDMS/NFP) and debris on the silicon nitride counterface ball after 50,000

cycles, 100,000 cycles, and 200,000 cycles.The experimental tests were carried out

under a normal load of 100 mN and a rotational speed of 500 rpm.

7.4.4.2. Wear behavior under lubricated conditions

Fig. 7.10 shows the analysis of wear track dimensions on the nano-

lubricant of PDMS, PDMS/NFP, and PDMS/PFP for 50,000 cycles, 100,000

cycles, and 200,000 cycles. The wear behavior of the tested specimens was

analyzed under the optical microscope, after the wear data tests, with a normal

load of 100 mN and a high rotational speed of 500 rpm. After coating with

PFPE nano-lubricant and testing the wear durability for 50,000 cycles, the

wear track was significantly enlarged on the PDMS flat surface compared to

PDMS under dry conditions. The observed wear dimension was 850 µm. Due

to the fluorine atoms in PFPE, the wear track could be reduced. However, the

fingerprint pattern reduced the wear debris on PDMS/PFP/PFPE and

PDMS/NFP/PFPE, until the wear durability of 50,000 cycles was reached.

Though there was no wear track on the samples, a small amount of wear

debris was attached to the counterface ball in the investigations of 50,000

cycles.

When increasing the wear life cycles to 100,000, the failure of wear

behaviors was observed on all tested samples. The wear track dimensions were

markedly increased; they were 903 µm, 340 µm, and 350 µm for PDMS/PFPE,

Chapter 7

208

PDMS/PFP/PFPE, and PDMS/NFP/PFPE, respectively. In the wear durability

tests for 100,000 cycles, a wear track was also observed on the fingerprint

patterns that were overcoated with PFPE. The wear track dimensions on

PDMS/PFPE were significantly increased to 985 µm when the long wear

durability tests (200,000 cycles) were conducted. Wear failures were found in

long wear durability tests. The dimensions of wear tracks were 985 µm, 360

µm, and 700 µm for PDMS/PFPE, PDMS/PFP/PFPE, and PDMS/NFP/PFPE,

respectively. The dimensions (µm) of wear tracks for the tested specimens

with PFPE (PDMS/PFPE, PDMS/PFP/PFPE, and PDMS/NFP/PFPE) under

the rotational tests of 50,000 cycles, 100,000 cycles, and 200,000 cycles are

shown in Fig. 7.11.

The PFPE thin film that was coated on the PDMS polymer showed

larger wear track dimensions than the non-coated layer on PDMS. PDMS has

many positive qualities, including a low Young’s modulus, low surface energy,

ease of use, felxible fabrication capabilities by replica molding, low cost, and

biocompatibility. Although PDMS is an attractive material in development

work for realistic applications, the drawback of PDMS is the swelling induced

in the materials by many common solvents. Three aspects of the solvent

compatibility problem are: the solubility of a solvent in PDMS, the solubility

of solutes in PDMS, and the dissolution of PDMS oligomers in solvents are

potential contaminants. Lee et al. [2003] studied the interaction of PDMS with

many solvents, acids, and bases and they observed that PDMS swelled the

least in water, nitromethane, dimethy sulfoxide, ethylene glycol,

perfluorotributylamine, perfluorodecalin, acetonitrile, and propylene carbonate;

Chapter 7

209

it swelled the most in diisopropylamine, triethylamine, pentane, and xylenes.

Surface hydrophobicity was also achieved due to the presense of fluorocarbon

on the PDMS elastomer.

Number of

cycles

Wear track

dimension

(µm)

Debris on the

counterface ball

Wear track on

the samples

50, 000 cycles

PDMS/PFPE

(850 µm)

PDMS/PFP/

PFPE

(0 µm)

PDMS/NFP/

PFPE

(0 µm)

100, 000 cycles

PDMS/PFPE

(903 µm)

PDMS/PFP/

PFPE

(340 µm)

PDMS/NFP/

PFPE

(350 µm)

PDMS/PFPE

(985 µm)

Chapter 7

210

200,000 cycles

PDMS/PFP/

PFPE

(360 µm)

PDMS/NFP/

PFPE

(700 µm)

Fig. 7.10. Optical images of wear tracks on the tested specimens with PFPE

(PDMS/PFPE, PDMS/PFP/PFPE ,and PDMS/NFP/PFPE) and debris on the silicon

nitride counterface ball after 50,000 cycles, 100,000 cycles, and 200,000 cycles. The

experimental tests were carried out under a normal load of 100 mN and a rotational

speed of 500 rpm.

Fig. 7.11. Dimensions (µm) of wear tracks for the tested specimens with PFPE

(PDMS/PFPE, PDMS/PFP/PFPE, and PDMS/NFP/PFPE) for the rotational tests of

50,000 cycles, 100,000 cycles, and 200,000 cycles.The experimental tests were

carried out under a normal load of 100 mN and a rotational speed of 500 rpm.

Based on the XPS analysis of the PFPE coated layer on PDMS (in Fig.

7.12), the wide scan spectrum of XPS showed fluorine atoms on the tested

specimen compared to PDMS samples without PFPE. Moreover, the

PDMS/PFP/PFPE

PDMS/NFP/PFPE

PDMS/PFPE

0

200

400

600

800

1000

50,000100,000

200,000

0

340360

0

350

700

850 903985

Wea

r tr

ack

dim

ensi

on

(μ

m)

PDMS/PFP/PFPE PDMS/NFP/PFPE PDMS/PFPE

Chapter 7

211

flurocarbon group (PFPE) films did not swell properly on the surface of the

PDMS elastomer. The PFPE layer did not demonstrate effective wear beahvior

on PDMS and significant wear tracks were observed. Therefore, the PFPE thin

layer is not suitable for coating on a PDMS elastomer, and it cannot improve

friction and wear behaviors.

Fig. 7.12. XPS wide spectrum for (a) PDMS/PFPE, (b) PDMS/NFP/PFPE, and (c)

PDMS/PFP/PFPE.

7.4.5. Effects of different normal loads

Fig. 7.13 shows the effects of applied normal loads on the friction of

the PDMS negative and positive fingerprint pattern specimens. Different

normal loads of 50 mN, 100 mN, 150 mN, 300 mN, and 400 mN were applied

to the tested specimens to investigate the friction data for the effects of

different normal loads. A rotational speed of 2 rpm and a sampling rate of 10

Hz were used to record the friction data. The highest friction coefficient of

1.13 was observed at 50 mN for PDMS/NFP. The coefficients of friction on

PDMS/NFP markedly decreased between 50 mN and 150 mN, and nearly

constant coefficients of friction were observed between 150 mN and 400 mN.

PDMS/NFP observed the lowest coefficient of friction (~ 0.44) at the

normal load of 150 mN. It was found that the coefficient of friction for

PDMS/NFP/PFPE PDMS/PFPE PDMS/PFP/PFPE

Fluorine (1691.2 cps eV) Fluorine (4074.2 cps eV) Fluorine (1938.8 cps eV)

(a) (b) (c)

Chapter 7

212

PDMS/PFP had the highest value at the high load of 400 mN. The coefficient

of friction for PDMS/PFP was slightly reduced from 0.3975 to 0.309 when the

normal load was increased from 50 mN to 150 mN (Range 1), and the

coefficients of friction were nearly constant between 150 mN and 300 mN

(Range 2). The coefficients of friction increased after the normal load of 300

mN (Range 3). The lowest coefficient of friction on PDMS/PFP was obtained

at 150 mN. The coefficients of friction for PDMS/PFP were lower than that of

PDMS/NFP for all tested normal loads. The results showed that the

coefficients of friction reduced with increasing normal loads until a critical

load. Moreover, the negative and positive fingerprint patterns on PDMS had a

significant role in reducing coefficient of friction between the normal loads of

50 mN and 150 mN.

Based on the previous results from the negative fingerprint pattern on

Si/SU-8, a decreasing trend of friction coefficients with different normal loads

was observed between 50 mN and 150 mN [Sandar et al. 2013]. The geometry

effect of surface textures (pillar and groove) on PDMS greatly reduced the

coefficients of friction with increasing normal loads [He et al. 2008]. Unal et

al. [2003] found that for pure PTFE and its composites, the coefficients of

friction decreased with increasing normal loads. Moreover, the coefficients of

friction decreased with increasing normal loads according to:

Equation 2.25: )1( nkW .

Stuart [1997] observed that the coefficients of friction for PEEK with

chloroform decreased until a critical load was reached, and thereafter, the

coefficients of friction dramatically increased with increasing normal loads.

Chapter 7

213

Several researchers have found that the friction coefficients of polymers

sliding against metal decreased with increasing normal loads [Unal et al. 2003

and Sirong et al. 2007]. Different observations were made by investigations of

the coefficient of friction with different normal loads, and coefficients of

friction for polymers (PTFE, POM, and PEI) increased with increasing normal

loads, due to the critical surface energy of these polymers [Unal et al. 2003].

Based on our experiments, for both the negative and positive

fingerprint patterns on PDMS, mixed behavior (a decreasing and increasing

trend) of friction coefficients was observed between the tested normal loads of

50 mN and 400 mN. A reduction in the coefficient of friction (Range 1) was

observed for both negative and positive fingerprint patterns when the normal

loads were not high (between 50 mN and 150 mN) and the proportionality

between the coefficient of friction. The normal load for PDMS/NFP and

PDMS/PFP followed Equation 2.25, until the critical load of 150 mN. After

the normal load of 150 mN (Range 2 – between 150 mN and 300 mN), the

coefficients of friction were nearly constant. Furthermore, the coefficients of

friction increased slightly for both PDMS/NFP and PDMS/PFP after a normal

load of 300 mN (Range 3), because of the critical surface energy and frictional

heat between the counterface ball and the PDMS textured surfaces. The real

contact area between the Si3N4 counterface ball and the PDMS fingerprint

textured surface increased with increasing normal loads. The shear stress may

have increased when the applied normal loads increased. However, surface

texture may reduce friction and maintain load carrying capacity. Therefore, the

coefficient of friction on PDMS was influenced by surface texture, and spatial

Chapter 7

214

texture density plays an important role in influencing the contact area and load

carrying capacity.

Fig 7.13. The coefficients of friction for the negative and positive fingerprint patterns

on PDMS versus the applied normal loads (50 mN, 100 mN, 150 mN, 300 mN, and

400 mN) at a constant speed of 2 rpm. The measurements were taken at 20 sliding

cycles.

7.5. Summary

The current chapter presented the effect of positive asperities on

PDMS on frictional forces. The negative and positive fingerprint patterns were

fabricated using PDMS positive and SU-8 negative fingerprint pattern moulds.

Friction studies with an Si3N4 counterface ball under a normal load of 100 mN

and a rotational speed of 2 rpm were carried out. It was found that PDMS/NFP

and PDMS/PFP could significantly reduce dry friction. In addition, reduction

rates of 19% and 25% for the coefficients of friction were observed for

PDMS/NFP and PDMS/PFP, respectively, compared to the PDMS flat surface.

The spatial texture densities of PDMS/NFP and PDMS/PFP were 0.414 and

0.483, respectively. PDMS/PFP had a low coefficient of friction with a high

spatial texture density (0.483). Reducing the spatial texture density increased

0

0.2

0.4

0.6

0.8

1

1.2

0 100 200 300 400 500

Normal Load (mN)

Coef

fici

ent

of

fric

tion

PDMS/NFP PDMS/PFP

Range 1 Range 2 Range 3

Chapter 7

215

the contact area on multiple asperities. According to the experimental results,

the spatial texture density played a primary role in determining the coefficients

of friction.

The wear studies that used a Si3N4 counterface ball with a normal load

of 100 mN and a high rotational speed of 500 rpm were carried out for 50,000

cycles, 100,000 cycles, and 200,000 cycles. The dimension of the wear track

on the PDMS flat surface markedly increased when the rotational cycles

increased. Both PDMS/NFP and PDMS/PFP significantly improved the wear

behavior and reduced the dimension of the wear track in the tests. No wear

failure occurred on either PDMS/NFP or PDMS/PFP for 100,000 cycles.

However, for the long wear tests of 200,000 cycles, there was a small amount

of wear debris on PDMS/NFP and PDMS/PFP. The protuberances of surface

patterns improved the wear behavior. In the wear studies with PFPE, the wear

track dimensions significantly increased and failure of wear behavior was

observed in the long tests of 100,000 cycles and 200,000 cycles. PFPE was

not suitable for coating on PDMS because fluorocarbons (PFPE) did not swell

PDMS.

A specific range of loads (50 mN to 400 mN) was investigated with

both PDMS/NFP and PDMS/PFP; for both, the coefficients of friction

decreased between 50 mN and 150 mN because of their contact area and

viscoelastic properties. The coefficients of friction were nearly constant

between 150 mN and 300 mN, and then they increased after a normal load of

300 mN. The small length between the two ridges of PDMS/NFP supported an

Chapter 7

216

increase in contact area on multiple asperities. The small spatial texture

density of PDMS/NFP provided high coefficients of friction for all normal

loads.

In conclusion, surface textures (fingerprint textures) can provide

effective friction and wear behavior on PDMS. Although the protruding part

of a surface pattern can reduce the real contact area, it is necessary to consider

spatial texture density, in order to fine-tune the textures. Moreover, the spatial

texture density reduces the coefficients of friction and it can alter the effects

of different normal loads.

Chapter 8

216

Chapter 8

Contact Area Analysis of the Textured Surfaces

Friction and wear behaviors are affected by contact area and contact

pressure. This chapter presents the calculations of the contact radius and the

contact area of surface texture, using the Hertzian contact theory. In order to

determine the contact behavior between the counterface ball and the sliding

surfaces, different surface geometries with positive and negative asperities

were used in the Hertzian contact calculation. Different counterface balls are

also involved in the analysis of the contact behavior of surface textures.

Chapter 8

217


The contact configuration at the conjunction between two sliding

surfaces is a vital point of control for friction and wear behaviors. In the

previous chapters, the friction and wear behaviors were investigated after

modifying surface patterns (positive asperity and negative asperity) on

polymers, such as SU-8 and PDMS. Surface patterns with negative and

positive fingerprints, micro-dots, and honeycombs were developed on SU-8.

Based on the experimental results, a reduction in friction occurred in the

presence of a positive asperity on SU-8. However, the negative asperity with

poles provided a higher coefficient of friction than the flat SU-8 surfaces. Due

to positive results from studies on positive asperity in tribology [Tay et al.

2011 and Singh et al. 2011], the positive asperities of the negative fingerprint

and positive fingerprint patterns were also developed on an elastomeric

polymer, PDMS, to observe their behavior. The protruding part of the

fingerprint patterns reduced the contact area and friction.

To determine the effect of contact area, the analysis of a Hertzian

contact on a single asperity is presented in this Chapter. The geometrical

effects on the elastic deformation properties have been considered with the

“Hertzian Theory of Elastic Deformation”. This Hertzian contact theory

relates the contact area of a sphere with a plane surface or the contact area

between two spheres to the elastic deformation properties of the materials. The

surface interactions, Van der Waals interactions or adhesive contact

interactions, are neglected in this Hertzian contact theory. The calculation of

Chapter 8

218

the contact radius and contact area on SU-8 and PDMS are analyzed and

surface patterns such as the negative fingerprints, positive fingerprints, micro-

dots, and honeycombs are included in the analysis. In this contact calculation,

the properties of SU-8 material, PDMS material, and the Si3N4 counterface

balls are presented in Table 8.1.

Table 8.1. The properties of SU-8 material, PDMS material, and the Si3N4

counterface ball.

Materials Young’s modulus (GPa) Poisson’s ratio

SU-8 5.25 0.22

PDMS 0.001 0.5

Si3N4 310 0.22

8.2. Calculation of the contact radius )(a and the area of contact )(A

on Si/SU-8

8.2.1. Analysis of the contact radius a and the area of contact A on the

negative fingerprint pattern of Si/SU-8

Fig. 8.1 shows a schematic presentation of single asperity contact

between the negative fingerprint pattern and the counterface ball (not to scale).

The radius of curvature was obtained by fitting a circle to the ridge geometry.

In this analysis of the negative fingerprint pattern on Si/SU-8, the Hertzian

contact theory was applied to obtain the radius of contact area )(a and the area

of contact )(A .

The width of the base of the finger ridge was 260 µm and the width of

the top surface of the finger ridge was 150 µm. The depth of the arc h was

measured as 1.5 µm in, Fig. 8.1, and these values were used in the calculation

Chapter 8

219

process for all loads (50 mN, 100 mN, 150 mN, 300 mN, and 400 mN) to

obtain the radius 2R of the circle. The radius 2R is determined by:

ZrhR 2 (8.1)

For the negative fingerprint pattern:

ZrR NFP µ5.12 (8.2)

Using the properties of the triangles, one may write:

2222 µ130 rR NFP (8.3)

2222 µ75µ5.1 rR NFP (8.4)

By solving Equation 8.3 and Equation 8.4, the value r obtained was 3.7576

mm. After substituting the value r into Equation 8.3, the radius of the fitted

circle NFPR 2 was calculated as 3.7598 mm. The effective radius of NFPR for

Si/SU-8/NFP was then calculated as follows:

NFPNSiNFP RRR

21

111

43

(8.5)

mm 3055.1NFPR

According to Hertz’s equation, the radius of the elastic contact a between

two spheres is:

Chapter 8

220

31

9

3-31

* 10 * 5.425 *4

10 * 1.3055 * 3

4

3

nNFPn F

E

RFa (8.6)

where

43

43

2

8

28

*

111

NSi

NSi

SU

SU

EEE

; by substituting the value of Young’s

modulus and Poisson’s ratio (From Table 8.1): GPa 425.5* E .

Fig. 8.1. A schematic presentation of single asperity contact between the negative

fingerprint pattern and the counterface ball (not to scale). The radius of curvature

was obtained by fitting a circle to the ridge geometry.

r

R2

Z

h=1.5 µm

130 µm

R2

An imaginary circle is fit to find

out the radius of curvature of the

finger print ridge.

R1 = 2 mm

h=1.5 µm Z

r R2

260 µm

r

Fn

Si3N4

150 µm

Chapter 8

221

The calculated radii of contact and the area of contact for single

asperity contact, under the static condition using Hertz’s equation, versus the

applied normal loads for Si/SU-8/NFP are as follows in Table 8.2:

Table 8.2. The radius of contact and area of contact for single asperity contact under

the static condition using Hertz’s equation versus the applied normal loads for Si/SU-

8/NFP.

Load – nF (mN) Si/SU-8/NFP

Radius of contact

[x10-6

m]

Contact area

[x10-9

m2]

50 20.82 1.36

100 26.23 2.16

150 30.03 2.83

300 37.83 4.5

400 41.64 5.45

8.2.2. Analysis of the contact radius and the area of contact A on the

positive fingerprint pattern of Si/SU-8

A schematic presentation of single asperity contact between the

counterface ball and Si/SU-8/PFP (not to scale) is shown in Fig. 8.2. The

observed dimensions under the microscope were 210 µm and 150 µm, for the

width of the base of the finger ridge and the width of the top surface of the

finger ridge, respectively. The depth of the arc h was measured as 1.25 µm.

Using Equation 8.1, the radius PFPR 2 of the circle for Si/SU-8/PFP was:

ZrR PFP µ25.12 (8.7)

Using the properties of the triangles, one may write:

2222 µ105 rR PFP (8.8)

a

Chapter 8

222

2222 µ50µ25.1 rR PFP (8.9)

The value of PFPR 2 was 3.412 mm, after solving Equation 8.8 and

Equation 8.9. The effective radius of PFPR for Si/SU-8/PFP was then

calculated as:

NFPNSiPFP RRR

21

111

43

(8.10)

Therefore, 2609.1PFPR mm.

According to Hertz’s equation, the radius of the elastic contact between

two spheres for Si/SU-8/PFP was:

31

9

3-31

* 10 * 5.425 *4

10 * 1.2609 * 3

4

3

nPFPn F

E

RFa (8.11)

where ; by substituting the value of Young’s

modulus and Poisson’s ratio (from Table 8.1): .

The calculated radii of contact and area of contact for single asperity

contact, under the static condition using Hertz’s equation, versus the applied

normal loads for Si/SU-8/PFP are as follows in Table 8.3:

a

43

43

2

8

28

*

111

NSi

NSi

SU

SU

EEE

GPa 425.5* E

Chapter 8

223

Fig. 8.2. A schematic presentation of a single asperity contact between the positive

fingerprint pattern and the counterface ball (not to scale). The radius of curvature

was obtained by fitting a circle to the ridge geometry.



8/PFP.

Load – (mN) Si/SU-8/PFP

Radius of contact [x10-

6 m]

Contact area [x10-9

m2]

50 20.58 1.36

100 25.93 2.11

150 29.68 2.78

300 37.4 4.39

400 41.6 5.32

nF

50 µm

105 µm

h=1.25 µm

r

R2

Z

R2



fingerprint ridge.

R1 = 2 mm

h=1.25 µm Z

r R2

210 µm

r

Fn

Si3N4

100 µm

Chapter 8

224

8.2.3. Analysis of the contacta radius and the area of contact on

the honeycomb pattern of Si/SU-8


between the honeycomb pattern and the counterface ball (not to scale), in the

calculation of the radius of contact area a .

Fig. 8.3. A schematic presentation of single asperity contact between the honeycomb

pattern and the counterface ball (not to scale).

The effective radius of HCR for Si/SU-8/HC was determined with the

following equation:

HCNSiHC RRR

21

111

43

(8.12)

The radius value of HCR for Si/SU-8/HC was 2HCR mm.

According to Hertz’s equation, the radius of the elastic contact a between a

hard hemisphere and a flat surface is:

a A

R1 = 2 mm

Fn

2a

Chapter 8

225

31

9

331

* 10*425.5*4

10*2*3

4

3

nHCn F

E

RFa (8.13)

where

43

43

2

8

28

*

111

NSi

NSi

SU

SU

EEE

; by substituting the value of Young’s

modulus and Poisson’s ratio (From Table 8.1): .

The radii of contact and the Hertzian contact area were calculated for

the honeycomb pattern using the above analysis for each load used. Table 8.4

provides the calculated results of the radii of contact and the contact area.



8/HC.

Load – (mN) Si/SU-8/HC

Radius of contact

[x10-6

m]

Contact area

[x10-9

m2]

50 24 1.81

100 30.24 2.87

150 34.61 3.76

300 43.61 5.97

400 48 7.24

8.2.4. Analysis of the contact radius and the area of contact on the

micro-dot pattern of Si/SU-8

The width of the base of the micro-dot and the width of the top surface

of the micro-dot were 100 µm and 80 µm, respectively. Those dimensions

were included in the calculation of the contact area radius and the contact area

of the micro-dot patterns. The calculation method used for Si/SU-8/MD (400

µm) was the same as that used for Si/SU-8/NFP and Si/SU-8/PFP. Using

GPa 425.5* E

nF

a A

Chapter 8

226

Equation 8.1 and the triangle method, the value of MDR 2 was calculated as

0.564 mm. By substituting the value of MDR 2 , the value of MDR was 0.434

mm.

The radius of the elastic contact between a hard hemisphere and a flat

surface, using Hertzian contact was determined by:

31

9

331

* 10*425.5*4

10*439.0*3

4

3

nMDn F

E

RFa

(8.14)

where ; by substituting the value of Young’s

modulus and Poisson’s ratio (From Table 8.1): .


between Si/SU-8/MD (400 µm) and the counterface ball (not to scale). Table

8.5 provides the radii of contact, the Hertzian contact area, and the contact

area.



8/MD (400 µm).

Load – (mN) Si/SU-8/MD(400 µm)


6 m]

Contact area [x10-9

m2]

50 14.48 0.65

100 18.24 1.05

150 20.88 1.37

300 26.31 2.17

400 28.96 2.63

a

43

43

2

8

28

*

111

NSi

NSi

SU

SU

EEE

GPa 425.5* E

nF

Chapter 8

227

Fig. 8.4 A schematic presentation of a single asperity contact between Si/SU-8/MD

(400µm) and the counterface ball (not to scale).

40 µm

50 µm

h=0.8 µm

r

R2

Z

R2



fingerprint ridge.

R1 = 2 mm

h=0.8 µm Z

r R2

100 µm

r

Fn

Si3N4

80 µm

Chapter 8

228

8.3. Calculation of the contact area radius and the area of contact

on PDMS


PDMS flat surface

Fig. 8.5 shows a schematic presentation of a single contact between

the PDMS flat surface and the counterface ball (not to scale). The calculated

radius of PDMSR for the PDMS flat surface was obtained as 2PDMSR mm.

Using Young’s modulus and Poisson’s ratio of PDMS and Si3N4 (from Table

8.1), the calculated value of *E was 1.33 MPa.

Fig. 8.5. A schematic presentation of single asperity contact between the PDMS flat

surface and the counterface ball (not to scale).

The radius of the elastic contact between a hard hemisphere and a

flat surface, using Hertzian contact, is determined by:

31

6

331

* 10*33.1*4

10*2*3

4

3

nPDMSn F

E

RFa (8.15)

)(a

)(A

a A

a

Fn

PDMS

R1 = 2 mm

2a

Chapter 8

229

The resulting data of the elastic contact radius and the area of contact between

the PDMS flat surface and the counterface ball under different normal loads

are shown in Table 8.6.

Table 8.6. The radius of contact and area of contact for single asperity contact at the

static condition using Hertz’s equation versus the applied normal loads for PDMS.

Load – (mN) PDMS

Radius of contact

[x10-6

m]

Contact area

[x10-9

m2]

50 383.47 461.97

100 483.15 733.35

150 553.07 960.97

300 696.82 1274.04

400 766.95 1847.92


PDMS positive fingerprint surface


between the PDMS positive fingerprint pattern and the 4 mm counterface ball

(not to scale). Using the properties of the triangles, one may write:

2222 µ5.112 rR PFPPDMS (8.16)

2222 µ39µ40 rR PFPPDMS (8.17)

The value of PFPPDMSR 2 was 163.94 µm, after solving Equation

8.16 and Equation 8.17. The effective radius of PFPPDMSR for PDMS/PFP

was then calculated by:

PFPPDMSNSiPFPPDMS RRR

21

111

43

(8.18)

nF

a A

Chapter 8

230

Therefore, 52.151PFPPDMSR µm.

According to Hertz’s equation, the radius of the elastic contact between

two spheres for PDMS/PFP is determined by:

31

6

6-31

* 10 * 1.33 *4

10 * 151.52 * 3

4

3

nPFPPDMSn F

E

RFa (8.19)

where

43

43

22

*

111

NSi

NSi

PDMS

PDMS

EEE

; by substituting the values of Young’s

modulus and Poisson’s ratio for PDMS (from Table 8.1): 33.1* E MPa.

The calculated radii of contact and the area of contact for single


applied normal loads for PDMS/PFP are demonstrated in Table 8.7:

Table 8.7. The radius of contact and the area of contact for single asperity contact

under the static condition using Hertz’s equation versus the applied normal loads for

PDMS/PFP.

Load – (mN) PDMS/PFP

Radius of contact

[x10-6

m]

Contact area

[x10-9

m2]

50 162.26 82.71

100 204.44 131.31

150 234.02 172.05

300 294.85 273.12

400 324.52 330.85

a

nF

Chapter 8

231

Fig. 8.6 A schematic presentation of single asperity contact between the PDMS

positive fingerprint pattern and the counterface ball (not to scale).


PDMS negative fingerprint surface


between the PDMS negative fingerprint pattern and the 4 mm counterface ball

(not to scale). Based on Fig. 8.7, using the triangle method, one may write:

a A

39 µm

112.5 µm

h=40 µm

r

R2

Z

R2

78 µm

R1 = 2 mm

h=40 µm

225 µm

r

Fn

Si3N4

PDMS



fingerprint ridge.

Z

r R2

Chapter 8

232

2222 µ85 rR NFPPDMS (8.20)

2222 µ5.42µ35 rR NFPPDMS (8.21)

After solving Equation 8.20 and Equation 8.21, the calculated value of

NFPPDMSR 2 was 103.96 µm. The effective radius of NFPPDMSR was 98.83

µm for PDMS/NFP, obtained by substituting into Equation 8.22:

NFPPDMSNSiNFPPDMS RRR

21

111

43

(8.22)

According to Hertz’s equation, the radius of the elastic contact

between two spheres for PDMS/NFP is:

31

6

6-31

* 10 * 1.33 *4

10 * 98.83 * 3

4

3

nNFPPDMSn F

E

RFa

(8.23)

where ; by substituting the values of Young’s

modulus and Poisson’s ratio for PDMS (from Table 8.1): MPa.

Table 8.8 shows the calculated radii of contact and area of contact for single


applied normal loads for PDMS/NFP.

a

43

43

22

*

111

NSi

NSi

PDMS

PDMS

EEE

33.1* E

Chapter 8

233

Table 8.8. The radius of contact and the area of contact for single asperity contact

under the static condition using Hertz’s equation versus the applied normal loads for

PDMS/NFP.

Load – (mN) PDMS/NFP


6 m]

Contact area [x10-9

m2]

50 140.72 62.21

100 177.3 98.76

150 202.95 129.4

300 255.71 205.42

400 281.44 248.84

Fig. 8.7 A schematic presentation of single asperity contact between the PDMS

negative fingerprint pattern and the counterface ball (not to scale).

nF

85 µm

R1 = 2 mm

h=35 µm

170 µm

r

Fn

Si3N4

PDMS



fingerprint ridge.

Z

r R2

42.5 µm

85 µm

h=35 µm

r

R2

Z

R2

Chapter 8

234

8.4. Different counterface effects on the contact radius and the

area of contact on Si/SU-8/NFP and Si/SU-8/PFP

In this section, the values of the contact area radii and the area of

contact were calculated for different samples (Si/SU-8/NFP and Si/SU-8/PFP)

using different counterface balls (2 mm, 4 mm, and 6 mm). The calculation

methods are the same as those used in the above sections.

Tables 8.9 to 8.10 show the radii of contact area and the area of contact

that was calculated on a single asperity.

a

A

Chapter 8

235

Table 8.9. The radius of contact and area of contact for single asperity contact under the static condition using Hertz’s equation versus the applied normal

loads with different counterface balls (2 mm, 4 mm, and 6 mm) for Si/SU-8/NFP.

Load–

(mN)

Si/SU-8/NFP

2 mm 4 mm 6 mm

Radius of

contact [x10-6

m]

Contact area

[x10-9

m2]

Radius of

contact [x10-6

m]

Contact area

[x10-9

m2]

Radius of

contact [x10-6

m]

Contact area

[x10-9

m2]

50 19.05 1.14 20.82 1.36 22.6 1.6

100 24 1.81 26.23 2.16 28.47 2.55

150 27.47 2.37 30.03 2.83 32.59 3.34

300 34.61 3.76 37.83 4.5 41.06 5.3

400 38.1 4.56 41.64 5.45 45.19 6.42

Table 8.10. The radius of contact and area of contact for single asperity contact under the static condition using Hertz’s equation versus the applied normal

loads with different counterface balls (2 mm, 4 mm, and 6 mm) for Si/SU-8/PFP.

Load–

(mN)

Si/SU-8/PFP

2 mm 4 mm 6 mm

Radius of

contact [x10-6

m]

Contact area

[x10-9

m2]

Radius of

contact [x10-6

m]

Contact area

[x10-9

m2]

Radius of

contact [x10-6

m]

Contact area

[x10-9

m2]

50 17.48 0.96 20.58 1.36 22.26 1.56

100 22.03 1.53 25.93 2.11 28.05 2.47

150 25.22 1.998 29.68 2.78 32.11 3.24

300 31.77 3.17 37.4 4.39 40.45 5.14

400 34.97 3.84 41.6 5.32 44.52 6.23

nF

nF

Chapter 9

236

Chapter 9

Numerical Simulations for SU-8 Textured Surfaces

Numerical simulation using the finite element method has helped to

solve problems with tribological behaviors. Surface texture with a positive

asperity was used to simulate the topographies of the SU-8 surfaces using

three dimensional (3D) finite element models.

Chapter 9

237


Numerical simulation is a very useful approach to predicting the

tribological behavior of surface texture. Numerical simulation using the finite

element method (FEM) is a versatile tool for resolving the stress, strain, and

reaction and contact forces regardless of the geometry of the bodies. The finite

element method has been used to explore new designs and understand contact

behavior and friction mechanisms. The aims of this chapter are to evaluate the

frictional behaviors and to compare and support the experimental results of

polymer textured surfaces.

In order to understand the frictional process and the nature of the

interactions between two surfaces, Tangena and Wijnhoven [1985]

investigated the 2D finite element models with a rigid asperity and an asperity

with an elastic-plastic material. They determined the normal and shear forces

and the friction coefficient, and they observed the normal force effects when

changing the adhesive friction on the interface between interacting asperities

from 0 to 0.1. Moreover, the shear force and the friction coefficient were

dependent on the radius of the rigid asperity. Komvopoulos [1988–1989]

presented a comprehensive finite element analysis of subsurface stress, and

deformation fields produced during normal contact conditions between a rigid

surface and a layered medium. He observed that the stress/strain field

produced in the compressed layered solid was mainly dependent on the ratio

of the layer thickness to the half-contact width )/( ah . Thus, the layer

Chapter 9

238

thickness and the mechanical properties of the layer and the substrate

significantly affected the contact pressure.

Kral et al. [1995] investigated the surface deformation characteristics

that resulted from the repeated elastic-plastic indentation of a half-space

covered with a harder and stiff layer using the finite element method. The high

pressure at the contact edge was promoted by thinner, stiffer, and harder layers,

and the surface stresses showed slight changes with subsequent load cycles for

the nonhardening material. Moreover, Kral and Komvopoulos [1996] studied

the surface stresses and strain fields while the indentation was implemented

and slid on the 3D finite element layered elastic-plastic half-space. It was

shown that the higher load case did not shake down to an elastic cycle and the

lower load cases shook down only in the substrate.

Faulker et al. [1998] evaluated the influence of a deformable indenter

and a rigid indenter on the contact pressures and stress and strains. The lowest

peak radial tensile stress for cracking of the film to the surface was observed

under the rigid indenter, and the maximum interfacial shear stress responsible

for debonding of the coating, increased as the flexibility of the indenter was

reduced. Ye and Komvopoulos [2003] performed a three dimensional finite

contact analysis of normal (indentation) and sliding contact of elastic-plastic

layered media, to clarify the role of residual stress in the surface layer and

identify the coefficient of friction at the contact region. It was observed that

the coefficient of friction and contact for sliding and indenting were mainly

Chapter 9

239

affected by stress in the surface layer. In addition, high friction sliding

promoted plasticity.

Researchers have focused on the analysis of surface geometry that has

a great influence on the tribological behaviors, although it is difficult to

simulate the roughness of a surface using the finite element method. Faulkner

and Arnell [2000] investigated the effects of surface roughness on the

coefficient of friction using the 3D finite element of combing two elastoplastic

hemispherical asperities into a statistical model. Ramachandra and Ovaret

[2000] examined the normal pressure distributions and the sub-surface stress

fields on continuous coating (flat surface) and discontinuous coating (sharp

edges, rounded edges and crown profiles) using a two dimensional numerical

model. Reduced normal pressure spikes were observed and a crowned profile

was present on the discontinuous coating. Moreover, it was observed that the

discontinuous coatings were an effective approach to reducing friction

between contacting bodies and reducing contact temperatures caused by

frictional heating.

Gong and Komvopoulous [2003] investigated the contact pressure

distribution, surface tensile stress, and surface equivalent plastic strain for

layered media using different meandered and sinusoidal surfaces and a two-

dimensional plane-strain finite element analysis of the normal and sliding

contact of elastic-plastic layered media was performed. Based on the finite

element solutions and using a best-fit approach, it was observed that the

pattern geometry has a strong influence on the maximum contact pressure. In

Chapter 9

240

2004, a three dimensional finite element model was developed with a rigid

sphere indenter [Gong and Komvopoulous, 2004] and a simulation was

completed after three complete loading cycles including indentation, sliding

and unloading. Moreover, Jackson and Green [2005] observed that based on

an analysis of a normalized 2D axis-symmetric finite element model, the fully

plastic average contact pressure or hardness varied with deformed contact

geometry.

Based on previous research studies that used the finite element method,

the contact pressure distributions, stress, and strain were primarily affected by

the surface patterns. The indention and sliding processes were carried out to

examine the frictional behavior. According to the previous studies, the

coefficient of friction on a textured surface has not been investigated for

rotations. In the present study, an investigation of coefficient of friction was

conducted based on a three dimensional finite element model, in order to

determine the effect of a positive asperity on an SU-8 polymer under the

rotational action.

9.2. Finite element method

The finite element method (FEM), a computational technique, is of

great importance for solving complex problems in mathematical and

engineering fields. FEM is widely used and applied to physical problems such

as stress analysis, fluid mechanics, diffusion and electrical and magnetic fields.

Although the finite element method has a drawback when solving problems in

complicated geometries, it can be used advantageously to find approximate

Chapter 9

241

solutions for boundary conditions and the behavior of the textured surfaces.

FEM can solve complex geometry problems by breaking into discrete

elements and easily modifying the surface geometry to obtain new

specifications.

Analysis of the finite element method is supported by commercial

computer programs such as ANSYS and ABAQUS software. Three basic

steps of finite element analysis, pre-processing, analysis and post-processing,

are carried out to analyze the surface contact mechanisms of the textured

surfaces. During pre-processing, surface geometry models are created using

computer aided design (CAD) software and then converted into ABAQUS

software to analyze the surface morphology. The surface geometry is mainly

composed of ‘elements’ that are connected at nodes. Two types of elements,

2D type modeling and 3D type modeling, are used to study the surface

mechanisms; 3D element type modeling can obtain more accurate results. The

material and structural properties are used in the finite element method. The

next step, ‘meshing’ of the surface geometry, is performed and a fine mesh is

supported to obtain a better solution. Finally, the results of simulation are

obtained using the finite element method in ABAQUS software.

The detailed finite element analysis steps in ABAQUS are:

(1) Pre-processing

a. Part – Create individual parts

b. Property – Create and assign material properties

Chapter 9

242

c. Assembly – Create and place all parts

d. Step – Define all analysis steps and the result

e. Interaction – Define the contact interaction

f. Load – Define and place all loads and boundary conditions

g. Mesh – Define the nodes and elements

(2) Analysis

a. Job – Submit the job for analysis

(3) Post-processing

a. Visualization – View and analyze the results

9.3. Modeling procedures

The numerical simulation of the patterned surfaces was carried out

using the finite element code, ABAQUS 6.11. In this study, the model was

developed based on the assumptions of Hertzian elastic contact. Fig. 9.1

shows the two-dimensional contact model of a flat surface under the usual

plane-strain assumption.

Chapter 9

243

Fig. 9.1. Contact problem geometry

9.3.1. Geometry of the 3D fingerprint model

Fig. 9.2 shows the three dimensional fingerprint patterns on SU-8 in

the XY plane. In this study, one half of the rigid sphere was modeled and the

deformable part of a finger pattern was created. The fingerprint model was

created with for a finger ridge of 260 μm in length, the top surface of the

finger ridge was 150 μm in length, the furrow was 226 μm in length and the

corner of finger ridge was 1.5 μm in depth. The SU-8 layer was assumed to

exhibit elastic material behaviors.

A discrete rigid sphere with a radius of 2 mm was modeled for the

rigid silicon nitride slider that slid on the fingerprint pattern. Fig. 9.3 shows

the geometry of a discrete rigid sphere. The parameters of fingerprint pattern

were obtained from the experiments. The finite element model used in this

research study consisted of a rigid sphere indenting and rotating over a

deformable textured surface.

Rigid Cylinder

Indenter

2a

d h Layer

Substrate ssE ,

llE ,

Chapter 9

244

Fig. 9.2. Three dimensional fingerprint patterns on SU-8.

.

Fig. 9.3. Three dimensional discrete rigid sphere.

9.3.2. Material properties

The fingerprint patterns were made from SU-8 material and the

patterns were in deformation modes with elastic behaviors. The rigid slider

was made from silicon nitride (Si3N4) material and the mechanical properties

of SU-8 and Si3N4 are listed in Table 9.1.

Table 9.1. Mechanical properties of SU-8 and Si3N4

Materials Young’s Modulus Poisson’s ratio

SU-8 5.25 GPa 0.22

Si3N4 310 GPa 0.22

9.3.3. Assembly and contact interaction between the fingerprint pattern and

the slider

To detect contact between the top surface of the fingerprint pattern and

the slider, the rigid slider was brought into contact with the fingerprint

patterned surface when the rotating contact was in a dry condition. In the

260 μm

226 μm 150 μm

1.684 mm

1.5 mm

80 μm

4 mm diameter

Chapter 9

245

current model, two general static steps, indenting and rotating were performed.

First, normal contact was simulated by advancing the slider toward the elastic

fingerprint pattern. In the second step, the slider was rotated to complete one

cycle (360o). Then, in order to define the interaction between the slider and the

fingerprint textured surface, the slider was set as a master surface and the top

layer of the fingerprint pattern was used as a slave surface. Fig. 9.4 shows the

assembly and contact behavior between the fingerprint pattern and the slider.

Moreover, to obtain more accurate results for the pressure and stress

concentration, a surface-to-surface discretization method was applied between

the slider and the contacting surface. The coefficient of friction was then set to

0.2.

Fig. 9.4. The assembly and contact behavior between the fingerprint pattern

and the indenter.

9.3.4. Loading and boundary condition

For the boundary condition, first, the base of the SU-8 deformable

surface was fixed with zero degrees of freedom. Then, the SU-8 top surface

freedom was set to any direction. The rigid slider was fixed as a rigid body

and two reference points were set, on the rigid slider and on the top layer of

SU-8. Second, a load of 100 mN was applied onto the top surface of the

fingerprint pattern was 100 mN and the slider was brought into contact with

the top layer of the surface in the negative Y axis. Third, the indenter rotated

Chapter 9

246

to complete one cycle of 360o and the velocity of the rotation was set 0.21

rad/time (2 rpm). Each analysis consisted of the thirty steps that were required

to complete one cycle. The effects of reaction force and stress concentration

were observed after rotating the slider on the SU-8 textured surface.

9.3.5. Meshing process of models

Fig. 9.5 shows the normal and rotating contact simulation that was

performed with the three dimensional finite element mesh. The mesh consisted

of three-dimensional 4-node linear hexahedral elements for the fingerprint

textured parts, and 4-node quadrilateral rigid elements for the slider. Different

numbers of elements were applied to the ridges and furrows of the fingerprint

pattern. The finite element mesh consisted of approximately 1,500 elements

for the small indenter part and 116,500 elements for the fingerprint pattern.

Fig. 9.5. Three-dimensional finite element mesh for the fingerprint textured pattern

and the slider.

Chapter 9

247

9.4. Results

In this section, the finite element results that were based on the

processes of indentation and rotation are presented and solutions are provided

for:

(1) The stress behavior of the patterned surface,

(2) The friction force and coefficient of friction that can be affected by

the patterned surface

The fingerprint texture parameters that were obtained from the

experiment, such as the width of the finger furrow, the length between two

finger ridges, and the size were included in this simulation. The quasi-static

contact simulations consisted of two sequential steps for loading and rotating

the slider on the fingerprint pattern. To study the effect of friction force and

the stress behaviors of the fingerprint textured surfaces, a friction coefficient

equal to 0.2 was used in this simulation. Then, in order to estimate the

coefficient of friction, the frictional force was divided by the normal load.

9.4.1. Analysis of the coefficient of friction on the fingerprint textured

surface

The contact behavior and stress concentration between the fingerprint

texture surface and the slider were investigated when the rotation of the slider

was rotated. In order to determine whether or not friction force correlates with

surface textures, ridges and furrows of fingerprint patterns were made on the

surface textures.

Chapter 9

248

Using a silicon nitride slider ball, a set of friction forces was taken.

The slider started indenting the corner of the ridges of the fingerprint pattern,

passing on the top surface, sliding down the grooves, and then rotating to

complete one cycle, stopping at the same position. The three selected friction

force curves (passing upwards a step, on a ridge, and passing down a step)

were chosen to understand the behaviors of friction on the fingerprint pattern.

From the sets of numerical simulations, a change of friction force and

coefficient of friction were observed. Based on the numerical simulation, the

average value of the friction coefficients was significantly decreased; and it

was observed at 0.03 for one cycle. The best coefficient of friction may have

occurred when the fingerprint pattern was placed on the SU-8.

9.4.1.1. Sliding up the step/the ridge

Based on the simulation result with the slider passing upwards on the

ridge, it was noted that the friction force increased and then the coefficient of

friction increased. Fig. 9.6 shows the coefficient of friction, curves and stress

concentration for the condition of sliding up the ridge that was carried out with

a normal load of 100 mN. The contact area between the slider and the

fingerprint pattern increased while sliding up the ridge. A high stress

concentration was obtained at the corner of the fingerprint and it helped to

increase the friction force and coefficient of friction.

9.4.1.2. Sliding down the step/ ridge

It was observed that the friction force and contact area decreased as the

slider passed down the ridge. In addition, the coefficient of friction decreased

Chapter 9

249

under the step-down condition. Fig 9.7 shows the coefficient of friction and

stress concentration curves for the condition of sliding down the ridge. Less

stress concentration was occurred while sliding down the ridge.

Fig. 9.6. The coefficient of friction and stress concentration curves for the condition

of sliding up the ridge.

Fig. 9.7. The coefficient of friction and stress concentration curves for the

conditon of sliding down the ridge.

9.4.1.3. Sliding on the top surface

When passing through the top surface of the fingerprint pattern, the

stick-slip effect occurred and the friction force on the top surface was higher

than it was at the positions of sliding up and sliding down. Fig. 9.8 shows the

0

0.05

0.1

0.15

0.2

0.25

0 50 100 150

Co

effi

cien

t o

f fr

icti

on

Rotation (Degree)

0

0.05

0.1

0.15

0.2

0.25

0 50 100 150

Coef

fici

ent

of

fric

tion

Rotation (Degree)

0

20

40

60

80

0 5 10 S

tres

s Step

Sliding up

Sliding down

Chapter 9

250

coefficient of friction curves for the condition on the top surface of the finger

ridge.

Fig. 9.8. The coefficient of friction curve for the condition on the top surface

of the finger ridge.

9.5. Discussion

Based on the friction force curves, the ridge of the fingerprint pattern

significantly influenced the friction force. The friction force also changed due

to the effect of stress concentration. On the other hand, the experiment in

Chapter 4 showed that the lowest coefficient of friction occurred on the

negative fingerprint pattern that was also carried out in the numerical

simulation. However, in the numerical simulation, a low coefficient of friction

of 0.03 was recorded for the fingerprint pattern. The average value of the

coefficients of friction from numerical simulation with the finite element

method was less than that of the experiments. Based on both the experiment

and the numerical simulation, the best coefficient of friction was observed on

0

0.05

0.1

0.15

0.2

0.25

0 50 100 150 200 250

Co

effi

cien

t o

f fr

icti

on

Rotation (Degree)

Chapter 9

251

the negative fingerprint pattern. The results of the numerical simulation helped

to explain.

Moreover, in the elastic deformation model, the friction force may be

caused primarily by the elastic deformation. While the slider passes down the

corner of the fingerprint pattern, the elastic deformation is reduced and the

contact area between the slider and the pattern is also reduced, because of the

relaxation of deformed material. In contrast, the elastic deformation increases

when the indenter slides upwards on the fingerprint pattern corner. Due to the

elastic properties, the effect of stick-slip was clearly observed.

It may be explained by another fact that the contact area of the ridges

was less than that of the flat SU-8 surface. Thus, the contact between the slider

and the pattern plays a primary role in reducing the coefficient of friction and

friction force. It is also important to mention that geometrical effects such as

height, width, and length of the pattern influence the coefficient of friction.

9.6. Conclusions of the numerical simulation

The three dimensional SU-8 fingerprint patterns and the rigid silicon

nitride indenter were developed in this numerical simulation using the finite

element code, ABAQUS 6.11. The main concerns in tribology, the friction

force and coefficient of friction, were investigated on the fingerprint pattern

under the rotating contact condition. It can be concluded:

(1) The average coefficient of friction observed was as 0.03.

Chapter 9

252

(2) The coefficient of friction and friction force increased at the step up

condition (passing upwards of the corner of the fingerprint pattern),

(3) The coefficient of friction and friction force decreased at the step down

condition (passing downwards of the corner of the fingerprint pattern).

It can also be concluded, based on the experiments and numerical

simulations that the parameters of the negative fingerprint texture were the

optimal values for the development on SU-8 and they provided the best

coefficient of friction.

Chapter 10

253

Chapter 10

Conclusions and Future Recommendations

This chapter presents the main conclusions from the present research

study and the future recommendations are included.

10.1. Conclusions

The main objectives of this thesis are to develop a new fabrication

technique for a fine tuned pattern (fingerprint pattern) on SU-8 and PDMS

polymers that are useful for MEMS and Bio-MEMS applications, to optimize

different surface textures (fingerprint, micro-dot and honeycomb) and enhance

the friction and wear properties of SU-8 and PDMS polymers, to understand

the positive and negative asperities of surface texture in tribology with regards

to the spatial texture density, to determine the effects of perfluoropolyether

(PFPE) on the textured specimens, to understand the effects of different

sliding counterfaces on friction and wear behaviours, to examine the real

contact area on a single asperity (using the Hertzian contact method) that can

help to optimize the friction, and to investigate the coefficient of friction using

numerical simulation with the finite element method.

The following conclusions are drawn from the present research:

I Surface texture effects on the coefficient of friction

Based on the experimental results, improved friction and wear

behaviors were observed on the negative fingerprint and positive fingerprint

Chapter 10

254

patterns in dry sliding tests, due to the presence of protrusions on the

fingerprint pattern. Optimal friction and wear behaviors were observed on the

negative fingerprint pattern (Si/SU-8/NFP).

Positive asperities on SU-8 (fingerprint and micro-dot patterns)

Si/SU-8/NFP showed the lowest coefficient of friction (~0.08),

compared to that of Si/SU-8 (~0.2) and Si/SU-8/PFP (~0.169).

The coefficient of friction for Si/SU-8/NFP was 2.5 times lower than

that of Si/SU-8.

The coefficient of friction (~0.198) at the optimized micro-dot pitch

length of 400 µm was slightly lower than that of the SU-8 untextured

surfaces (~0.2).

Negative asperities on SU-8 (honeycomb pattern)

The coefficient of friction of Si/SU-8/HC (~0.41) was almost two

times higher than that of the untextured surface (Si/SU-8).

II Spatial texture density on the coefficient of friction


The large spatial texture density (0.53) of Si/SU-8/NFP decreased the

real contact area on multiple asperities and reduced the friction.

Chapter 10

255

The small spatial texture density (0.42) of Si/SU-8/PFP provided a

large contact area on multiple asperities and a high coefficient of

friction was observed.

In the observation of different micro-dot patterns, the minimum

coefficient of friction (~0.198) was observed at the optimized spatial

texture density of 0.75 for a 400 µm pitch length with the micro-dot

pattern.

III Perfluoropolyether (PFPE) effects on friction and wear

Perfluoropolyether, an effective coated layer, reduced the coefficient of

friction and enhanced the wear life cycles of the textured surfaces compared to

the unlubricated specimens.


The best wear life of 60,000 cycles was obtained for Si/SU-

8/NFP/PFPE and the coefficient of friction of Si/SU-8/NFP/PFPE was

lower than that of Si/SU-8/NFP. The wear life of Si/SU-8/NFP/PFPE

was nearly thirty times higher than that of Si/SU-8/NFP.

The wear life significantly improved to 30,000 cycles for Si/SU-

8/PFP/PFPE.

The highest wear life cycles (25,000 cycles) were observed for Si/SU-

8/MD (500 µm)/PFPE among different pitch lengths with the micro-

dot pattern.

Chapter 10

256

Negative asperities on SU-8 (honeycomb pattern)

Overcoating of a PFPE top layer on the honeycomb pattern increased

the wear durability of 22,500 cycles; the wear life cycle was improved

by eighteen times for Si/SU-8/HC/PFPE.

IV Effects of different sliding counterfaces

The contact area was highly dependent on the friction when using

different sliding counterfaces and the friction and wear behaviors were altered

by the textured surfaces. The best friction and wear behaviors were observed

for the tested specimens such as Si/SU-8, Si/SU-8/NFP and Si/SU-8/PFP,

while using the 4 mm counterface ball on the tested specimens.

When using the large 4 mm and 6 mm counterface balls on the

textured surfaces, increasing the spatial texture density decreased the

coefficient of friction and Si/SU-8/NFP provided the lowest coefficient

of friction.

The small 2 mm counterface ball slid on the bottom surface of the

furrows, enlarging the contact area and increasing the adhesion forces

and the coefficient of friction.

V Numerical simulation with the finite element method

Numerical simulation with the finite element method has helped

researchers to understand the friction behavior and stress concentration of the

textured surfaces.

Chapter 10

257

The average coefficient of friction observed was as 0.03.

The coefficient of friction increased at the step upwards condition

(passing upwards at the corner of the fingerprint pattern).

The coefficient of friction decreased at the step down condition

(passing downwards at the corner of the fingerprint pattern)

VI Surface texture effects on PDMS

PDMS/NFP and PDMS/PFP reduced the coefficient of friction under

dry friction. Reduction rates of 19% and 25% for the coefficients of friction

were observed for PDMS/NFP and PDMS/PFP, respectively, compared to the

PDMS flat surface. Both PDMS/NFP and PDMS/PFP significantly improved

the wear behavior and reduced the dimension of the wear track in the tests.

However, in the wear studies with PFPE, the wear failures were observed in

the long tests of 100,000 cycles and 200,000 cycles and PFPE was not suitable

for coating on PDMS.

10.2. Future work

For mass production, it may not be suitable to use a finger as a mould.

However, the fingerprint pattern on PDMS could be used as master piece and

then the fingerprint pattern could be fabricated on SU-8 using polymer jet

printing. According to the experiments and numerical simulation, the contact

area plays a vital role in affecting on the tribological behaviors. Further

development should be carried out including;

Chapter 10

258

1. Use of other sliding conditions, such as pin-on-disc or flat-on-disc

conditions, to investigate the tribological behaviors.

2. Investigation of other lubricants on the textured surfaces, or

composites of SU-8 textured surfaces, and further studies on other

lubricants or composites are necessary.

3. Further studies on the friction with different rotational speeds on the

textured surfaces should be revealed.

4. Use the finite element method that is a good reveal to predict fine

patterns in tribology.

Reference

259

Reference

Abgrall, P., Conedera, V., Camon, H., Gue, A. and Nguyen, N. (2007)

‘SU-8 as a structural material for labs-on-chips and microelectromechanical

systems’, Electrophoresis, 28, 4539 – 4551.

Advincula, R. C., Brittain, W. J., Caster, K. C. and Ruhe, J. (2004)

‘Polymer brushes: synthesis, characterization and applications’, Wiley-VCH

Verlag GmbH & Co.

Andrew, D., Jesper, V. C., Krystian, L. W., Erica, B. H., Jack, G., Paul, M.

H., Jonathan, D. S. and Duncan, P. H. (2014), ‘Nanosecond laser texturing

for high friction applications’, Optics and Laser in Engineering, 62, 9 – 16.

Aronoff, M. I. and McNeil, M. (1992) ‘Textured air bearing surface’, U.S.

Patent 5,079,657, January 7, 1992.

Baumgart, P., Krajnovich, D.J., Nguyen, T.A., and Tam, A.G. (1995)

‘A new laser texturing technique for high performance magnetic disk

drives’, IEEE Transactions on Magentics, 31(6), 2946-2951.

Belanger, M. C. and Marois, Y. (2001) ‘Hemocompatibility,

biocompatibility, inflammatory and in vivo studies of primary reference

materials low-density polyethylene and polydimethylsiloxane: a review’, J

Biomed Mater Res, 58, 467–77.

Bhushan, B. (1999) ‘Wear and mechanical characterization on micro-to

picoscales using AFM’, Int. Mater. Rev, 44, 105-177.

Bhushan, B. and Jung, Y. C. (2007) ‘Wetting study of patterned surfaces for

superhydrophobicity’, Ultramicroscopy, 107, 1033 – 1041.

Bowden, F. P. and Tabor, D. (1950) ‘Friction and Lubrication of Solid’,

Clarendon Press, Oxford.

Bowden, F.P. and Tabor, D. (1973) ‘Friction: An introduction to tribology’,

Anchor Books, New York.

Bowden, F.P. and Tabor, D. (1986) ‘The friction and lubrication of solids,’

Clarendon Press, Oxford.

Briscoe, B. J. (1981) ‘Wear of polymers: an essay on fundamental aspects’,

Tribology International, 14, 231-243.

Briscoe, B. J. and Sinha, S. K (2002) ‘Wear of polymers’, Proc. Of the Inst.

Of Mech. Eng, 216, 401-413.

Briscoe, B. J. and Tabor, D. (1978) ‘Friction and wear of polymers: the role

of mechanical properties’, The British Polymer Journal, 10, 74-78.

Burton, Z. and Bhushan, B. (2005) ‘Hydrophobicity, adhesion, and friction

properties of nanopatterned polymers and scale dependence for micro- and

nanoelectromechanical systems’. Nano Lett 5, 1607–1613.

Reference

260

Burwell, J. T (1957) ‘Survey of possible wear mechanisms’, Wear, 1, 119-

141.

Campo, A. del. and Greiner, C. (2007) ‘SU-8: a photoresist for high-aspect-

ratio and 3D submicron lithography’, J. Micromech. Microengineering, 17, 81

– 95.

Cassie, A. B. D. and Baxter, S. (1944) ‘Wettability of porous surfaces’,

Trans. Faraday Soc., 40, 546-551.

Curtis, J. M. and Colas, A. (2004) ‘Silicone Biomaterials: history, chemistry

& medical applications of silicones’, In Biomaterials Science, 2nd

Edition,

Elsevier: London, UK, Dow Corning (R), 80-86.

Crowninshield, R. D. and Zolman, A. R. (1989) ‘Prosthesis with enhanced

surface finish. In: USPTO. US’.

Chang, W.R., Etsion, I. and Bogy, D. B. (1987) ‘Static friction coefficient

model for metallic rough surfaces’, Journal of Tribology,110, 57-63.

Chaudhury, M. K. and Whitesides, G. M. (1991) ‘Direct measurement of

interfacial interactions between semispherical lenses and flat sheets of

poly(dimethylsiloxane) and their chemical derivatives’, Langmuir, 7, 1013 –

1025.

Derjaguin, B. V., Muller, V. M. and Toporov, Y. P. (1975) ‘Effect of

contact deformations on the adhesion particles’, J. Colloid Interface Sci, 77,

91-103.

Dumitru, G., Romano, V., Weber, H. P., Haefke, H., Gerbig, Y. and

Pfluger, E. (2000) ‘Laser microstructuring of steel surfaces for tribological

applications’, Applied Physics A, 70, 485 – 487.

Duncan, C., Weisbuch, F., Rouais, F., Lazare, S. and Baquey, Ch. (2002)

‘Laser microfabricated model surfaces for controlled cell growth’, Biosensors

& Bioelectronics, 17, 413–426.

Etsion, I. and Michael, O. (1994) ‘Enhancing sealing and dynamic

performance with partially porous mechanical face seals’, Tribology

Transactions, 37(4), 701-710.

Etsion, I. and Burstein, L. (1996) ‘A model for mechanical seals with regular

microsurface structure.’, Tribology Transactions, 39(3), 677-683.

Etsion, I., Kligerman, Y. and Halperin, G. (1999) ‘Analytical and

experimental investigation of laser-textured mechanical seal faces’, Tribology

Transactions, 42, 511 – 516.

Etsion, I. and Halperin, G., (2002) ‘A laser surface textured hydrostatic

mechanical seal’, Tribology Transactions, 45(3), 430-434.

Reference

261

Etsion, I. (2004) ‘Improving tribological performance of mechanical

components by laser surface texturing’, Tribology Letters, 17, 733 - 737.

Etsion, I. (2005) ‘State of the art in laser surface texturing’, Journal of

Tribology, 127, 248 – 253.

Faulkner, A. and Arnell, R. D. (2000) ‘The development of a finite element

model to simulate the sliding interaction between two, three-dimensional,

elastoplastic, hemispherical asperities’, Wear, 242, 114-122.

Faulkner, A., Tang, K.C., Sen, S. and Arnell, R.D. (1998) ‘Finite element

solutions comparing thenormal contact of an elastic-plastic layered medium

under loading by (a) a rigid and (b)a deformable indenter’, Journal of Strain

Analysis for Engineering Design, 33(6), 411-418.

Fitzsimmons, V. G. and Zisman, W. A. (1958) ‘Thin films of

polytetrafluoroethylene (Teflon) resin as lubricants and preservative coatings

for metals’, Ind. Eng. Chem, 50, 781-784.

Friedrich, K. (1986) ‘Friction and wear of polymer composites’, Composite

Materials Series, I, Elsevier, Amsterdam, 434-443.

Fritzson, D. (1990) ‘Friction of elastomer composites: influence of surface

temperature, sliding speed and pressure’, Wear, 139, 17 – 32.

Green, A. P. (1955) ‘Friction between unlubricated metals: a theoretical

analysis of junction model’, Proceedings of the Royal Society of London

Series A, 228, 191.

Gong, Z. Q. and Komvopoulos, K. (2003) ‘Effect of surface patterning on

contact deformation of elastic-plastic layered media’, Journal of Tribology,

125(1), 16-24.

Gong, Z. Q. and Komvopoulos, K. (2004) ‘Mechanical and

thermomechanical elastic-plastic contact analysis of Layered media with

patterned surfaces’, Transactions of the ASME. Journal of Tribology, 126, 9-

17.

Halhouli-Al, A. T., Kampen, I., Krah, T. and Buttgenbach, S. (2008)

‘Nanoindentation testing of SU-8 photoresist mechanical properties’,

Microelectronic Engineering, 85, 942-944.

Halling, J. (1976) ‘A contribution to the theory of friction’, Wear, 37, 169-

184.

Hammerstrom, L. and Jacobson, S. (2008) ‘Designed high friction surfaces

– Influence of roughness and deformation of the counter face’, Wear, 264, 807

– 814.

He B., Chen W., and Wang J. (2008) ‘Surface texture effect on friction of a

microtextured poly(dimethylsiloxane) (PDMS)’, Tribo Lett, 31, 187 – 197.

Reference

262

Hertz, H. (1882) ‘On the Contact of Elastic Solids’, J. reine und angewandte

Mathematik, 92, 156–171.

Irschick, D-J., Austin, C-C., Petren, K., Fisher, R-N., Losos, J-B. and

Ellers, O. (1996) ‘A comparative analysis of clinging ability among pad-

bearing lizards’, Biological Journal of the Linnean Society, 59, 21–35.

Ishihara, H., Yamagami, H., Sumiya, T., Okudera, M., Inada, A., Terada,

A. and Nakamura, T. (1994) ‘Contact start/stop characteristics on

photolithographic magnetic recording media’, Wear 172, 65–72.

Ito, M., Shinohara, T., Yanagi, K. and Tanaka, K. (1990) ‘Influence of

surface texture of a thin-film rigid disk on frictional properties at the head-disk

interface’, IEEE Trans. Mag., 26, 2502 – 2504.

Jackson, R. L. and Green, I. (2005) ‘A finite element study of elasto-plastic

hemispherical contact against a rigid flat’, Transactions of the ASME. Journal

of Tribology, 127, 343-354.

Jost, H. P. (1990) ‘Tribology - Origin and future’, Wear, 136, 1-17.

Jost, H. P. (1966) ‘Lubrication (tribology) education and research’, Jost

Report, (Department of Education and Science, HMSO, London) 4.

John, A. R. and Ralph, G. N. (2005) ‘Recent Progress in Soft Lithography’,

Materialstoday, Volume 8, Issue 2, 50-56.

Johnson, K. L., Kendall, K. and Roberts, A. D. (1971) ‘Surface energy and

the contact of elastic solids’, Proc. R. Soc. London, 324, 301-313.

Julthongpiput, D. Design of nanocomposite polymer coatings for MEMS

applications. PhD thesis, Iowa State University, 2003.

Kitoh, M. and Honda, Y. (1995) ‘Preparation and tribological properties of

sputtered polyimide film’, Thin Solid Film, 271, 92-95.

Komvopoulos, K., (1988) ‘Finite element analysis of a layered elastic solid in

normal contact with a rigid surface’, Journal of Tribology, 110(3), 477-485.

Komvopoulos, K., (1989) ‘Elastic-plastic finite element analysis of indented

layered media’, Transactions of the ASME. Journal of Tribology, 111(3), 430-

9.

Kovalchenko, A., Ajayi, O., Erdemir, A., Fenske, G. and Etsion, I. (2005)

‘The effect of laser surface texturing on transitions in lubrication regimes

during unidirectional sliding contact’, Tribology International, 38, 219-225.

Kragelskii, I. V. (1982) ‘Friction and Wear’, Pergamon Press, Elmsford.

Kral, E. R., Komvopoulos, K. and Sogy, D. B. (1995) ‘Finite element

analysis of repeated indentation of an elastic-plastic layered medium by a rigid

Reference

263

sphere. II. Subsurface results’, Transactions of the ASME. Journal of Applied

Mechanics, 62, 29-42.

Kral, E. R. and Komvopoulos, K. (1996) ‘Three-dimensional finite element

analysis of subsurface stresses and shakedown due to repeated sliding on a

layered medium’, Transactions of the ASME. Journal of Applied Mechanics,

63(4), 967-973.

Lahey, F. H. (1946) ‘Results from using Vitallium tubes in biliary surgery’,

read by Pearse, H.E. before the American Surgical Association, Hot Springs,

VA. Ann. Surg. 124, 1020-1026.

Lancaster, J, K. (1969) ‘Abrasive wear of polymers’, Wear, 14, 223 -239.

Lee, J. Ng., Park, C. and Whitesides, G-M. (2003) ‘Solvent compatibility of

poly(dimethylsiloxane)-based microfluidic device’, Anal. Chem, 75, 6544-

6554.

Li, J., Zhou, F. and Wang, X. (2011) ‘Modify the friction between steel ball

and PDMS disk under water lubrication by surface texturing’, Meccanica, 46,

499-507.

Liu, H., Imad-Uddin Ahmed, S. and Scherge, M. (2001)

‘Microtriobological properties of silicon and silicon coated with diamond like

carbon, octadecyltrichlorosilane and stearic acid cadmium salt films: A

comparative study’, Thin Solid Films, 381, 135-142.

Liu, W., Zhou, F., Yu, L., Chen, M., Li, B., and Zhao, G. (2002)

‘Preparation and tribological investigation of thin silicone films’, J. Mater.

Res., 17, 2357 – 2362.

Liu, H. and Bhushan, B. (2003) ‘Nanotribological characterization of

molecularly thick lubricant films for applications to MEMS/NEMS by AFM’,

Ultramicroscopy, 97, 321-340.

Liu, H. and Bhushan, B. (2003) ‘Adhesion and friction studies of

microelectromechanical systems/nanoelectromechanical systems materials

using a novel microtriboapparatus’, Journal of Vacuum Science and

Technology A, 21, 1528-1538.

Lorenz, H., Despont, M., Fahrni, N., Labianca, N., Rnaud, P. and Vettiger,

P. (1997) ‘SU-8: a low-cost negative resist for MEMS’, J. Micromech.

Microeng. 7, 121-124.

Lotters, J. C., Olthuis, W., Veltink, P-H. and Bergveld, P. (1997) ‘The

mechanical properties of the rubber elastic polymer polydimethylsiloxane for

sensor applications’, J. Micromechanics Microeng, 7,145-147.

Ludema, K. C. and Tabor, D. (1966) ‘The friction and visco-elastic

properties of polymeric solids’, Wear, 9, 329-348.

Reference

264

Lui, H. and Bhushan, B. (2003) ‘Nanotribological characterization of

molecularly thick lubricant films for applications to MEMS/NEMS by AFM’,

Ultramicroscopy, 97, 321-340.

Mack, C. A. (1985) ‘PROLITH: A Comprehensive Optical Lithography

Model’, Optical Microlithography IV, Proc., 538, 207-220.

Mcdonald, J. C. and Whitesides, G. M. (2002) ‘Poly(dimethylsiloxane) as a

material for fabricating microfluidic devices’, Acc. Chem. Res., 35, 491–499.

McFarlane, J. S. and Tabor, D. (1950) ‘Relation between friction and

adhesion’, Proceedings of the Royal Society of London A, 202, 244-253.

Minn, M. and Sinha, S. K. (2008) ‘DLC and UHMWPE as hard/soft

composite film on Si for improved tribological performance’, Surface &

Coatings Technology, 202, 3698-708.

Myshkin, N. K., Kim, C. K. and Petrokovets, M. I. (1997) ‘Introduction to

Tribology’, CMG Publishers, Seoul.

Myshkin, N. K., Petrokovets, M. I. and Kovalev, A.V. (2005) ‘Tribology of

polymers: Adhesion, friction, wear and mass-transfer’, Tribology International,

38, 910 – 921.

Nicholaos, G. D. and Polycarpou, A. A. (2008) ‘Tribological performance of

PTFE-based coatings for air-conditioning compressor’, Surface & Coatings

Technology, 203, 307-316.

Opondo, M. N. and Bessell, T. J. (1982) ‘Wear of rigid PVC in relation to

water pump designs’, Tribology International, 15, 75-83.

Overney, R. M., Takano, H., Fujihira, M., Paulus, W., and Ringsdorf, H. (1994), Anisotropy in friction and molecular stick-slip motion, Phys. Rev. Lett,

72, 3546-3549.

Pan, C-T., Yang, H., Shen, S-C., Chou, M-C. and Chou, H-P. (2002) ‘A

low-temperature wafer bonding technique using patternable material’, J.

Micromech. Microeng, 12, 611–615.

Parry, A. O., Swain, P. S. and Fox, J. A. (1996) ‘Fluid adsorption at a non-

planar wall: roughness-induced first order wetting’, J. Phys: Condens. Matter,

8, 659-666.

Piruska, A., Nikcevic, I., Lee, S. H., Ahn, C., Heineman, W. R., Limbach,

P. A. and Seliskar, C. J. (2005) ‘The autofluorescence of plastic materials

and chips measured under laser irradiation’, Lab on a Chip, 5, 1348-1354.

Reference

265

Persson B. N. J. (1998) ‘On the theory of rubber friction’, Surface Science,

401, 445 – 454.

Pettersson, U. and Jacobson, S. (2003) ‘Influence of surface texture on

boundary lubricated sliding contacts’, Tribology International, 36, 857-864.

Pham, D. C., Na, K., Yang, S., Kim, J. and Yoon, E. S. (2009)

‘Microtribological properties of topographically-modified polymeric surfaces

with different pitches’, Journal of the Korean Physical Society, 55, 1416-1424.

Pooley, C. M. and Tabor, D. (1972) ‘Friction and molecular structure: the

behaviour of some thermoplastics’, Proceedings of the Royal Society of

London: Series A, 329, 251-274.

Rabinowicz, E. (1976) ‘Wear’, Materials Science and Engineering, 25, 23-28.

Ramachandra, S. and Ovaert, T. C. (2000) ‘Effect of coating geometry on

contact stresses in twodimensional discontinuous coating’, Transactions of the

ASME. Journal of Tribology, 122, 665-71.

Rozman, M. G., Urbakh, M. and Klafter, J. (1996) ‘Stick-slip motion and

force fluctuations in a driven two-wave potential’, Phys. Rev. Lett, 77, 683.

Ryk, G., Kligerman, Y. and Etsion, I. (2002) ‘Experimental investigation of

laser surface texturing for reciprocating automotive components’, Tribology

Transactions, 45, 444-449.

Samad, M. A., Satyanarayana, N. and Sinha, S. K. (2010) ‘Tribology of

UHMWPE film on air-plasma treated tool steel and the effect of PFPE

overcoat’, Surface & Coatings Technology, 204, 1330-1338.

Sandar. M. M., Minn, M., Yaping, R., Satyanarayana, N., Sinha, S. K.

and Bhatia, C. S. (2013) ‘Friction and wear durability studies on the 3D

fingerprint and honeycomb textured SU-8 surfaces’, Tribology International,

60, 187-197.

Santner, E. and Czichos, H. (1989) ‘Tribology of polymers’, Tribology

international, 22, 103 – 109.

Saravanan, P., Satyanarayana, N., Siong, P. C., Duong, H. M and Sujeet,

K. S., (2013) ‘Tribology of self-lubricating SU-8+PFPE composite based lub-

tape’, Procedia Engineering, 68, 497 – 504.

Satyanarayana, N. and Sinha, S. K. (2005) ‘Tribology of PFPE overcoated

self-assembled monolayers deposited on Si surface’, J. Phys. D: Appl. Phys,

38, 3512-3522.

Satyanarayana, N., Sinha, S. K. and Ong, B. H. (2006) ‘Tribology of a

novel UHMWPE/PFPE dual-film coated onto Si surface’, Sensors and

Actuators A, 128, 98-108.

Reference

266

Satyanaryana, N., Gosvami, N., Sinha, S. K. and Srinivasan, M. P. (2007)

‘Friction, adhesion and wear durability studies of ultra-thin PFPE overcoated

3-Glycidoxypropyltrimethoxy silane SAM on Si surface’, Philosophical

Magazine, 87, 3209-3227.

Satyanarayana, N., Sinha, S. K. and Lim, S. C, (2009) ‘Highly wear

resistant chemisorbed polar ultra-high-molecular-weight polyethylene thin

film on Si surface for micro-system applications’, Journal of Material

Research, 24, 3331 – 3337.

Schneider, F., Draheim, J., Kamberger, R. and Wallrabe, U. (2009)

‘Process and material properties of polydimethylsiloxane (PDMS) for optical

MEMS’, Sensors and Actuators, 151, 95-99.

Shooter, K. V. and Tabor, D. (1952) ‘The frictional properties of plastics’,

Proc. Phys. Soc, 65, 661-671.

Singh, R. A., Yoon, E.S., Kim, H. J., Kong, H., Park, S., Jeong, H. E. and

Suh, K. Y. (2007) ‘Enhanced tribological properties of lotus leaf-like surfaces

fabricated by capillary force lithography’, Surface Engineering, 23, 161-164.

Singh, R. A., Yoon, E. S., Kim, H. J., Kim, J., Jeong, H. E. and Suh, K. Y. (2007) ‘Replication of surfaces of natural leaves for enhanced micro-scale

tribological property’, Material Science and Engineering C, 27, 875-879.

Singh, R. A., Siyuan, L., Satyanarayana, N., Kustandi, T. S. and Sinha, S.

K. (2011) ‘Bio-inspired polymeric patterns with enhanced wear durability for

microsystem applications’, Material Science and Engineering C, 31, 1577–83.

Singh, R. A., Satyanarayana, N. and Sinha, S. K. (2012) ‘Bio-inspired

advanced materials for reducing friction and wear in MEMS devices’,

Advanced Materials Research, 545, 359 – 363.

Siripuram, R. B. and Stephens, L. S. (2004) ‘Effect of deterministic asperity

geometry on hydrodynamic lubrication’, Journal of Tribology, 126, 527-534.

Sirong, Y., Zhongzhen, Y. and Yiu-Wing M. (2007). ‘Effects of SEBS-g-

MA on tribological behavior of Nylon 66/organoclay Nanocomposites’,

Tribology International, 40, 855-862.

Stuart, B. H. (1997) ‘Surface plasticization of poly (ether ether ketone) by

chloroform’, Polymer Testing, 16, 49-57.

Suh, M-S., Chae, Y-H., Kim, S-S., Hinoki, T. and Kohyama, A. (2010),

‘Effect of geometrical parameters in micro-grooved crosshatch pattern under

lubricated sliding friction’, Tribology International, 43, 1508 – 1517.

Ranjan, R., Lambeth, DN., Tromel, M., Goglia, P. and Li, Y. (1991) ‘Laser

texturing for low-flying-height media’, Journal of Applied Physics, 69, 5745–

5747.

Reference

267

Tabor, D. (1981) ‘Friction-The present state of our understanding’, Journal of

Lubrication Technology, 103, 169-179.

Tagawa, N. and David, B. B. (2002) ‘Air Film Dynamics for Micro-Textured

Flying Head Slide Bearings in Magnetic Hard Disk Drives’, Transaction of

ASME, 124, 568-574.

Tangena, A.G. and Wijnhoven, P. J. M. (1985) ‘Finite element calculations

on the influence of surface roughness on friction’, Wear, 103(4), 345-54.

Tam, A. C., Chuang, M. C., Cheng, D. C. and Wu, A. (1989) ‘Disk with

patterned overcoat and flat magnetic layer’, IBM Tech. Discl. Bull., 32, 5B,

264.

Tanaka, K. (1984) ‘Kinetic friction and dynamic elastic contact behavior of

polymers’, Wear, 100, 243–262.

Tanaka, K. and Miyata, T. (1977) ‘Studies on the friction and transfer of

semicrystalline polymers’, Wear, 41, 383-398.

Tay, N. B., Minn, M. and Sinha, S. K. (2011) ‘Polymer Jet Printing of SU-8

Micro-Dot Patterns on Si Surface: Optimization of Tribological Properties’,

Tribology Letters, 42, 215-222.

Tayebi, N. and Polycarpou, A.A. (2005) ‘Reducing the effects of

adhesion and friction in microelectromechanical systems (MEMSs)

through surface roughening: comparison between theory and experiments’,

Journal of Applied Physics, 98(7), 73528-73531.

Turner, M. P., Dowell, N., Turner, S. W. P., Kam, L., Isaacson, M.,

Turner, J. N., Craighead, H. G. and Shain, W. (2000) ‘Attachment of

astroglial cells to microfabricated pillar arrays of different geometries’,

Journal of Biomedical Materials Research, 51, 430–441.

Unal, H. and Mimaroglu, A. (2003) ‘Friction and wear behaviour of unfilled

engineering thermoplastics’, Mater Design, 24, 183–187.

Wakuda, M., Yamauchi, Y., Kanzaki, S. and Yasuda, Y. (2003) ‘Effect of

surface texturing on friction reduction between ceramic and steel materials

under lubricated sliding contact’, Wear, 254(3-4), 356-63.

Wang, R-H., Raman, V., Baumgart, P., Spool, A. M. and Deline, V. (1997)

‘Tribology of laser textured disks with thin overcoat’, IEEE Transactions on

Magnetics, 33, 3184-3186.

Wang, X., Kato, K., Adachi, K. and Aizawa, K. (2003) ‘Loads carrying

capacity map for the surface texture design of SiC thrust bearing sliding in

water’, Tribology International, 36(3), 189-197.

Reference

268

Wang, Q.J. and Zhu, D. (2005) ‘Virtual texturing: modeling the performance

of lubricated contacts of engineered surfaces’, Journal of Tribology, 127, 722–

728.

Wang, X., Yu, H. and Huang, W. (2010) ‘Surface texture design for

different circumstances,’ TriboBr-2010 and ITS – IFToMM 2010. In:

Proceedings of the 2nd Interna- tional Tribology Symposium of IFToMM.

Wenzel, R. N. (1936) ‘Resistance of solid surfaces to wetting by water’,

Industrial and Enginerring Chemistry, 28, 988-994.

Whitesides, G. M. (2006) ‘The origins and the future of microfluidics’,

Nature, 442, 368 – 373.

Williams, J. A. and Le, H. R. (2006) ‘Tribology and MEMS’, Journal of

Physics D: Applied Physics, 39, 201-214.

Willis, E. (1985) ‘Surface finish in relation to cylinder liners’, Wear, 109(1-4),

351-366.

Varenberg, M., Halperin, G. and Etsion, I. (2002) ‘Different aspects of the

role of wear debris in fretting wear’, Wear, 252(11-12), 902-910.

Yamada, S. (2003) ‘Layering transitions and tribology of molecularly thin films

of poly(dimethylsiloxane)’, Langmuir, 19, 7399-7405.

Ye, N. and Komvopoulos, K. (2003) ‘Effect of residual stress in surface

Layer on contact deformation of elastic-plastic Layered media’, Transactions

of the ASME. Journal of Tribology, 125(4), 692-699.

Yoon, E-S., Singh, R. A., Oh, H-J. and Kong, H. (2005) ‘The effect of

contact area on nano/micro-scale friction’, Wear, 259, 1424-1431.

Yoon, E. S., Singh, R. A., Kong, H., Kim, B., Kim, D.-H., Jeong, H. E. and

Suh, K. Y. (2006) ‘Tribological properties of bio-mimetic nano-patterned

polymeric surfaces on silicon wafer’, Tribology Letters, 21, 31 – 37.

Yu, X. Q., He, S. and Cai, R. L. (2002) ‘Frictional characteristics of

mechanical seals with a lasertextured seal face’, Journal of Materials

Processing Technology, 129, 463-466.

Yu, L., Tay, F. E. H., Xu, G., Chen, B., Avram, M. and Iliescu, C. (2006)

‘Adhesive bonding with SU-8 at wafer level for microfluidic devices’, Journal

of Physics: Conference Series, 34, 776–781.

Zhou, L., Kato, K., Umehara, N. and Miyake, Y. (2000) ‘The effects of

texture height and thickness of amorphous carbon nitride coating of a hard

disk slider on the friction and wear of the slider against a disk’, Tribology

International, 33, 665–672.

Documents

SANDAR MYO MYINT · My sincere thanks also go to Dr. Su Zhoucheng, Dr. Song Shaoning, Dr. Muhammad Ridha and Mr. Alvin from Applied Mechanics Group. Special thanks to my colleagues,