 

LIBRARY

MiChiQan State

 
  

University

PLACE IN RETURN BOX to remove this checkout from your record.

To AVOID FINES return on or before date due.

MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

|

DATE DUE

I

DATE DUE

 

'0 5 Ext/12H @005

 

 

{439‘

(.x

DEC 0 1 2M?

 

\wéaau;

 

 

'JLI 170.2]‘5
Uh“

 

 

 

 

  
 
  

 

6/01 c-JCIRC/Datooue.p65«p.15

 

Microwave Plasma Assisted Chemical Vapor Deposition of

Ultra-nanocrystalline Diamond Films

VOLUME I

By

Wen-Shin Huang

A DISSERTATION

Submitted to

Michigan State University

in partial fulﬁllment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Electrical and Computer Engineering

2004

ABSTRACT

Microwave Plasma Assisted Chemical Vapor Deposition of

Ultra-nanocrystalline Diamond Films

By

Wen-Shin Huang

Nlicrowave plasma assisted ultra—nanocrystalline diamond ﬁlm deposition was

investigated using hydrogen deﬁcient, carbon containing argon plasma chemistries with

MSU-developed microwave plasma reactors. Ultra-nanocrystalline diamond ﬁlm deposi-

tion on mechanically scratched silicon wafers was experimentally explored over the fol-

lowing input variables: (1) pressure: 60-240Torr, (2) total gas ﬂow rate: 101-642 sccm, (3)

input ruicrowave power 732-1518W, (4) substrate temperature: 500°C-7700C, (5) deposi-

tion time: 2-48 hours, and (6) N2 impurities 5-2500 ppm. H2 concentrations were less than

9%, while CH4 concentration was O.17-l.85%. It was desired to grow ﬁlms uniformly

over 3” diameter substrates and to minimize the grain size.

Large, uniform, intense, and greenish-white discharges were sustained in contact

with three inch silicon substrates over a 60-240 Torr pressure regime. At a given operating

pressure, ﬁlm uniformity was controlled by adjusting substrate holder geometry, substrate

position, input microwave power, gas chemistries, and total gas ﬂow rates. Film ultra-

nanocrystallinity and smoothness required high purity deposition conditions. Uniform

ultra-nanocrystalline ﬁlms were synthesized in low leak-rate system with crystal sizes

ranging from 3-30 nm. Films with 11—50 nm RMS roughness and respective thickness val-

ues of 1-23 pm were synthesized over 3” wafers under a wide range of different deposition

conditions. Film RMS roughness 7 nm was synthesized with thickness of 430 nm. Film

uniformities of almost 100% were achieved over three inch silicon wafers. UV-Raman and

XRD characterization results indicated the presence of diamond in the synthesized ﬁlms.

Optical Emission Spectroscopy measurements showed that the discharge gas temperature

was in excess of 2000 K.

The synthesized ﬁlms are uniformly smooth and the as grown ultra-nanocrystalline

diamond can be used for a high frequency SAW device substrate material. IR measure-

ments indicated that the ﬁlms have a wide-band, low-loss transmission window.

Cepyright by

Wen-Shin Huang

2004

To my mother and my husband

ACKNOWLEDGEMENTS

The author wishes to express her sincere appreciation to Professor

Jes Asmussen, Jr. for his encouragement and support for this thesis

research. Thanks are also due to the other members of the author’s guid-

ance committee: Professor Donnie Reinhard, Professor Timothy Grotjohn,

Professor Brage Golding, and Professor Martin A. Crimp. The author is

grateful to Dr. Dieter Gruen and the late Dr. Alan Krauss for the introduc-

tion to ultra-nanocrystalline diamond. A special note of thanks is

extended to Dr. Baokang Bi for training on AFM and FESEM, for his col-

laborative efforts in the research of SAW devices based on UNCD and for

valuable discussions. In addition, the author owes a debt of gratitude to

Dr. Rui Huang and Dr. Liang Zeng for training and sharing their knowl-

edge on XRD and TEM, respectively. The author also thanks Dr. Suilong

Jiao for assistance with TEM photos and Dr. Anirudha Sumant for assis-

tance with Raman spectra. The author thanks Brian Wright for his tech-

nical support and for assistance with SiC seal substrate holder design.

The author acknowledges use of the Keck Microfabrication Facility

and the National Science Foundation for support under MRSEC program

grant DMR-9809688 for the ﬁnancial assistance which made this thesis

research possible.

vi

TABLE OF CONTENT

LIST OF TABLES ......................................... xiii

LIST OF FIGURES ........................................ xv

1

Introduction ....................................... 1

l . 1

Motivation ........................................... 3

1 .2

Research Objectives ..................................... 4

1 .3

Dissertation Outline ..................................... 4

1

Background ....................................... 6

2. 1

General Introduction ..................................... 6

2.2

Diamond Synthesis ...................................... 8

2.2. 1

Polycrystalline Diamond Synthesis ...................... 13

2.2. l . 1

Methyl Radical and Acetylene Growth Processes ......... 13

2.2. 1 .2

Polycrystalline Diamond Competitive Growth Model ...... 16

2.2.2

UItra-nanocrystalline Diamond Synthesis ................. 18

2.2.2. 1

Carbon Dimer Growth Processes For The Synthesis of

UItra-nanocrystalline Diamond Films ...............

18

2.2.2.2

Ultra-nanocrystalline Diamond Re-nucleation Growth Model 2 1

2.3

Nanocrystalline Diamond Film Deposition Techniques and Conditions ..... 22

2.4

Nanocrystalline Diamond Applications ........................ 35

2.4. 1

IR Optical Transmission Window ....................... 35

2.4.2

Tribological Application-Seal Coatings ................... 45

vii

2.4.3

SAW Devices Based on Nanocrystalline Diamond ............. 54

3

Description of Experimental Systems and Procedures ............ 58

3. 1

Introduction ......................................... 58

3.2

Experimental systems ................................... 58

3.2. 1

Microwave Power Supply and Waveguide / Transmission System .

.

. 60

3.2.2

Vacuum Pump and Gas Flow Control System ................ 62

3.2.2. 1 MSU-MPACVD System I ....................... 62

3.2.2.2 MSU-MPACVD System 11 ...................... 64

3.2.3

Computer Control System ............................ 68

3.2.3. 1 MSU-MPACVD System I ....................... 68

3.2.3.2 MSU-MPACVD System 11 ...................... 70

3.2.4

Reactor Geometry ................................. 72

3.2.4. 1

Microwave Cavity Plasma Reactor .................. 72

3.2.4.2

Substrate Heating Stage ........................ 79

3.2.5

Microwave Cavity Plasma Reactor General Operating Field Map in

Ultra-nanocrystalline Diamond Synthesis .................. 82

3.3

Experimental Procedures ................................. 86

3.3. 1

Seeding Procedures ................................ 86

3.3. l . l

Photoresist Seeding ........................... 86

3.3. 1.2

Scratch Seeding ............................. 89

3.3. 1.3

Comparison of Photoresist Seeding and Scratch Seeding .

.

.

. 91

3.3.2

Start-up and Shut-down Procedures ...................... 97

3.3.2. 1

System I .................................. 97

3.3.2.2

System H ................................. 99

3.4

Experimental Parameter Space ............................ 10 1

3.5

Measurement Methodologies ............................. 106

3.5. l

Reactor Performance .............................. 106

3.5. l . 1

Total Growth Rate ........................... 106

3.5. l .2

Average Growth Rate ......................... 106

3.5. 1.3

Speciﬁc Yield ............................. 106

viii

3.5. 1.4

Carbon Conversion Efﬁciency ................... 106

3.5.2

Film Characteristics .............................. 107

3.5.2. 1

Film Texture .............................. 107

3.5.2.2

Film Roughness and Morphology ................. 108

3.5.2.3

Film Quality .............................. 108

3.5.2.4

Crystal Size ............................... l 10

3.5.2.5

Optical Property ............................ l 12

4

MSU-MPACVD Reactor Experimental Output Variables -

Reactor Performance (Y 1) ............................ 1 13

4. 1

Introduction ........................................ 1 13

4.2

System I .......................................... 1 19

4.2. 1

Introduction ................................... 1 19

4.2.2

Total Growth Rate=g(Ar/H2/CH4, t, ft, p, Pabs) .............. 120

4.2.3

Speciﬁc Yield=g(Ar/I-12/CH4, t, ft, p, Pabs) ................. 127

4.2.4

Carbon Conversion Efﬁciency=g(Ar/HZ/CH4, t, ft, p, Pabs) ...... 134

4.3

System 11 .......................................... 141

4.3. 1

Introduction ................................... 1 4 1

4.3.2

Total Growth Rate = g(Ar/H2/CH4, t, ft, p, Pabs,TS) ........... 141

4.3.3

Speciﬁc Yield =g(Ar/I-12/CH4, t, ft, p, Pabs) ................ 152

4.3.4

Carbon Conversion Efﬁciency =g(Ar/H2/CH4, t, ft, p, Pabs) ...... 158

4.4

Controlled Nitrogen Impurity Study By System H ................. 164

4.5

Summary .......................................... 169

5

MSU-MPACVD Reactor Experimental Output Variables -

Reactor Performance (Y1) ............................ 174

5. 1

Introduction ........................................ 174

5.2

System I .......................................... 1 77

5.2. 1

Film Texture=g(Ar/I-12/CH4, t, ft, p) ..................... 177

5.2.2

Film Roughness=g(Ar/Hz/CH4, t, f , p) ................... 183

5.2.3

Film Quality=g(t, p) .............................. 186

5.2.4

Morphology and Crystal Size=g(Ar/Hz/CH4, t, f,, p) .......... 188

5.3

System 11 .......................................... 202

5.3.1

Film Texture=g(Ar/H2/CH4, ft, p) ...................... 202

5.3.2

Film Roughness=g(Ar/Hz/CH4, t, ft, p) ................... 206

5.3.3

Film Quality=g(Ar/Hz/CH4, t, p) ...................... 2 1 1

5.3.4

Morphology=g(Ar/112/CH4, t, ft, p) ..................... 2 1 5

5.3.5

Growth Mechanism & Crystal Size=g(Ar/Hz/CH4) ........... 238

5.4

Controlled Nitrogen Impurity Study By System 11 ................. 251

5.5

Summary .......................................... 259

5.5. 1

System I ...................................... 259

5.5.2

System 11 ..................................... 260

6

Plasma Diagnostics ........ I ........................ 266

6. 1

Introduction ........................................ 266

6.2

Background Fundamental of C2 Emission Spectroscopy ............. 267

6.3

Gas Kinetic Temperature Measurement Theory .................. 270

6.4

Experimental Method and Measurement ....................... 272

6.5

Experimental Results .................................. 275

6.5. 1

C2 Emission Intensity ............................. 275

6.5.2

Gas Temperature ................................ 284

6.5.3

CN Emissions versus Controlled Nitrogen Impurities .......... 294

6.6

Summary ........................................ 295

7

Ultra-nanocrystalline Diamond Film Application ............ 297

7. 1

Introduction ...................................... 297

7.2

IR Optical Transmission Window ........................ 298

7.2. 1

Introduction ................................... 298

7.2.2

Experimental Method ............................. 300

7.2.3

Theoretical Calculation and Transmission Measurement ........ 303

7.2.3. 1

Theoretical Calculations ....................... 303

7.2.3.2

IR Transmission Measurements .................. 306

7.2.4

Experimental Results .............................. 307

7.2.4. 1

Introduction ............................... 307

7.2.4.2

Single Layer One Side With Ultra-nanocrystalline Diamond

Optical Coatings ............................ 307

7.2.4.3

Single Layer Both Side With Ultra-nanocrystalline Diamond

Optical Coatings ............................ 315

7.2.4.4 Two Layer One Side Diamond/SiOZ/Si/SiOZ

Optical Coatings ............................ 318

7.2.4.5 Two Layer Both Side Diamond/SiOZ/Si/SiOleiamond

Optical Coatings ............................ 320

7.2.5

Summary ..................................... 322

7.3

Tribological Application-Seal Coatings ..................... 323

7.3. 1

Introduction ................................... 323

7.3.2

Experimental Deposition Method ...................... 325

7.3.3

Experimental Measurements and Results .................. 327

7.3.4

Summary ..................................... 333

7.4

SAW Devices Based on Ultra-nanocrystalline Diamond .......... 334

7.4. 1

Introduction ................................... 334

7.4.2

Experimental Method ............................. 335

7.4.3

Theoretical Calculations and Measurements ................ 338

7.4.4

Experimental Results .............................. 338

7.4.5

Summary ..................................... 339

8

Summary ....................................... 340

8. 1

Introduction ........................................ 340

8.2

The Performance Comparison Between Systems I and II ............ 342

8.3

Summary of Experimental Performance ....................... 344

8.4

Plasma Diagnostics Using System 11 ......................... 348

8.5

Controlled Nitrogen Impurity Study by System 11 ................. 349

xi

8.6

Ultra—nanocrystalline Diamond Film Application ................ 350

8.7

Recommendations for Future Research ....................... 351

Appendix ....................................... 354

9. 1

System I .......................................... 354

9.2

System 11 .......................................... 357

9.3

OES ............................................. 36 1

9.4

Gas Temperature Measurements ........................... 364

9.5

X-ray Diffraction Crystallography .......................... 366

9.6

Atomic Force Microscopy ............................... 367

10

References ...................................... 369

xii

LIST OF TABLES

TABLE 2.1:

Some Properties of Natural, CVD Diamond And The Applications ........... 6

TABLE 2.2:

Nanocrystalline Diamond Film Deposition Methodology and

Film Characteristics ................................................................................... 30

TABLE 2.3:

Polycrystalline Diamond Film Deposition Methodology and

The Transmittance ...................................................................................... 37

TABLE 2.4:

Nanocrystalline Diamond Film Deposition Methodology and

The Transmittance ...................................................................................... 43

TABLE 2.5:

Diamond Film Deposition Methodology and The Friction Coefﬁcient ..... 47

TABLE 2.6:

SAW Velocities .......................................................................................... 55

TABLE 2.7:

Calculated SAW Velocities [Naka 1995] ................................................... 56

TABLE 3.1:

Parameter Space For Experiments Performed at System I ...................... 104

TABLE 3.2:

Parameter Space For Experiments Performed at System 11 ..................... 105

TABLE 4.1:

Parameter Space for Experiments Performed on System I ...................... 117

TABLE 4.2:

Parameter Space for Experiments Performed on System 11 ..................... 118

TABLE 4.3:

Reproducible Experiment Data for System I ........................................... 170

TABLE 4.4:

Reproducible Experiment Data for System II .......................................... 171

TABLE 4.5:

Quantitative Comparison of The Reactor Performance of

Systems I and H ....................................................................................... 172

xiii

TABLE 5.1:

Parameter Space For Experiments Performed on System I ..................... 175

TABLE 5.2:

Parameter Space for Experiments Performed on System 11 ..................... 176

TABLE 5.3:

Film Uniformity Study ............................................................................. 265

TABLE 6.1:

Experimental Parameters [Goyl 1998], [Goy2 1998] ............................. 268

TABLE 8.1:

Experimental Performance of Systems I & II .......................................... 343

TABLE 8.2:

Crystal Size Comparison of Ultra—nanocrystalline Diamond Films ........ 347

TABLE 9.1:

Experiment Data for System I .................................................................. 354

TABLE 9.2:

Experiment Data for System 11 ................................................................ 3 57

TABLE 9.3:

Controlled N2 Impurity Study by System 11 ............................................ 360

TABLE 9.4:

OES Data ................................................................................................. 361

TABLE 9.5:

CN Study .................................................................................................. 363

xiv

LIST OF FIGURES

FIGURE 2.1:Diamond Cubic Crystal Structure ..........................................................

....... 9

FIGURE 2.2:Graphite Crystal Structure

FIGURE 2.3:Phase Diagram by Bachmann et a1. [Bach 1991]

FIGURE 2.4:Polycrystalline Diamond Film

10

12

17

FIGURE 2.5: Diamond(110) Surface Growth Mechanism with C2

As A Growth Species ............................................................................

20

FIGURE 2.6:Ultra—nanocrystalline Diamond Film

FIGURE 3.1:MSU-MPACVD System

22

59

FIGURE 3.2:Microwave Power Supply and Waveguide/Transmission System for

MSU—MPACVD Reactor .....................................................................

61

FIGURE 3.3:Vacuum Pump and The Gas Flow Controller System for

MSU-MPACVD Reactor] ...................................................................

63

FIGURE 3.4:Vacuum Pump and the Gas Flow Controller system for

MSU-MPACVD Reactor 1] ..................................................................

FIGURE 3.5:Computer Control System for MSU-MPACVD Reactor]

FIGURE 3.6:Computer Control System for MSU-MPACVD Reactor 1]

FIGURE 3.7:The Cross Sectional View of the MCPR

FIGURE 3.8:Substrate Temperature Measurement Method

67

69

71

76

77

XV

FIGURE 3.9:Thermally Floating Substrate Holder Set Up

78

FIGURE 3.10:Substrate Heating Stage Set Up ...........................................................

...... 79

FIGURE 3.11:The Substrate Heating Stage

81

FIGURE 3.12:The MPACVD Operating Field Map Under Thermally Floating

Substrate Holder Set Up ......................................................................

..... 83

FIGURE 3.13:P-Pabs-Ts Relationship under Thermally Floating

Substrate Holder Set Up ......................................................................

..... 85

FIGURE 3.14:Photoresist Seeding

FIGURE 3.15:Scratch Seeding

88

90

FIGURE 3.16:Photoresist Seeding Film Morphology and Roughness

Inspected by AFM ................................................................................

..... 92

FIGURE 3.17:Scratch Seeding Film Morphology and Roughness

Inspected by AFM ...............................................................................

..... 95

FIGURE 3.18:Relationship Between Input, Internal, and Output Parameters In A

MPCVD-System.

FIGURE 4.1:MCPR Block Diagram for The Experiments

FIGURE 4.2:Total Growth Rate vs. Hydrogen Concentration

FIGURE 4.3:Total Growth Rate vs. Deposition Time

FIGURE 4.4:Total Growth Rate vs. Total Gas Flow Rate

FIGURE 4.5:Tota1 Growth Rate vs. Deposition Pressure

FIGURE 4.6:Tota1 Growth Rate vs. Absorbed Power

FIGURE 4.7:Speciﬁc Yield vs. Hydrogen Concentration

FIGURE 4.8:Speciﬁc Yield vs. Deposition Time

xvi

102

116

121

122

123

125

126

128

129

FIGURE 4.9:Speciﬁc Yield vs. Total Gas Flow Rate

FIGURE 4.10:Speciﬁc Yield vs. Deposition Pressure

FIGURE 4.11:Speciﬁc Yield vs. Microwave Power

FIGURE 4.12:Carbon Conversion Efﬁciency vs. Hydrogen Concentration

FIGURE 4.13:Carbon Conversion Efﬁciency vs. Deposition Time

FIGURE 4.14:Carbon Conversion Efﬁciency vs. Total Gas Flow Rate

FIGURE 4.15:Carbon Conversion Efﬁciency vs. Deposition Pressure

FIGURE 4.162Carbon Conversion Efﬁciency vs. Microwave Power

FIGURE 4.17:Total Growth Rate vs. H2 Flow Rate

FIGURE 4.18:The Inﬂuence of H2 ﬂow rate on TS and PM,S

FIGURE 4.19:Tota1 Growth Rate vs. Deposition Time

FIGURE 4.20:Total Growth Rate vs. Total Gas Flow Rate

FIGURE 4.21:The Inﬂuence of Total Gas Flow Rate on TS and PabS

FIGURE 4.22:Total Growth Rate vs. Pressure

FIGURE 4.23:Tota1 Growth Rate vs. Microwave Power

FIGURE 4.24:Speciﬁc Yield vs. H2 Flow Rate

FIGURE 4.25:Speciﬁc Yield vs. Deposition Time

FIGURE 4.26:Speciﬁc Yield vs. Total Gas Flow Rate

FIGURE 4.27:Speciﬁc Yield vs. Pressure

FIGURE 4.28:Speciﬁc Yield vs. Microwave power

FIGURE 4.29:Carbon Conversion Efﬁciency vs. H2 Flow Rate

FIGURE 4.30:Carbon Conversion Efﬁciency vs. Deposition Time

FIGURE 4.31:Carbon Conversion Efﬁciency vs. Total Gas Flow Rate

xvii

130

132

133

135

136

137

139

140

142

144

145

146

148

150

151

152

153

154

156

157

158

159

160

FIGURE 4.32:Carbon Conversion Efﬁciency vs. Pressure

FIGURE 4.33:Carbon Conversion Efﬁciency vs. Microwave Power

FIGURE 4.34:Total Growth rate vs. N2 Impurities

FIGURE 4.35:The Inﬂuence of N2 Impurities on TS and Pabs

FIGURE 4.36:Speciﬁc Yield vs. N2 Impurities

FIGURE 4.372Carbon Conversion Efﬁciency vs. N2 Impurities

FIGURE 5.1:](220)/I(111) vs. Hydrogen Concentration

FIGURE 5.2:](220)/](111) vs. Deposition Time

FIGURE 5.3:](220)/I(111) vs. Total Gas Flow Rate

FIGURE 5.4:](220)II(111) vs. Deposition Pressure

FIGURE 5.5:RMS Roughness vs. Hydrogen Concentration

FIGURE 5.6:RMS Roughness vs. Deposition Time

FIGURE 5.7:RMS Roughness vs. Deposition Pressure

FIGURE 5.8:Raman Spectra

FIGURE 5.9:Raman spectra

FIGURE 5.10:Film Morphology vs. Gas Chemistry

FIGURE 5.11:Film Morphology vs. Gas Chemistry

FIGURE 5.12:Film Morphology vs. Deposition Time

FIGURE 5.13:Film Morphology vs. Deposition Pressure

FIGURE 5.14:](220)/I(111) vs. H2 Flow Rate

FIGURE 5.15:1(220)/I(111) vs. Total Gas Flow Rate

FIGURE 5.16:1(220)/I(l 1 1) vs. Deposition Pressure

FIGURE 5.17:RMS Roughness vs. H2 Flow Rate

xviii

162

163

164

166

167

168

178

179

180

182

183

184

185

186

187

190

193

96

200

202

203

205

206

FIGURE 5.18:RMS Roughness vs. Deposition Time

FIGURE 5.191RMS Roughness vs. Total Gas Flow Rate

FIGURE 5.20:RMS Roughness vs. Deposition Pressure

FIGURE 5.21 :Raman Spectra

FIGURE 5.222Raman Spectra

FIGURE 5.23:Raman spectra

FIGURE 5.242Film Morphology vs. H2 Flow Rate

FIGURE 5.25:Film Morphology vs. Deposition Time

FIGURE 5.26:Film Morphology vs. Total Gas Flow Rate

FIGURE 5.272Film Morphology vs. Deposition Pressure

FIGURE 5.28:Film Morphology vs. Deposition Pressure

FIGURE 5.29:Bright Field TEM Image

FIGURE 5.30:Bright Field(a) & Dark Field(b) TEM Images

FIGURE 5.31:Dark Field TEM Image

FIGURE 5.32:Dark Field TEM Image by ANL [Huan 2000]

FIGURE 5.33:High Resolution TEM image by ANL [Huan 2000]

FIGURE 5.34:A Histogram of Figure 5.32 by ANL [Huan 2000]

FIGURE 5.352Dark Field TEM Image

FIGURE 5.36:Bright Field TEM Image

FIGURE 5.37:](220)/I(111) vs. Nitrogen Impurities

FIGURE 5.38:RMS Roughness vs. Nitrogen Impurities

FIGURE 5.39:Film Morphology vs. Nitrogen Impurities

FIGURE 6.1:Carbon Dimer d3I'] -a3]'1 Swan Band Emission

xix

207

208

210

212

213

214

216

221

224

228

234

239

24]

242

244

245

246

248

250

251

252

254

269

FIGURE 6.2:Experimental Set-up for OES Measurements

FIGURE 6.3:C2 Emission Intensity vs. H2 Flow Rate

FIGURE 6.4:C2 Emission Intensity vs. Total Gas Flow Rate

FIGURE 6.5:C2 Emission Intensity vs. Pressure

FIGURE 6.6:C2 Emission Intensity vs. Microwave Power

FIGURE 6.7:Growth Rate vs. C2 Emission Intensity

FIGURE 6.8:Gas Temperature vs. H2 Flow Rate

FIGURE 6.9:Growth Rate vs. Gas Temperature

FIGURE 6.10:Gas Temperature vs. Total Gas Flow Rate

FIGURE 6.11:Gas Temperature vs. Pressure

FIGURE 612sz Temperature vs. Microwave Power

FIGURE 6.13:Growth Rate vs. Gas Temperature

FIGURE 6.14:Growth Rate vs. Substrate Temperature

FIGURE 6.15:CN Emission Intensity vs. Nitrogen Impurities

274

276

277

280

281

283

284

285

286

288

289

291

293

294

FIGURE 7.1:A Sketch of Cross-Section View of UNCD Coated Si Substrates

302

FIGURE 7.2:Optical Transmission of Single Layer One Side

FIGURE 7.3:Optical Transmission of Single Layer One Side

FIGURE 7.4:Optical Transmission of Single Layer One Side

FIGURE 7.5:Optical Transmission of Single Layer One Side

FIGURE 7.6:Optical Transmission of Single Layer One Side

FIGURE 7.7:Optical Transmission of Single Layer

FIGURE 7.8:Optical Transmission of Single Layer Both Sides

FIGURE 7.9:Optical Transmission of 'IVvo Layers One Side

XX

308

310

311

313

314

316

317

319

FIGURE 7.10:Optical Transmission of Two Layers

FIGURE 7.11:Sea1 Coating Substrate Holder Setup

FIGURE 7.12:An U1tra~nanocrystalline Diamond Coated SiC Seal

321

326

328

FIGURE 7.13:Surface Proﬁle Inspected by Interferometric Imaging [Huan 2001]

329

FIGURE 7.14:Surface Proﬁle Inspected by Optical Microscopy

FIGURE 7.152SEM Images of UNCD on a Si Substrate [Bi 2002]

332

337

xxi

1

Introduction

Many attempts to control polycrystalline diamond’s microstructure have been

made but success has been limited to the control of crystal morphology through inﬂuenc-

ing the crystal’s preferred orientation (ﬁlm texture). Recently it was discovered that the

microstructure of diamond can be manipulated so the crystal size decreases from microns

to nanometers [Kond 1990]. One of the most common diamond thin ﬁlm deposition pro—

cesses is microwave plasma assisted chemical vapor deposition (MPACVD). In contrast to

other diamond synthesis processes, i.e. hot ﬁlament CVD, plasma torch, combustion syn-

thesis, etc., MPACVD has a number of advantages. MPACVD is excellent for

investigating the fundamental science of the synthesis of nanocrystalline diamond ﬁlms

because (1) it is easy to operate, (2) it is a simple technology, (3) it can be operated in a

wide range of pressure and power, (4) it produces high quality ﬁlms with reproducibility,

and (5) it is reliable and safe for long-time experiments.

The research described in this thesis is concerned with developing a microwave

plasma assisted chemical vapor deposition process and methodology that enables

ultra-nanocrystalline diamond ﬁlms to be uniformly deposited over large areas. The

MSU-invented and -developed reactors are the outcome of a long-term research effort

[Zhan 1993], [Zhan 1994], [Kuo 1997], [Khat 1997] and are utilized in this dissertation.

Building upon the experimental results of Gruen et a1. [Grue 1998], microwave plasma

assisted ultra-nanocrystalline diamond ﬁlm deposition is investigated using hydrogen

poor, carbon containing argon plasma chemistries. In order to better understand the

argon-hydrogen-methane plasma discharges, the plasma species concentrations and the

gas temperature of the deposition conditions were studied by optical emission

spectroscopy.

This exploratory investigation is concerned with depositing smooth, nanometer

sized diamond ﬁlms using a MSU-MPACVD reactor. Experimental operations was carried

out between 60-240 Torr. This research has extended the reactor operation pressure from

20-150 Torr [Zhan 1994], [Kuo 1997] to the higher pressures of 240 Torr. The work of this

thesis develops experimental methods that allow (1) the creation of large, uniform, mostly

argon discharges over a 60-240 Torr pressure regime and (2) the synthesis of uniform,

smooth ultra-nanocrystalline diamond ﬁlms on 3” diameter substrates. When ﬁrst reported

during 1999-2000, these experiments were the ﬁrst in the world to produce uniform

ultra-nanocrystalline diamond over 3” diameter substrate areas. Some of the results have

been presented at international conferences [Huan 1999], [Huan 2000] and two patent dis-

closures have been submitted to MSU. As part of this thesis [Asmu 2002], [Bi 2002], the

use of ultra-nanocrystalline diamond for a few important applications was investigated.

For example, applications such as IR optical transmission windows, tribological coatings

(SiC chemical process pump seal), and SAW (surface acoustic wave) devices were brieﬂy

investigated. The successful fabrication and operation of ultra-nanocrystalline diamond

ﬁlms as SAW device substrates, in collaboration with the Physics Department at MSU, has

been published [Bi 2002].

l.l

Motivation

Microwave plasma assisted chemical vapor deposition of diamond thin ﬁlms has

been extensively investigated by many research groups throughout the world. These

experimental investigations have shown that diamond ﬁlms can be synthesized from disso-

ciated CH4/H2 gas mixtures. As synthesized, these ﬁlms typically consist of 1-100

micron-sized grains and hence they have a surface roughness that is of the same order as

the grain size. The rough surface usually requires costly and time—consuming post process-

ing such as polishing and thus this additional step and associated cost often prevents these

ﬁlms from being used in many applications. Thus, it is desirable to develop new methods

to synthesize diamond ﬁlms with very small grains and a smooth surface.

The nucleation and growth of ultra-nanocrystalline diamond ﬁlms from a “hydro-

gen poor” argon-hydrogen-methane microwave discharges has been demonstrated by

Gruen et. a1 [Zhou] on substrate areas of less than 25 cm2. A practical importance is that

ultra-nanocrystalline diamond grown from Ar/112/CH4 gases has advantages in many

applications as compared to the more common HZ/CH4 grown polycrystalline diamond

ﬁlms. Ultra-nanocrystalline diamond ﬁlms have smaller RMS (root mean square) rough-

ness i.e., 20~40 nm vs. 1-100 micron for conventional CVD polycrystalline diamond, and

a lower ﬁlm wear coefﬁcient, and counterface wear coefﬁcient [Erde 1999], which reduces

frictional energy losses over polycrystalline diamond ﬁlms. Thus the ultra-nanocrystalline

ﬁlms may be the preferred diamond ﬁlms for many applications where smooth diamond

coatings are desired. One potential application is in wear coatings such as pump seal coat-

ings. The smaller crystal size has the beneﬁt of smaller gaps, which result in smaller seal

leakage. As a consequence, it reduces hazardous emissions.

1.2

Research Objectives

The objectives of the research covered in this dissertation were to (1) develop

microwave plasma assisted process methodologies that synthesize smooth ultra-nanocrys-

talline diamond ﬁlms uniformly over large areas using MSU-MPACVD (Michigan State

University-Microwave Plasma Assisted Chemical Vapor Deposition) reactors, (2) evaluate

the performance of the MSU-MPACVD reactor for ultra-nanocrystalline diamond synthe-

sis by investigating the reactor conditions and geometry over a wide range of experimental

parameters, (3) characterize the ultra-nanocrystalline diamond ﬁlm quality and properties

versus various deposition conditions, (4) develop a better understanding of the plasma,

species concentrations, and gas temperature during the deposition process by performing

OES (Optical Emission Spectrum) measurements, and (5) explore the use of ultra-nanoc-

rystalline diamond in a few important applications, i.e. IR optical transmission window,

tribological coatings (SiC chemical process pump seal), and SAW (surface acoustic wave)

device.

1.3

Dissertation Outline

Chapter 2 presents background and reviews the related literature. Chapter 3

describes the experimental systems and the experimental procedures; identiﬁes the impor-

tant input, internal and output variables; reports the parameter space used in this research;

and describes the measurement methodology for reactor performance and ﬁlm character-

ization. Chapters 4 through 6 summarize the experimental results of this investigation. In

particular, Chapter 4 experimentally determines the relationships of input, internal and

output variables and relates the ﬁlm deposition results with the reactor performance over

the range of the experimental parameters, and reports the inﬂuence of adding N2 impuri-

ties on ultra-nanocrystalline diamond synthesis. Chapter 5 presents the deposited ﬁlm

quality and properties versus the experimental variables. Chapter 6 presents the OES

investigation of the microwave plasma during ﬁlm deposition process and particularly

relates the C2 concentrations with the deposition conditions, and also determines gas tem-

perature versus deposition conditions. In Chapter 7, the application to an IR optical

transmission window, tribological coatings (SiC chemical process pump seal), and a SAW

device are explored. Chapter 8 summarizes the conclusions of the research in this thesis

and presents some speculations on future research/development of the ultra-nanocrystal-

line diamond ﬁlrn deposition using microwave technology.

2

Background

2.1

General Introduction

Throughout the world, researchers have long been attracted to diamond’s unique

properties. They include extreme hardness, high thermal conductivity, a wide spectral

optical transmission range, and chemical resistance. In addition, diamond has semicon-

ductor characteristics such as high breakdown electrical ﬁelds, a large band gap, and high

electron and hole mobility. Table 1.] lists some of diamond’s properties and some applica-

tions related to them [Angu 1989], [Grae 1992], [Grot 1994], [Kani 1993], [Liu 1995],

[Mori 1993], [Orr 1988], [Pan 1993], [Pier 1993], [Qadr 1993], [Shio 1990], [Spit 1981],

[Tamh 1992], and [Vavi 1992].

TABLE 2.]: SOME PROPERTIES OF NATURAL, CVD DIAMOND AND THE

APPLICATIONS

 

Property

Type Ila

CVD

Comparison

Application

  

 

Hardness (Gpa)

57-104

50- 100

2. 1 (Cu)

Abrasive Coatings

 

Thermal Conductivity

20—23

10—21

3.99(Cu)

Heat sink for electronic

31.5(SiC)

for cutting tools

(W/cle) at 298K

& microwave power devices

 

Electrical Resistivity

1016

10124016

Heat sink for electrical

(ohm-cm)

device

 

Thermal Expansion Coefﬁ-

0.8-1.2

~2.0

17(Cu)

Microelectronic circuit

cient (x10’6/K)

between 298-473K

5.5(Si02)

boards, Semiconductor sub-

mates

 

Band Gap (eV)

5.45

~5.5

1.124(Si)

Semiconductor & micro-

(GaAs)

wave power devices

 

Optical Transparency

 

 

 

UV(~230nm)

‘0

rmcrowave

(mm)

 

Visible to

mid-IR

 

 

Optical coating,

electro-optical device

 

TABLE 2.] (cont’d)

 

Property

Type Ila

Comparison

Application

 

 

Rel. Dielectric

5.7

5.6

1 1.7(Si)

RF electronic device

Constant at 45 MHz to 20

(GaAs)

GHz

 

Lattice Electron Mobility

1900-

1350-

1350(Si)

High speed electronic

(V/cmz)

2200

1500

device

 

Lattice Hole Mobility

1600

10- 1000

480(Si)

High speed electronic

(Vlemz)

device

 

Breakdown Field (V/m)

107

3x105(Si)

Semiconductor device

  
 
 

Semiconductor device

3.5x105(GaAs)

Electron Saturation

2.7x107

2.7x 107

1X107(Si)

Velocity (cm/sec)

l

 

 

 

 

Ever since Tennant discovered that diamond is a crystalline form of carbon in

1797, researchers have attempted to synthesize it [Tenn 1797]. Since the early 1980’s,

applications and deposition processes for thin polycrystalline diamond ﬁlms have been

pursued. At Michigan State University (MSU), East Lansing, Michigan, past reactor

research by Asmussen et a1. [Zhan 1994] has led to some MSU-owned and -patented

microwave plasma technologies. One outcome of this activity is the MSU-MPCVD

(Microwave Plasma Chemical Vapor Deposition) reactor, which is used for diamond syn-

thesis experiments. The experimental and theoretical knowledge developed in earlier

investigations with this reactor has been applied to diamond ﬁlm deposition. Extensive

experimental studies using HZICH4 gas mixtures to deposit diamond ﬁlms are available for

this reactor conﬁguration including uniformity of coating, optimization of growth rate,

and control of morphology [Zhan 1993], [Kuo 1997], and [Khat 1997].

As potential applications of CVD diamond are continuously developed, it is

expected that this emerging diamond technology will an have an important impact on

space, defense, science, engineering, and commercial technologies. For example, nowa-

days, tooling engineers and tool fabricators are incorporating CVD diamond products into

their product lines. As a result, an end user can buy more reliable and more readily avail-

able thick-ﬁlm diamond inserts for the consistent production of highly accurate parts.

CVD diamond offers signiﬁcantly longer tool life and increased machining productivity

because of its pureness, hardness, and. rigidness, which gives it a low coefﬁcient of fric-

tion, great abrasion resistance, high thermal conductivity, and good chemical and thermal

stability.

However, the conventional HZ/CH4 CVD polycrystalline diamond synthesis meth-

odologies exhibit columnar growth. The grain size increases rapidly with the thickness of

the ﬁlm, and usually the larger the grains the rougher the surface of the ﬁlms. The surface

roughness requires additional post processing to smooth the ﬁlms and thereby increases

the cost of the ﬁlms being used in many wear and cutting tool applications. Thus, it is

desirable to develop new methods to synthesize small crystalline and smooth diamond

ﬁlms.

2.2 Diamond Synthesis

This section brieﬂy describes the atomic and crystalline structure of diamond, as

well as the competing forms of carbon, such as graphite or amorphous phases, which

might be produced during the synthesis process.

Two basic crystal structures, cubic symmetry and hexagonal symmetry, are found

in diamond. The dominant crystal structure in both natural and synthetic diamond is cubic

structure because of the slightly lower energy (0.1-0.2 eV/carbon atom) [Pier 1993]. The

diamond cubic crystal structure (Fig.2.1) consists of two interpenetrating face-centered

cubic (FCC) lattices, displaced from each other by one quarter of a body diagonal. Each

carbon atom is tetrahedrally coordinated (using sp3 atomic orbitals), creating strong,

directed sigma bonds with its four neighboring carbon atoms. The bond length and lattice

constant are 1.54 and 3.56 angstroms, respectively;

      

--_

--

—_

"'-—

—

--

. lst lattices

displaced lattice

FIGURE 2.1 : DIAMOND CUBIC CRYSTAL STRUCTURE

Diamond symmetry covalent structure:

each carbon atom combines with four other tetrahedra

Graphite is the most common form of carbon. Each in-plane carbon atom com-

bines with its three neighbors using hybrid sp2 atomic orbits, with a covalent 6 bond

length of 1.42 angstroms. The repeating layers are van der Waals bonded, perpendicular to

the planes with a 3.35 angstrom lattice constant (Fig. 2.2).

FIGURE 2.2 : GRAPHITE CRYSTAL STRUCTURE

Graphite stacking-sheet structure:

in-layer carbon atoms form a two dimensional network of regular hexagons

Diamond is stable once formed, but the formation of diamond is less likely than

graphite or other possible phases of carbon. During the formation, diamond is thermody-

namically stable relative to graphite only at high pressure. Graphite to diamond conversion

faces a considerable kinetic barrier, although the free energy difference between them is

only 0.02 eV [Liu 1995]. In 1953, direct conversion from graphite to diamond, under high

pressure (6Gpa) and high temperature (~13000C), using a liquid solvent catalyst, was

developed by H. Liander in Sweden [Liu 1995]. In 1949, the ﬁrst successful attempts to

form CVD (chemical vapor deposition) diamond at low pressure were achieved by W.G.

Eversole [Ever 1958], who discovered that diamond could be deposited on a substrate

from a CO/C02 mixture or a hydrocarbon gas. In 1968, more research was reported by

Derjaguin et a1. [Derj 1968] and Angus et a1. [Angu 1968], but the low growth rate of this

method (less than 0.1 micrometer per hour) was discouraging and prevented it from

achieving industrial success. The breakthrough in synthesizing diamond at low pressure

 

10

came when Spitsyn et al. [Spit 1981] and Matsumoto et a1. [Mats 1982] succeeded at pro—

ducing diamond ﬁlm growth at higher rates (up to 5 micrometers per hour) at high

temperature from hydrocarbon-hydrogen gas mixtures.

 

P.K. Bachmann et. al. [Bach 1991] described the diamond growth region with the

C-H-O gas phase diagram, shown in Figure 2.3. It is apparent that diamond can be grown

from a wide range of gases mixed with hydrogen, including methane (CH4), ethane

(C2H6), ethylene (C2H4), acetylene (CZHZ), and carbon dioxide (C02). The diagram is

independent of deposition technique but not independent of temperature. Bachmann con-

cluded that diamond synthesis depends on the overall content of C, O, and H in the input

gases and the speciﬁc carbon containing input gases are not important. He also described

“above temperature of 1300 °C, only graphitic carbon can be deposited and below temper-

ature of 400 °C, amorphous carbon tends to grow”.

11

 

FIGURE 2.3 : PHASE DIAGRAM BY BACHMANN ET AL. [BACH 1991]

Identiﬁcation of the possible gas combinations in reported successful

polycrystalline-diamond-growth experiments. [Bach 1991]

12

 

2.2.1 Polycrystalline Diamond Synthesis

2.2.1.1 Methyl Radical and Acetylene Growth Processes

The current knowledge of diamond growth is described in this section. In 1995,

Huimin Liu and David S. Dandy divided the typical growth process associated with the

CVD assisted deposition of polycrystalline diamond ﬁlms into several steps [Liu 1995]: (i)

incubation period, (ii) three dimensional nucleation of an individual crystal on a substrate

surface, (iii) termination of surface nucleation and three dimensional growth of an individ-

ual crystal, (iv) coalescence of an individual crystal and the formation of a continuous

ﬁlm, and (v) the growth of a continuous ﬁlm.

Experimental observations found that diamond did not nucleate directly on a

non-diamond substrate surface. The nucleation takes places on an intermediate layer

which is formed between diamond and the non-diamond substrate during the incubation

period. This layer is formed by chemical interactions of activated gas species with the sub-

strate surface. Such an intermediate layer provides nucleation sites for diamond crystalline

growth.

13

The surface nucleation process may include the following events [Liu 1995]:

a) Atoms from the gas species impinge upon the deposition substrate and are adsorbed

onto the substrate surface.

b) The adsorbed atoms may desorb and diffuse over the substrate surface and may bond

to other surface atoms.

c) As time progresses, the concentrations of absorbed atoms increase and clusters are

formed.

(1) These clusters grow or decay according to statistical ﬂuctuation of the adsorbed atom

concentrations.

e) There exists a critical cluster size above which the cluster is more likely to survive and

grow than to decay.

Those clusters whose size exceeds the critical size during the concentration ﬂuctu-

ation are called the stable clusters, which provide suitable sites for growth either from the

direct impingement of atoms from the gas species or from the continued migration of the

adsorbed atoms.

Nuclei are formed and crystals grow independently of each other during the nucle-

ation step. Nucleation will stop when crystals have occupied all the available nucleation

sites. Three dimensional growth of individual crystals continues until crystals coalescence

and then a continuous ﬁlm forms.

The chemical mechanism of CVD diamond growth is not completely understood.

BUI it is well established that H2 and hydrocarbon gases such as CH4, CH3, CZHZ, etc.

Play an important role in the quite complex CVD growth environment. C.C. Battail, D.J.

Sl'Olovitz, and J.E. Butler proposed a possible growth theory by studying the morphologies

0f diamond ﬁlms from three dimensional atomic-scale simulations of chemical vapor dep-

l4

osition in atmosphere containing H, H2, CH3 and various partial pressures of CZHZ [Batt

1997]. The diamond growth species are believed to be (3sz and CH3. Some important

facts about the diamond growth on diamond surfaces are summarized in (a) through (g)

below:

a)

Diamond grows by the chemisorption of hydrocarbons on the substrate surface. The

chemical environment (which affects the rate at which chemisorption occurs at each

surface site) and the surface crystal structure (which determines how chemisorbed

hydrocarbon must be coordinated to form diamond) play interdependent roles in gov-

erning growth behavior.

b)

C)

Among {111},{ 110}, and {100} faces, the growth of the {111} face is the most sensi-

tive to C2H2 partial pressure and the {100} face is the least sensitive.

A single C2H2 molecule contributes two neighboring C atoms (CH3 contributes only

one) when it chemisorbs. Therefore, the presence of C2H2 effectively reduces the req-

uisite number of chemisorbed hydrocarbons for the deposition into neighboring sites

to nucleate diamond on a ﬂat surface.

d)

At low C2H2 partial pressures, {110} ﬁlms grow slowly, and {111} ﬁlms grow very

slowly, because the nucleation of a monolayer on ﬁat {110} and {111} faces requires

the bonding of two and three neighboring chemisorbed C atoms, respectively.

Once a monolayer has been initiated, for the {111} surface subsequent growth of the

monolayer requires bonding of at most two (at some sites only one) neighboring

chemisorbed C atoms, and for the {110} requires only one C atom. So, the growth of

the {111 } and {110} layers is faster than their initial nucleation.

Low C2H2 concentration yields low growth rates on both {111} and {110} faces,

because at low (3sz partial pressures, the growth of {111} and {110} facets is con-

trolled by the limited nucleation of monolayers on atomically smooth facets. Thus,

growth in Csz-poor environments occurs primarily from CH3.

g)

CH3 might be responsible for the major growth events, while CZHZ controls the

growth rate by controlling the nucleation of growth layers.

15

Gas mixtures of high hydrogen concentrations are commonly used to provide an

in-situ graphite etching. At less than or equal to one atmosphere, in which CVD systems

operate, graphite codeposition results from the higher probability of graphite formation

than diamond or other possible phases of carbon. In the CVD process, hydrogen atoms

play a crucial deposition role [May 1995]. For example:

a) H-atoms remove the ‘dangling’ carbon bonds on the growing diamond surface and

stop them from cross-linking and constructing graphite-like surfaces.

b) Atomic hydrogen etches both diamond and graphite, but unlike other forms of carbon,

diamond’s growth rate exceeds its etching rate. This is believed to be the reason that

diamond growth predominates over growth of other forms of carbon under this condi-

tion.

c) H abstraction reactions with stable hydrocarbon molecules produce highly reactive

carbon-containing radical species, such as methyl (CH3), which diffuses to the sub-

strate surface and reacts with it, thereby forming the C-C bonds necessary to propagate

the diamond lattice.

Atomic H helps stabilize the diamond phase by preferentially etching spZ-bonded

carbon and passivating radical sites on the growing surface [But] 1993], [Good 1996].

2.2.1.2

Polycrystalline Diamond Competitive Growth Model

Diamond ﬁlms produced by regular HZ/CH4 CVD methodologies results in 1-100

micro-sized polycrystals. In CVD-polycrystalline-diamond ﬁlms, the texture is usually

Perpendicular to the substrate. The Van der Drift [Van] competitive growth model has two

assumptions: ( 1) the absence of secondary nucleation, and (2) crystalline morphology is

independent of crystalline orientation. Polycrystalline diamond ﬁlms grown from ran-

domly oriented nuclei exhibit columnar growth, which is evidence of “evolutionary

Selection” by this model. Competitive growth between differently oriented grains deter-

16

mines the subsequent ﬁlm growth. With increasing ﬁlm thickness, only those crystals with

the fastest vertical growth rate will survive and determine the ﬁnal texture and surface

morphology of the ﬁlm.

The surface morphology is often very important in industrial applications. A typi-

cal SEM plan-view image of a HZ/CH4 CVD diamond ﬁlm, shown in Figure 2.4, displays

the often rough surface of polycrystalline diamond ﬁlm.

FIGURE 2.4 : POLYCRYSTALLINE DIAMOND FILM

The evolutionary selection of the crystal growth and the determination of the surface

morphology by the survived crystals with the fastest vertical growth rate.

Deposition condition: Pressure=120 Torr, H22CH4=90:10, ft=223 sccm and deposition

time :8 hours by system I.

 

2.2.2 Ultra-nanocrystalline Diamond Synthesis

2.2.2.1 Carbon Dimer Growth Processes For The Synthesis of Ultra-nanocrystalline

Diamond Films

Ultra-nanocrystalline diamond ﬁlms have crystal sizes on the order of nanometers.

These crystal sizes are usually much smaller than polycrystalline diamond ﬁlms synthe-

sized by Hz-CH4 chemistries, such as shown in Figure 2.4. Nanocrystalline diamond is

often grown in a lower hydrogen concentration environment. Noble gases, such as Ar, are

sometimes used in place of H2. When a noble gas such as Ar is added to a HZ/CH4 dis-

charge, the plasma chemistry changes. That is, ionization, dissociation, and species

concentrations are greatly modiﬁed and thereby changes the emission intensity of various

species of plasma. As H2 input is reduced and replaced by Ar, C2 emission is greatly

increased. This increase in C2 emission is interpreted as being due to the increased C2

ground state population in the plasma since excitation and quenching (competing de-exci-

tation mechanism) rates can be expected to change only marginally under these conditions

[Grue 1995]. This rise in C2 population is correlated with the observed increase in growth

rate, and thus supports that C2 is a growth species for UNCD synthesis. Thus, dicarbon

(C2) is believed to be the key growth species [Grue 1995] for ultra-nanocrystalline dia-

mond growth instead of methyl (CH3) and acetylene (CZHZ), which are believed to be the

important species in traditional CH4/1-12 polycrystalline diamond growth.

A two-step C2 addition mechanism for growth from a CH4 precursor in Ar plasmas

for a (110)-(1x1):H unreconstructed surface was proposed by Gruen et a1. [McCa 1998],

18

as shown in Figure 2.5. The ultra-nanocrystalline diamond growth process is summarized

as follows:

a) One C2 adds to the unreconstructed monohydride surface by inserting itself into ﬁrst

one C-H surface bond without abstraction of the terminating hydrogen bond (stepl) (at

the Hartree-Fock level of theory, C2 can insert into a CH bond of methane with no

activation barrier, forming a stable H3C-CH=C molecule).

b) The C2 molecule then rotates about the newly formed bond to insert its other carbon

into the OH bond across from it, thus forming a (100)-oriented surface dimer row

(step 2), producing an adsorbed ethylene-like structure (11]).

C)

A subsequent C2 molecule then inserts itself into the adjacent surface C-H bond, paral-

lel to the newly inserted surface C2 dimer (III) to produce a surface with two adjacent

ethylene-like groups (V), (steps 3 and 4).

d) The original state of the (110) surface is ﬁnally recovered by the formation of a CC

single bond between adjacent ethylene-like groups and produces a new layer on the

diamond surface via step 5 (formation of a singlet diradical, structure VI, directly from

structure V) This direct insertion growth mechanism for C2 is unique in that it is not

dependent on the abstraction of hydrogen atoms from the surface. Specially, the path

for the formation of a C-C single bond between adsorbed, two-carbon moieties via

step 5 does not involve any gas-phase atomic hydrogen.

19

 

\ H H\ /

C; .H

H\

C.

H

H

\CZC/

:c

c: —> / C

c : —> /

\

\

1 :C\

“2 :C\ f:

I

1]

1H

3 C2

FIGURE 2.5 : DIAMOND (110) SURFACE GROWTH MECHANISM

WITH C2 AS A GROWTH SPECIES

A schematic representation of the proposed mechanism for direct two-

step C2 addition to the (110)-(1x1):H diamond surface [McCa 1998].

In 1999, D.M. Gruen reported: (1) the reaction of a singlet C2 with the C=C double

b0nd of the C9H12 cluster gives carbene structures, which lead to the formation of new

diamond critical nuclei during growth. (2) On the other hand, the reaction of singlet C2

With the HC-CH single bond or CH bonds of the C9H14 cluster results

in a

Cyclobutene—like geometry, which leads to growth on the (100) surface in a series of steps.

(3) The nucleation rates increase dramatically under conditions where small fractions of

the reconstructed (100) surface are unhydrided and C2 concentrations in the plasma reach

20

levels of 1012 cm'3, and (4) low hydrogen content plasma favor these conditions [Grue

1999].

2.2.2.2

Ultra-nanocrystalline Diamond Re-nucleation Growth Model

In contrast to CVD polycrystalline diamond ﬁlms (grown by the traditional

method involving CH4/H2 mixtures), which frequently have a crystal-facet morphology

with rough surfaces (Van der Drift’s “survival of the largest”), ultra-nanocrystalline dia-

mond ﬁlms (grown from carbon containing Ar plasma) have a “featureless” morphology

with RMS surface roughness on the order of a few to 10’s of nanometers. Unlike polycrys-

talline diamond ﬁlms, ultra-nanocrystalline ﬁlms are grown by continuous re-nucleation.

An ultra-nanocrystalline growth model explains the structural evolution of the ﬁlm based

on a substrate seeded with diamond nuclei that grow isotropically. Very high heteroge-

neous renucleation rates (lOlOcm’Zsec'l) ensure that growth occurs and results in the

formation of smooth, phase-pure ultra-nanocrystalline diamond ﬁlms. This high second—

ary nucleation rate make the transition from microcrystalline to nanocrystalline diamond

ﬁlms. A typical SEM plan-view image, shown in Figure 2.6, illustrates the surface mor-

Phology of a UNCD ﬁlm.

21

FIGURE 2.6 : ULTRA-NANOCRYSTALLINE DIAMOND FILM

The renucleation growth and the determination of

the surface morphology by the small crystallinity.

Deposition condition: Pressure=120 Torr, Ar:I-]2:CH4=100:4:1, ft=105 sccm,

and deposition time=8 hours by system H.

2.3

Nanocrystalline Diamond Film Deposition Techniques

and Conditions

This section describes the past experimental work associated with nanocrystalline

diamond ﬁlm synthesis. In general, most investigations do not use hydrogen poor, carbon

containing argon discharges and most of the synthesized ﬁlms have crystal sizes greater

than 50 nm. Thus this thesis will refer to those ﬁlms as nanocrystalline ﬁlms. On the other

hand, when a high concentration of Ar discharge is combined with a 0.5-2% CH4 and zero

to a few percentage of H2, the resulting crystal sizes are usually smaller than 50 nm and

often are as small as 10 or less nm. Due to the different input gas mixtures, plasma chem-

 

22

istries, and the very small resulting crystal sizes, this thesis will refer to the deposited ﬁlms

as ultra-nanocrystalline diamond (UNCD) ﬁlms.

Ever since nanocrystalline diamond was recognized as an important structural

class in the diamond family, a number of different deposition techniques and conditions

have been employed to synthesize nanocrystalline diamond ﬁlms. Among these are hot ﬁl-

ament CVD [Lin 2000], [Wang 2004], TMHFCVD (time modulated hot ﬁlament chemical

vapor deposition) [Ali 2003], RF plasma CVD reactor [Nied 2001], biased hot ﬂame [H01]

1998], remote tubular microwave CVD [Erz 1993], DC are plasma [Kono 1995], [Nist

1997], dual ECR (electron cyclotron resonance)-rf plasma [Zarr 1997], condensing carbon

ions [Dava 1996], and microwave plasma CVD reactors [2qu 1995], [Yosh 2001], [Hong

2002], [Xu 2001], [Bhus 1998], [Grue 1998], [Chen 2000], [Shar 2001], [Shar 2003],

Yang 2002], [Li 2002]. A summary of the deposition techniques, conditions, and results of

the synthesis of nanocrystalline diamond ﬁlms is listed in Table 2.2. Some of these investi-

gations are brieﬂy described below.

Lin et al. [Lin 2000] studied the inﬂuence of argon concentration in the transition

of the morphology of diamond crystal from microcrystalline to nanocrystalline using an

argon-methane-hydrogen mixture in a conventional hot ﬁlament chemical vapor deposi-

tion system. A transitional boundary between polycrystalline and nanocrystalline occurred

when 95.5% argon concentration was used.

The nanocrystalline diamond ﬁlms were deposited on 2” Si (100) wafers by

decreasing the deposition pressure and increasing the acetone in hydrogen ratio with a hot

23

 

ﬁlament technique [Wang 2003]. Averages grain sizes of approximately 4-8 nm were

achieved. However, HRTEM (high resolution transmission electron microscopy) images

of the nanocrystalline diamond ﬁlms show grains are surrounded with amorphous

graphite.

Ali et al. fabricated the nanocrystalline diamond ﬁlms on cemented tungsten car—

bide (WC-Co) substrates using a new growth technique, identiﬁed as TMHFCVD (time

modulated hot ﬁlament chemical vapor deposition) [Ali 2003]. The modulated methane

ﬂow promoted the secondary nucleation of nanometer crystalline diamond.

With a RFPACVD (radio frequency plasma assisted CVD) reactor, an appropriate

thickness of nanocrystalline diamond layer as an anti-abrasive coating on cemented car-

bide substrates without post mechanical polishing can improve the friction coefﬁcient in

sliding against wood [Nied 2001]. The results of this study are helpful in the selection of

the optimum thickness of the nanocrystalline diamond ﬁlms to be coated on the cemented

carbide tools to improve cutting of the mills used in the wood industry.

Nanocrystalline diamond coated spherical cemented carbide substrates have been

tested in air, lubricant (water and oil) sliding against ball-bearing steel, cemented carbide,

stainless steel, titanium, and aluminum [H01] 1998]. The smooth nanocrystalline diamond

coatings show exceptionally low friction and wear when slid against ball-bearing steel,

cemented carbide, stainless steel in air and lubricant (water or oil). But when sliding

against titanium and aluminum, the nanocrystalline diamond coatings need to be polished

slightly to have the similar behavior.

24

Erz et a]. fabricated nanocrystalline diamond optical transparent ﬁlms on silicon

and quartz substrates using a methane-oxygen-hydrogen mixture by a remote tubular

microwave CVD [Erz 1993]. In this investigation, they used different diamond powder

with grain sizes ranging from 0.01—3 um to enhance diamond nucleation on the substrates.

The results showed that a nucleation density up to 3x1010 nuclei cm’2 was achieved by

scratching the substrates with 10 nm diamond powder. The high nucleation density led to a

ﬂat diamond ﬁlm with a smooth surface. However, by increasing the ﬁlm thickness from 1

pm to 10 um, the surface roughness increases more than 6 times (30 i 10 nm to 200 nm).

Konov et al. [Kono 1995] and Nistor et a1. [Nist 1996] grew ﬁne-grain diamond

ﬁlms on silicon substrates using a methane-hydrogen argon gas mixture with ﬁxed argon

ﬂow (50 sccm) and varied the methane ﬂow from 5—50 sccm and the hydrogen ﬂow from

45—0 seem in a DC. are discharge plasma. After the ultrasonic seeding process (ultrasonic

seeding with 5 nm sized diamond powder suspension in ethanol), pulsed excimer laser

irradiation generated by an excimer KrF laser (pulse duration 15 nanosecond) was used to

remove the undesirable non-uniformities in the surface distribution of the seeded crystals,

while leaving the uncoalesced particles for subsequent growth undisturbed. The two-step

seeding procedure led to highly smooth ﬁlms owing to the irradiation on the pretreated

substrates by laser assisted disintegration of the coalesced seeds and removal of large resi-

due diamond particles [Kono 1995]. The improvement of the growth of nanocrystalline

diamond ﬁlms was obtained by the combination of the uniformly seeded substrates and a

high methane concentration (50% of argon-methane—hydrogen mixture) [Nist 1996].

25

HRTEM and EELS (electron energy loss spectroscopy) characterization results of

the ﬁlms synthesized with a dual ECR-rf plasma by Zarrabian et al. showed the 4-30 nm

diamond crystallinities were embedded in, amorphous matrix, DLC (diamond like carbon)

[Zarr 1997]. Hexagonal diamond crystals that were changed to cubic diamond crystals

were observed when the bias voltage was increased.

Nanophase diamond ﬁlms were deposited on stainless steel substrates using con-

densing carbon ions carrying kinetic energy of the order of kiloelectronvolts, generated

from the laser ablation of graphite at target intensity near 1011 W cm’2 by Davanloo et al.

[Dava 1996]. Deep penetration of carbon atoms into the steel substrates was shown within

the interfacial layers examined by TEM images. The measured mechanical properties (low

friction coefﬁcient [C011 1994] and a hardness, in excess of 78 GPa [C011 1993], [Coll

1994]) conﬁrmed that the nanophase diamond ﬁlms were suited to a wide range of indus-

trial applications.

Zuiker et a1. [Zuik 1995] and Gruen at al. [Grue 1998] grew nanocrystalline and

ultra-nanocrystalline diamond ﬁlms with an argon-carbon (C60 in argon) microwave

plasma and controlled the diamond crystal microstructure by argon additions to meth-

ane-hydrogen microwave plasma discharges in a microwave plasma CVD reactor (ASTeX

PDS-l7). It was found that nanometer sized diamond could be synthesized with either a

C60 or a CH4 carbon precursor [Zuik 1995]. Cross-section and plan-view SEM images

showed that the morphology, grain size, and growth mechanism are affected by the ratio of

argon to hydrogen in the gas mixture. The transition from microcrystalline to nanocrystal-

line, which depended on the ratio of argon to hydrogen, was conﬁrmed by X-ray

26

diffraction and Raman spectroscopy. The nanocrystalline diamond was synthesized at an

Ar/H2 volume ratio of 9 and CH4 volume percentage of 1%. The ultra-nanocrystalline dia-

mond was synthesized at 0-2% of H2 and 1% of CH4 (vol%). An OES (Optical Emission

Spectroscopy) study reveal that the C2 dimer concentration is promoted signiﬁcantly by

increasing the argon concentration [Grue 1998].

A nanocrystalline diamond ﬁlm was synthesized on a 4” Si(lOO) wafer with hydro-

gen ﬂow rate of 100 sccm and methane ﬂow rate of 10 sccm using a microwave plasma

CVD system. The silicon substrate was scratched twice by dry diamond powders with the

sizes of 250 nm and 5 nm respectively [Yosh 2001]. The high nucleation density, approxi-

mately 1x10ll cm’z, led to a smooth (RMS=8.4 nm by atomic force microscopy) and

ﬁne-grain (about 10 nm observed by ﬁeld emission scanning electron microscopy) dia-

mond ﬁlm with 3.5 pm in thickness. The FI‘IR (fourier transform infrared spectrometer)

spectra showed C-H bands: sp3-CH2 symmetric stretch at 2850 cm"1 and sp3-CH2 asym-

metric stretch at 2925 cm'l, in the ﬁlm. Hong et al. used the same technique and similar

conditions to deposit nanocrystalline diamond ﬁlms also on a 4” Si(100) wafer, but modi-

ﬁed the two-step scratch seeding procedure with dry diamond powders of the sizes of 1

11m and 5 nm for tribological characteristics study [Hong 2002]. A slightly thinner ﬁlm

(2.2 um thick) with approximately same size crystals (10-15 nm) showed a very close sur-

face roughness value (10 nm).

Bhusari et a1. [Bhus 1998] and Chen et a1. [Chen 2001] deposited diamond ﬁlms

with grain sizes ranging from 4 nm to a few hundreds of nanometers in methane, hydro-

gen, and oxygen gas mixture by an AsTex 5 kW microwave reactor. The growth results of

27

the quartz substrates pretreated, with two different diamond-powder sizes, 4 nm and 0.1

pm, were compared. The ultrasmooth and highly transparent nanocrystalline diamond

ﬁlms were coated on the quartz substrates (1) scratched by 4 nm sized diamond powder

and with low ( 520%) methane concentration, and (2) scratched by 0.1 pm sized diamond

powder and with high (220%) methane concentration. According to the in situ OES

(optical emission spectroscopy) study, the C2 dimer continued to increase as methane con-

centration increased, while other hydrocarbon (CH+, C3, CH) species that decreased

signiﬁcantly as methane concentration increased. Thus, it was speculated that C2 may be

the predominant growth species at higher methane fractions [Chen 2001].

Sharda et al. compared the optical properties of microcrystalline and nanocrystal-

line diamond ﬁlms fabricated on silicon substrates by microwave plasma chemical vapor

deposition with a mixture of 5% methane in hydrogen. The substrate was pretreated with

bias enhanced nucleation. The nanocrystalline diamond ﬁlm grown at 700 °C had a very

high optical absorption coefﬁcient, i.e. >104 cm'l (higher than that of the microcrystalline

diamond ﬁlm), even though it was smoother than microcrystalline diamond ﬁlm. Never-

theless, the nanocrystalline diamond ﬁlm grown at 600 °C, was smoother, had 78%

transmittance in the infrared region, and thus had demonstrated a potential for application

as optical windows [Shar 2001]. A homogeneous refractive index within the nanocrystal-

line diamond ﬁlms and negligible changes at the interface between the ﬁlm and the

substrate were observed [Shar 2003].

Transparent nanocrystalline diamond protective coatings on quartz glasses were

grown with a MPECVD technique in methane—hydrogen-oxygen plasma discharges [Yang

28

2002]. The substrates were ultrasonically polished with 0.5 um diamond powder prior to

deposition. The nucleation density was generally higher than 1011 cm'2 as revealed by

TEM examination. By 3% of methane fraction, with coatings of 1 um in thickness, and

about 12 nm surface roughness, an optimal transmittance of 65% in the visible light region

was achieved.

Nanocrystalline diamond thin ﬁlms deposited on both sides of silicon wafers with

a ﬂow rate ratio of methane to hydrogen ﬁxed at 1:100 in a MPECVD reactor had shown

that the diamond coatings greatly enhanced the infrared transmission of silicon wafers [Li

2002]. A maximum transmittance of 85% at the wavenumber of 2200 cm‘1 (at 4.55 pm)

was achieved with 0.5 pm and 0.3 um ﬁlm thickness on each side of the silicon wafer and

approximately 27 nm surface roughness.

29

M
L
I
F
D
N
A
Y
G
O
L
O
D
O
H
T
E
M
N
O
I
T
I
S
O
P
E
D
M
L
I
F
D
N
O
M
A
I
D
E
N
I
L
L
A
T
S
Y
R
C
O
N
A
N
:
2
.
2
E
L
B
A
T

C
I
T
S
I
R
E
T
C
A
R
A
H
C

S 

.
R
.
S

)
m
n
(

l
a
t
s
y
r
C

e
z
i
S

)
m
n
(

)
m
u
(

)
C
°
(

)
W
(

)
r
r
o
T
(

)
m
c
c
s
(

)
r
e
d
w
o
p

m
l
i
F
t

S
T

r
e
w
o
P

w
o
l
F

t
n
e
m
t
a
e
r
t
e
r
P

e
t
a
R

y
r
t
s
i
m
e
h
C

s
a
G

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

 

6
.
2

A
N

s
a
i
b

0
7
8

t
n
e
m
a
l
ﬁ

  H

D
V
C
F

c
i
n
o
s
a
r
t
l
u

4
H
C
3
2
H
I
I
A

)
0
0
1
(
i
S

3
.
0
5
4

0
1
-
5

5
.
1
1

9
-
0
5
8

t
n
e
m
a
l
ﬁ

0
5
-
0
3

  H

D
V
C
F

2
H
I
3
H
C
O
C
3
H
C

m
p
5
.
0

A
N

0
5
<

-
l
i
a
v
a

t
o
n

s
i

e
l
b
a

3
.
1

:
9
1
:
0
3
4
=

g
n
i
h
c
t
a
r
c
s

]
0
0
0
2

n
i
L
[

0
5

.
p
m
e
t

C
°
0
0
3
2
-
0
0
9
1

d
e
h
c
t
a
r
c
s

0
0
1
:
)
4
-
3
(
=

e
t
s
a
p

d
n
o
m
a
i
d

)
0
0
1
(
i
S

]
3
0
0
2

g
n
a
W
[

0
3

0
0
1

A
N

t
n
e
m
a
l
ﬁ

  T

D
V
C
F
H
M

o
C
-
C
W

c
i
n
o
s
a
r
t
l
u

A
N

2
H
Z
4
H
C

5
7
3
.
0

A
N

0
0
9

5
1
2
1

t
u
p
n
i

r
e
w
o
p

  R

D
V
C
A
P
F

0
0
1
-
0
5

4
H
C

A
N

C
C

1
>

W
0
0
3
r
e
w
o
p

)
8
9
-
6
9
(
:
)
4
-
2
(
=

g
n
i
h
s
i
l
o
p

]
3
0
0
2

i
l
A
[

30

7
5
o 1
t

5 

0
0
0
3

]
1
0
0
2

d
e
i
N
[

0
3

A
N

A
N

0
5
8

r
e
w
o
p

e
m
a
ﬂ

0
6
7

0
6
9

1

2
H
2
C
3
2
0

m
p

1

C
C

t
o
H

d
e
s
a
i
B

0
3

t
i

e
l
b
a
l
i
a
v
a

t
o
n

s
i

0
3
0
1
:
0
3
9
=

d
e
h
s
i
l
o
p

]
8
9
9
1

]
1
0
H
[

e
m
a
l
F

0
2

0
0
1

0
0
1

A
N

0
5
5

0
0
8

 

 

 

5
.
7
3

0
3
<

8
-
0
0
6

0
5
-
0
3

 

 

  
  R
  

n
o
i
s
n
e
p
s
u
s

s
e
g
r
a
h
c
s
i
d

:
)
0
-
5
4
(
:
0
5
=

4
H
C
:
2
H
:
r
A

m
u
3
-
1
0
.
0

D
V
C
M
T

d
e
h
c
t
a
r
c
s

l
o
h
o
c
l
a

d
e
t
u
l
i
d

2
0
3
2
H
I
4
H
C

1
:
7
9
:
2
=

z
t
r
a
u
q

c
r
a
C
D

o
n
o
K
[

 

)
0
5
-
5
(

a
i
d
e
m

-
e
u
q
a

]
3
9
9
1

]
5
9
9
1

z
r
E
[

 

 

 

A
N

m
n

5

i
S

n
i

s
u
o

 

n
i

-

5
!

I
I
I
V
I
-
H
A

l
‘

\
n

 

 

e
u
n
i
t
n
o
C

2
.
2

e
l
b
a
T

.
R
S

)
m
n
(

3
1
>

0
3

<

l
a
t
s
y
r
C

e
z
i
S

m
n
(

) 

)
C
°
(

r
e
w
o
P

)
W
(

)
r
r
o
T
(

)
m
c
c
s
(

)
r
e
d
w
o
p

w
o
l
F

t
n
e
m
t
a
e
r
t
e
r
P

e
t
a
R

y
r
t
s
i
m
e
h
C

s
a
G

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

 

—
—
=
—
r
=
=

=
=
 

0
5
0
1

0
0
7

A
N

0
0
1

0
0
1

4
H
C
:
z
s
r
A

m
n

5

i
S

e
r
a
C
D

:
)
0
-
5
4
(
:
0
5
=

n
i

n
o
i
s
n
e
p
s
u
s

)
0
5
-
5
(

l
o
n
a
h
t
e

]
6
9
9
1

t
s
i
N
[

s
e
g
r
a
h
c
s
i
d

3
-
1

A
N

A
N

A
N

0
5
—
0
1

  C

d
r
o
c
r
o
b
r
o
a
(

g
n
i
s
n
e
d
n
o
c

g
n
i
s
n
e
d
n
o

s
s
e
l
n
i
a
t
S

-
r
a
c

A
N

s
n
o
i

n
o
b

)
e
r
o

l
e
e
t
S

s
n
o
I

n
o
b
r
a
C

f
d
n
a

]
6
9
9
1

a
v
a
D
[

0
1

4
3

5
.
3

4
.
8

A
N

7
-
0
7
6

  M

&
m
n
0
5
2

D
V
A
P

2
H
I
4
H
C

0
1
1

”
4

0
3

7
.
1

0
0
1

0
5
8

0
0
5
1

0
0
3
-
0
1

  M

D
V
A
P

Z
H
I
I
A

2
0
1

m
p

1
.
0

i
S

 

0
1

5
1
-
0
1

2
.
2

6
-
0
4
6

0
0
2
2
-
0
0
8
1

4
2

A
N

2
H
Z
4
H
C

&
m
p

1

i
S

D
V
C
P
W
M

31

0
8

0
0
1
:
0
1
=

d
e
h
c
t
a
r
c
s
m
n

5

]
2
0
0
2

g
n
o
H
[

0
5

0
0
1
:
0
1
=

m
n

5

)
0
0
1
(
i
S

]
1
0
0
2

h
s
o
Y
[

d
e
h
c
t
a
r
c
s

9
.
4
1
=

.
e
v
a

C
°

0
0
6

:
0
6
C

2
:
0
0
1
=

d
e
h
s
i
l
o
p

]
5
9
9
1

k
i
u
Z
[

.
e
v
a

 

 

 

5
0
1

1
4

0
0
1

0
0
1

0
5
8

0
5
8

.
e
v
a

 

 

 

0
0
5
1

0
0
5
1

9
.
4
1
=

9
.
4
1
=

0
0
3
-
0
1

0
0
3
-
0
1

 

  
  M
  M

d
e
h
s
i
l
o
p

d
e
h
s
i
l
o
p

D
V
A
P

D
V
A
P

0
1
:
0
9
=

:
0
6
C

k
i
u
Z
[

]
5
9
9
1

k
i
u
Z
[

]
5
9
9
1

2
:
8
9
=

0
3
6

2
H
I
I
A

Z
H
I
I
A

c
°

0
0
1

m
p

C
0

0
3
6

:
0
6
C

m
p

1
.
0

1
.
0

i
S

0
0
1

i
S

 

 

 

 

I
n
t
-
n
u
t
.
.
.

I

.
0
.

Q

 

.
R
.
S

)
m
n
(

l
a
t
s
y
r
C

e
z
i
S

)
m
n
(

m
l
i
F
t

)
m
1
1
(

)
C
°
(

r
e
w
o
P

)
W
(

w
o
l
F

t
n
e
m
t
a
e
r
t
e
r
P

)
r
r
o
T
(

)
m
c
c
s
(

)
r
e
d
w
o
p

e
t
a
R

y
r
t
s
i
m
e
h
C

s
a
G

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

 

e
u
n
i
t
n
o
C

2
.
2

e
l
b
a
T

5
2
1

A
N

0
0
1

0
5
8

0
0
5
1

  M

4
H
C
:
2
H
:
r
A

D
V
A
P

8
.
0
0
1

n
r
u

]
.
0

i
S

8
.
0
:
0
1
:
0
9
=

C
°
0
2

:
0
6
C

d
e
h
s
i
l
o
p

]
5
9
9
1

k
i
u
Z
[

3
5

3
.
2

A
N

0
0
1

0
5
8

0
0
5
1

  M

4
H
C
:
2
H
:
r
A

D
V
A
P

3
.
0
0
1

m
p

1
.
0

i
S

8
0
5
1
:
8
9
=

C
°
0
2

:
0
6
C

d
e
h
s
i
l
o
p

]
5
9
9
1

k
i
u
Z
[

0
1

5
4

0
0
1

A
N

0
5
8

0
0
8

  M

4
H
C
i
n
i
f
A

D
V
A
P

m
u

3
0
1

1
.
0

i
S

0
0
1

0
0
8

0
0
2
1

4
8
.
8
1

0
3
-
0
1

  M

4
H
C
:
2
H
z
r
A

D
V
A
P

0
0
1

m
p

1
.
0

i
S

0
0
1

0
0
8

0
0
2
1

9
1
.
4
5

0
5
-
0
3

  M

4
H
C
:
2
H
:
r
A

)
0
0
1
(
i
S

D
V
A
P

0
0
1

m
p

1
.
0

1
:
9
:
0
9
=

d
e
h
s
i
l
o
p

]
8
9
9
1

e
u
r
G
[

C
°

0
2

:
0
6
C

1
:
2
:
0
0
1
=

d
e
h
s
i
l
o
p

]
5
9
9
1

k
i
u
Z
[

32

1
:
2
:
7
9
=

d
e
h
s
i
l
o
p

]
8
9
9
1

e
u
r
G
[

l

A
N

0
0
1

0
0
2

0
0
8

0
2
-
3

  M

4
H
C
3
2
H
Z
I
A

D
V
A
P

0
0
1

m
u

1
.
0

i
S

8
2

2
2

0
6

a
-
4

8
.
0

0
0
1

w
e
f

0
5
7

0
0
2

0
0
0
1

0
0
0
2

6
-
0
9
5

0
0
4
-
0
7

6
.
0
-
5
.
0

 

 

 

 

 

 

  
  
  M

l
.
0
:
)
9
.
9
7
-
9
.
5
9
(

 

 

d
e
h
c
t
a
r
c
s

g
n
i
h
s
i
l
o
p

 

 

c
i
n
o
s
a
r
t
l
u

d
n
o
m
a
i
d

D
V
C
E
P

2
0
1
2
H
I
4
H
C

D
V
C
P
M

:
)
0
2
-
4
(
=

m
p
5
.
0

z
t
r
a
u
q

n
i
z
s
C

n
e
h
C
[

s
u
h
B
[

]
0
0
0
2

]
]
0
0
2

e
t
s
a
p

]
8
9
9
1

m
n
4

4
8
:
6

0
0
2

A
N

u
X
[

i
S

1

=

l
:
0
:
9
9
=

d
e
h
s
i
l
o
p

]
8
9
9
1

e
u
r
G
[

 

.
R
.
S

)
n
m
(

l
a
t
s
y
r
C

e
z
i
S

)
m
n
(

m
l
i
F
t

r
e
w
o
P

)
m
u
(

)
C
°
(

)
W
(

)
r
r
o
T
(

w
o
l
F

e
t
a
R

)
m
c
c
s
(

)
r
e
d
w
o
p

t
n
e
m
t
a
e
r
t
e
r
P

y
r
t
s
i
m
e
h
C

s
a
G

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

 

0
0
2

-
6

w
e
f

a
-
4

6
.
0
-
5
.
0

6
-
0
9
5

0
0
0
1

2
2

0
0
2

2
0
2
2
H
Z
4
H
C

m
p

1
.
0

z
t
r
a
u
q

D
V
C
P
M

0
0

1

l
.
0
:
)
9
.
7
5
-
9
.
9
7
(

:
)
2
4
-
0
2
(
=

c
i
n
o
s
a
r
t
l
u

g
n
i
h
s
i
l
o
p

]
8
9
9
1

s
u
h
B
[

]
0
0
0
2

n
e
h
C
[

e
u
n
i
t
n
o
C

2
.
2

e
l
b
a
T

a

7
1

0
3

0
0
0
1

0
0
6

-
w
e
f

  M

)
0
0
1
(
i
S

D
V
C
P

A
N

d
e
s
a
i
B

2
H
I
4
H
C

0
1

w
e
f

a

0
0
1
:
5
=

.
d
e
c
n
a
h
n
E

V
0
6
2
-

]
1
0
0
2

r
a
h
S
[

]
3
0
0
2

r
a
h
S
[

0
3

a

4
3

0
0
7

-
w
e
f

0
0
0
1

  M

)
0
0
1
(
i
S

D
V
C
P

d
e
s
a
i
B

A
N

2
H
I
4
H
C

w
e
f

a

0
0
1

0
0
1
:
5
=

.
d
e
c
n
a
h
n
E

]
1
0
0
2

r
a
h
S
[

V
0
6
2
-

]
3
0
0
2

r
a
h
S
[

0
3

5
.
0

A
N

A
N

0
5
6

0
0
5
2

  M

D
V
C
E
P

c
i
n
o
s
a
r
t
l
u

2
H
Z
4
H
C

i
S
-
n

A
N

0
0
1

:

1
=

g
n
i
h
s
i
l
o
p

]
2
0
0
2

i
L
[

2
.
0
:
8
.
6
9
:
3
=

c
i
n
o
s
a
r
t
l
u

g
n
i
h
s
i
l
o
p

]
2
0
0
2

g
n
a
Y
[

0
3

0
0
5
1

0
0
5

2
1

5
6

0
.
l
~

  M

D
V
C
W

z
t
r
a
u
q

0
0
2
~

m
p

5
.
0

I
‘
2
0
3
2
H
I
4
H
C

33

7
2

0
3

0
3

0
3

3
.
0

A
N

A
N

3
.
1

9
.
0

&

3
.
1

A
N

A
N

A
N

0
5
6

0
5
6

0
5
6

 

 

 

0
0
5
2

0
0
5
2

0
0
5
2

&
5
.
0

 

 

 

  
  M
  M
  M

c
i
n
o
s
a
r
t
l
u

c
i
n
o
s
a
r
t
l
u

g
n
i
h
s
i
l
o
p

g
n
i
h
s
i
l
o
p

g
n
i
h
s
i
l
o
p

D
V
C
E
P

D
V
C
E
P

D
V
C
E
P

c
i
n
o
s
a
r
t
l
u

 

0
0
1
:
1
=

]
2
0
0
2

]
2
0
0
2

]
2
0
0
2

2
H
I
4
H
C

2
H
I
4
H
C

2
H
I
4
H
C

 

 

 

 

i
S
-
n

A
N

i
S
-
n

i
S
-
n

A
N

A
N

0
0
1

:
1
=

0
0
1

:
1
=

i
L
[

i
L
[

i
L
[

r
e
h
t
e
l
y
t
u
b
o
n
o
m
l
o
c
y
l
g
l
y
h
t
e

%
0
9
d
n
a

H
O
l
C
H

%
0
1

f
o

n
o
i
t
u
l
o
s

h
t
i
w

g
n
i
h
s
i
l
o
p

y
l
l
a
c
i
t
y
l
o
r
t
c
e
l
e

:
a

e
r
u
t
c
u
r
t
s

s
u
o
h
p
r
o
m
a

y
b

d
e
d
n
u
o
r
r
u
s

e
r
e
w

s
l
a
t
s
y
r
c

d
n
o
m
a
i
d

e
n
i
l
l
a
t
s
y
r
c
o
n
a
n

:
D
C
N
*

s
t
n
e
m
i
r
e
p
x
e
f
o

s
r
u
o
h

2

r
e
t
f
a

d
e
d
d
a

s
a
w

2
0

:
*
Z
O
:
2
H
:
4
H
C

e
d
i
b
r
a
c

n
e
t
s
g
n
u
t

d
e
t
n
e
m
e
c

:
o
C
-
C
W

e
d
i
b
r
a
c

d
e
t
n
e
m
e
c

:
C
C

r
e
d
w
o
p

e
d
i
x
o

m
u
n
i
m
u
l
a

d
e
n
i
c
l
a
c
m
m
9

h
t
i
w

g
n
i
h
s
i
l
o
p

d
n
a

g
n
i
p
p
a
l

:
b

r
e
d
w
o
p

e
d
i
x
o

m
u
n
i
m
u
l
a

d
e
n
i
c
l
a
c
m
p

3

h
t
i
w

g
n
i
h
s
i
l
o
p

d
n
a

g
n
i
p
p
a
l

:
0

r
e
d
w
o
p

e
d
i
x
o

m
u
n
i
m
u
l
a

d
e
n
i
c
l
a
c
m
p

1

h
t
i
w

g
n
i
h
s
i
l
o
p

d
n
a

g
n
i
p
p
a
l

:
1
(

s
e
l
c
i
t
r
a
p

C
i
S

e
z
i
s

t
i
r
g

0
0
0
1

h
t
i
w

g
n
i
h
s
i
l
o
p

r
e
l
b
m
u
t

:
e

e
r
u
t
a
r
e
p
m
e
t

e
t
a
r
t
s
b
u
s

:
5
T

s
s
e
n
h
g
u
o
r

e
c
a
f
r
u
s

:
.
R
.
S

s
s
e
n
k
c
i
h
t

m
l
ﬁ

:
m
l
i
F
t

r
A

y
b

d
e
w
o
l
l
o
f

,
n
i
m

0
6

r
o
f

n
o
i
t
u
l
o
s

e
n
a
h
t
e
m

d
n
a

e
n
a
h
t
e
o
r
o
u
ﬂ
i
r
t
o
r
o
l
h
c
i
r
t

1
:
1

e
r
u
p
-
a
r
t
l
u

n
i

g
n
i
t
a
c
i
n
o
s

,
g
n
i
n
a
e
l
c

l
a
r
e
n
e
g

:
f

n
i
m

0
3

r
o
f

g
n
i
n
a
e
l
c

a
m
s
a
l
p

34

 

2.4

Nanocrystalline Diamond Applications

2.4.1 IR Optical Transmission Window

Proper coating protection has become a necessity because many optical compo-

nents are operated in a hostile environment, such as infrared windows for aerodynamic

applications. By combining the good optical transparency of diamond and its exceptional

physical, mechanical, and chemical properties, like high strength and chemical inertness,

diamond ﬁlms make excellent protective coatings for optical elements. Thus, during the

past 15 years, researchers have given much effort to investigating the deposition of dia-

mond ﬁlms for optical coatings. Table 2.3 summarizes the past experiments concerned

with the deposition of polycrystalline diamond for optical application. The different depo-

sition methodologies that were employed for polycrystalline diamond deposition are

summarized along with the transmission performance of the resulting structures. The

results of a number of these investigations are brieﬂy described below.

A low temperature (~400 °C) plasma enhanced chemical vapor deposition process

(by varying the time duration of the plasma pulse and the number of the pulses to control

the substrate temperature) was developed to coat diamond on fused quartz slides. It used a

methane, hydrogen, and oxygen gas mixture as deposition gases [Ong 1989]. The ﬁlms’

transmission is over 60% in the range of 0.6-2 um wavelength. In 1990, Wang et a1.

noticed that by etching away the substrate and polishing the rough surfaces of polycrystal-

line diamond, the transmittance of the free-standing diamond ﬁlm was increased from the

range of 14-45% to the range of 64-75% (4000 cm'1 to 600 cm‘l) [Wang 1990]. Sun et al.

[Sun 1992] produced another free-standing diamond ﬁlm by etching away the silicon sub-

35

strate. Without polishing the diamond surface, the mean transmission was 71% around a

wavelength of 20 um and 57% around a wavelength of 2.5 pm. Another investigation

which makes use of the traditional CH4/H2 polycrystalline diamond growth achieved

as-deposited, non-polished CVD polycrystalline thin diamond ﬁlms with near idea optical

transmission throughout the visible and near- to mid— infrared [Ulcz 1998] and [Rein

2000]. Tang et a]. reported that the transmittance of the polished diamond ﬁlms (by

another thick polycrystalline diamond ﬁlm) could be increased more than one order in

magnitude when the surface roughness were decreased from 3.2 pm to 0.55 pm and the

other from 5.2 pm to 1.35 um [Tang 2003].

36

E
H
T
D
N
A
Y
G
O
L
O
D
O
H
T
E
M
N
O
I
T
I
S
O
P
E
D
M
L
I
F
D
N
O
M
A
I
D
E
N
I
L
L
A
T
S
Y
R
C
Y
L
O
P

:
3
.
2
E
L
B
A
T

C
N
A
T
T
I
M
S
N
A
R
T

E 

l
a
t
s
y
r
C

e
z
i
S

.
R
.
S

e
c
n
a
t
t
i
m
s
n
a
r
T

m
a
r
t

S
T

r
e
w
o
P

w
o
l
F

e
t
a
R

s
a
G

y
r
t
s
i
m
e
h
C

t
n
e
m
t
a
e
r
t
e
r
P

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

)
m
n
(

)
m
-
l
l
(

)
C
°
(

)
W
(

)
r
r
o
T
(

m
n
(

) 

)
m
c
c
s
(

)
r
e
d
w
o
p

m
u
0
2

t
a

%
1
7
.

A
N

=
.
e
v
a

0
0
9

1
.
4

t
n
e
m
a
l
i
F

  H

D
V
C
F

e
t
s
a
p
m
p

)
0
0
1
(
i
S

5
.
0

=
2
H
Z
4
H
C

0
0
1

5
.
7

0
0
0
2

m
t
.
]
5
.
2

t
a

%
7
5
1
2

C
0

0
0
1
2

5
2
9
9
2
5
7
.
0

d
e
h
c
t
a
r
c
s

]
2
9
9
1

n
u
S
[

0
2

0
0
4

2
9
.
0

%
0
6
r
e
v
o

  M

D
V
C
P
W

=
2
0
2
2
H
I
4
H
C

l
-
5
.
0

d
e
s
u
f

m
u

0
0
1

0
3

0
0
2

0
0
4

5
2
.
1

0
0
4
~

%
0
6
r
e
v
o

  M

D
V
C
P
W

=
2
0
1
2
H
I
4
H
C

d
e
s
u
f

1
-
5
.
0

n
r
u

0
3

0
0
1

)
m
u

8
3
.
1
-
2
.
1
(

)
m
p

9
.
1
-
6
.
l
(
&

2
.
0
:
5
.
9
9
:
3
.
0

d
e
h
c
t
a
r
c
s

z
t
r
a
u
q

]
9
8
9
1

g
n
O
[

)
m
u

2
-
6
.
0
(

2
.
0
:
5
.
9
9
:
3
.
0

d
e
h
c
t
a
r
c
s

z
t
r
a
u
q

]
9
8
9
1

g
n
O
[

%
5
4
.
4
1
3

1
2

A
N

0
9
7

0
0
0
1

n
o
r
c
i
m

  M

D
V
C
P

d
e
h
c
t
a
r
c
s

=
2
H
Z
4
H
C

0
0
1

0
9

i
S

%
5
7
-
4
6
b

0
2

A
N

A
N

0
9
7

0
0
0
1

   M

D
V
C
P

d
e
h
c
t
a
r
c
s

=
2
H
I
4
H
C

0
0
1

0
9

i
S

  

)
m
u

7
.
6
1
-
5
.
2
(

5
.
9
9
:
5
0

]
0
9
9
1

g
n
a
W
[

A
N

A
N

)
m
u

7
.
6
1
-
5
.
2
(

%
9
6
-
1
6
”

6
.
5
1

0
8
9

0
7
6

0
9

0
0
1

=
2
H
I
4
H
C

d
e
h
c
t
a
r
c
s

i
S

D
V
C
P
M

6
9
:
4

]
0
9
9
1

g
n
a
W
[

)
m
u

7
.
6
1
-
5
.
2
(

5
.
9
9
:
5
0

]
0
9
9
1

g
n
a
W
[

37

 

%

C
°

0
5

0
5

0

0
1

t
a

t
a

A
N

0
5
1

A
N

0
7
1

0
0
9

0
5
5

0
0
1
2

0
0
0
1

0
2
3
3

0
0
0
3

=
.
e
v
a

)
m
p

=
.
e
v
a

9
-
0
0
8

9
-
0
0
8

%
9
5
“

%
4
5
3

 

 

 

 

m
u
0
2

m
u
5
.
2

 

t
n
e
m
a
l
i
F

 

0
1
-
5
.
2
(

0
0
4
-
0
0
2
3

0
0
4
-
0
0
2
3

3
4
.
1
2
-
9
7
.
9
1
c

%
3
6
.
0
-
5
2
.
0

  H
  M
0  
 

e
t
s
a
p
m
p
5
.
0

D
V
C
A
P

5
7
.
8
9
:
5
2
.
l

D
V
C
A
P
M

d
e
h
c
t
a
r
c
s

]
3
0
0
2

 

g
n
a
T
[

 

)
0
0
1
(
i
S

 

 

D
V
C
F

 

=
2
H
I
4
H
C

=
2
H
Z
4
H
C

=
2
H
I
4
H
C

]
2
9
9
1

1
-
0
0
1

1
-
0
0
1

0
0
1
:
4

0
0
1
:
4

]
3
0
0
2

g
n
a
T
[

n
u
S
[

A
N

A
N

0
0
1

0

0
1

A
N

A
N

i
S

5
.
7

i
S

\

l
i
l
o
-
n
o
.
0
.

l
t
§
i
§
n
u

\

  

 

e
u
n
i
t
n
o
C

3
.
2

e
l
b
a
T

l
a
t
s
y
r
C

e
z
i
S

.
R
.
S

e
c
n
a
t
t
i
m
s
n
a
r
T

m
l
i
F
t

r
e
w
o
P

)
m
n
(

)
n
u
t
(

)
C
°
(

)
W
(

)
r
r
o
T
(

w
o
l
F

e
t
a
R

s
a
G

y
r
t
s
i
m
e
h
C

t
n
e
m
t
a
e
r
t
e
r
P

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

m
n
(

) 

)
r
e
d
w
o
p

)
m
c
c
s
(

 

=
.
e
v
a

0
0
2
5

%
1
.
0
-
4
0
.
0

5
3
3

9
-
0
0
8

0
0
4
-
0
0
2
3

1
-
0
0
1

A
N

=
2
H
I
4
H
C

A
N

i
S

D
V
C
A
P
M

%
7
1
-
8
c

0
5

A
N

0
1
3

0
5
3
1

0
0
0
3

9
-
0
0
8

0
0
4
-
0
0
2
3

0  M

D
V
C
A
P

=
2
H
Z
4
H
C

1
-
0
0
1

0
0
1
:
4

A
N

0
1

i
S

A
N

]
3
0
0
2

g
n
a
T
[

.
6
V
8

4
.
0

A
N

5
-
0
5
4

2
-
0
1

%
6
6
-
3
6

  M

D
V
C
A
P

d
e
h
c
t
a
r
c
s

2
H
Z
4
H
C
*

A
N

A
N

i
S

0
5

0
1

0
0
1
:
4

]
3
0
0
2

g
n
a
T
[

0
2
$

.
e
v
a

5
-
0
5
4

4
.
0

A
N

2
-
0
1

%
9
8
-
0
8

0  M

2
H
2
4
I
—
I
C
*

D
V
C
A
P

d
e
h
c
t
a
r
c
s

A
N

A
N

i
S

)
m
p

5
-
3
(

]
0
0
0
2

n
i
e
R
[

.

0
2
S

6
V
3

A
N

5
1
.
0

%
8
9
-
5
8

5
-
0
5
4

2
-
0
1

0  M

i
S
I
Z
O
i
S

D
V
C
A
P

d
e
h
c
t
a
r
c
s

2
H
Z
4
H
C
*

A
N

A
N

1

 

 

 

 

 

 

5
.
1

0
0
3

0
0
6

5
-
3
(

)
m
p

0
0
0
1

5
-
3
(

)
m
u

=
.
e
v
a

>
-
0
0
7

)
m
u

0
0
2
$

>
-
0
0
0
1

 

2
-
2
.
0
(

%
0
8
0
“

%
0
9
=
.
e
v
a
&

  M
  

.
d
e
c
n
a
h
n
E

)
O
O
l
(
i
S

D
V
C
P

V
0
0
2
-

]
3
0
0
0
2

=
2
H
I
4
H
C

>
-
)
0
0
1

 

d
e
s
a
i
B

]
1
0
0
2

]
3
0
0
2

n
i
e
R
[

r
a
h
S
[

r
a
h
S
[

>
-
0
3

)
0
0
1

A
N

0
5

:
2
(

 

 

 

 

:

l
(

)
m
p

5
-
3
(

]
0
0
0
2

n
i
e
R
[

.
6
V
3

A
N

2
-
0
1

5
1
.
0

5
-
0
5
4

%
4
7
-
7
6

  M

r
s
y
o
r
s

D
V
C
A
P

d
e
h
c
t
a
r
c
s

2
H
I
4
H
C
*

A
N

A
N

0
0
2
$

%
6
8
=
.
e
v
a
&

)
m
p

5
-
3
(

)
m
p

5
-
3
(

]
0
0
0
2

n
i
e
R
[

38

d
e
h
s
i
l
o
p

s
i

e
l
p
m
a
s

e
h
t

d
n
a

y
a
w
a

d
e
h
c
t
e

s
i

e
t
a
r
t
s
b
u
s

e
h
t

r
e
t
f
a

d
e
r
u
s
a
e
m

e
c
n
a
t
t
i
m
s
n
a
r
t

,
m
l
ﬁ

g
n
i
d
n
a
t
s
-
e
e
r
f

:
b

y
a
w
a

d
e
h
c
t
e

s
i

e
t
a
r
t
s
b
u
s

e
h
t

r
e
t
f
a

d
e
r
u
s
a
e
m

e
c
n
a
t
t
i
m
s
n
a
r
t

,
m
l
ﬁ

g
n
i
d
n
a
t
s
-
e
e
r
f

:
a

d
e
h
s
i
l
o
p

s
i

e
c
a
f
r
u
s

s
’
d
n
o
m
a
i
d

e
h
t

r
e
t
f
a

d
e
r
u
s
a
e
m

e
c
n
a
t
t
i
m
s
n
a
r
t

:
c

d
e
d
d
a

s
a
w

2
0
C
,
s
e
s
a
c

e
m
o
s

n
i

:
2
H
:
4
H
C
*

C
0

0
0
1
1

t
a

n
e
g
y
x
o

y
r
d

y
t
i
r
u
p

h
g
i
h

h
t
i
w

e
c
a
n
r
u
f

a

n
i

n
w
o
r
g

y
l
l
a
m
r
e
h
t

s
a
w

r
e
y
a
l

2
0
i
S

k
c
i
h
t
-
m
n

0
5
2

:
Z
O
i
S

e
r
u
t
a
r
e
p
m
e
t

e
t
a
r
t
s
b
u
s

:
5
T

s
s
e
n
h
g
u
o
r

e
c
a
f
r
u
s

:
.
R
.
S

s
s
e
n
k
c
i
h
t

m
l
ﬁ

:
m
l
i
F
t

39

Due to the signiﬁcant impact of the surface roughness on optical transmission of

diamond coated samples, i.e. the reﬂected and transmitted light from the rough surfaces

are scattered in various directions, which leads to poor transparency. Consequently, in

recent years, the effort of depositing diamond optical coatings has been directed toward

nanocrystalline diamond. The smooth surfaces of nanocrystalline diamond ﬁlms have

enhanced the optical transmittance of as-grown diamond ﬁlms for optical coating applica-

tions. Table 2.4 summarizes the past experiments concerned with the deposition of

nanocrystalline diamond for optical application. The different deposition methodologies

that were employed for nanocrystalline diamond deposition are summarized along with

the transmission performance of the resulting structures. The results of a number of these

investigations are brieﬂy described below.

In 1998, the optical transparency of nanocrystalline diamond ﬁlms grown using

methane, hydrogen, and oxygen gas mixture in a microwave reactor with different sizes of

diamond powder for substrate (quartz) pretreatment was investigated [Bhus 1998]. A few

discoveries was made by Bhusari et a1. and Chen et a1. [Bhus 1998], [Chen 2000] and

summarized below.

(1) The optical transmittance and surface roughness data exhibited some crossover

behavior.

(2) The ﬁlms reached the saturation of optical transmittance of the ﬁlms around 80-84%

even though the surface roughness continued to decrease from 13 nm to 6 nm.

40

(3) as long as the sp2 bonded carbon in the ﬁlm was minimum, the surface roughness was

the dominant factor that controls the optical transmission of the ﬁlms,

(4) the grain size of the nanocrystalline diamond ﬁlms did not have as much inﬂuence as

the surface roughness on the optical transparency, and

(5) smooth and highly transparent nanocrystalline diamond ﬁlms could be synthesized

with (i) low methane concentration (between 4-20%) on ﬁner-diamond-powder (4

nm)-scratched quartz substrates, and (ii) high methane concentration (between

20-42%) on coarser-diamond-powder (0.1 um)-scratched quartz substrates. The opti-

cal transmittance was around 80% at a wavelength 700 nm and was achieved on the

quartz substrate.

The comparison of the optical properties of free-standing diamond and nano-dia-

mond was performed by Sharda et al. [Shar 2001]. Although the nano-diamond grown at

600 0C had a somewhat higher absorption coefﬁcient, its lower surface roughness (17 nm

of nano-diamond versus 60 nm of polycrystalline diamond) gave it the potential for appli-

cations as optical windows in the near IR region (transmittance close to 78%).

A 0.1 rim-thick nanocrystalline diamond protective coating on quartz glass, syn-

thesized by microwave chemical vapor deposition system with methane and hydrogen gas

mixtures, achieved 65% of optical transmittance in the visible light range when the surface

roughness was controlled under 12 nm [Yang 2002].

41

Jing et al. used a microwave plasma enhanced chemical vapor deposition tech-

nique to grow nanocrystalline diamond on both sides of a silicon wafer with thickness of

0.3 pm and 0.5 pm, and reported a maximum transmittance of 85% at the wavenumber of

2200 cm" (~45 um) [Jing 2002].

,

.
.

_
.
.

—
4
ﬁ
P
“

42

E
H
T
D
N
A
Y
G
O
L
O
D
O
H
T
E
M
N
O
I
T
I
S
O
P
E
D
M
L
I
F
D
N
O
M
A
I
D
E
N
I
L
L
A
T
S
Y
R
C
O
N
A
N

:
4
.
2
E
L
B
A
T

C
N
A
T
T
I
M
S
N
A
R
T

E 

l
a
t
s
y
r
C

e
z
i
S

.
R
S

)
m
n
(

e
c
n
a
t
t
i
m
s
n
a
r
T

m
l
i
F
t

s
T

r
e
w
o
P

)
m
u
(

)
C
°
(

)
W
(

)
r
r
o
T
(

w
o
l
F

e
t
a
R

s
a
G

y
r
t
s
i
m
e
h
C

t
n
e
m
t
a
e
r
t
e
r
P

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

m
n
(

) 

)
m
c
c
s
(

)
r
e
d
w
o
p

o
/
c
a
m
a

a

7
1

1
.
1

0
0
6

-
w
e
f

0
0
0
1

  M

D
V
C
P

d
e
s
a
i
B

2
H
Z
4
H
C

)
0
0
1
(
i
S

A
N

0
3

w
e
f

a

)
m
u

2
-
2
.
0
(

0
0
1
:
5
=

.
d
e
c
n
a
h
n
E

]
1
0
0
2

r
a
h
S
[

a
-
4

w
e
f

0
0
0
1

6
-
0
9
5

6
.
0
-
5
.
0

%
4
8
-
0
3

  M

.
0
&
m
n
4

D
V
C
P

2
0
2
2
H
I
4
H
C

z
t
r
a
u
q

0
0
2

m
p

2
2

1

a
-
4

w
e
f

0
0
0
1

6
-
0
9
5

6
.
0
-
5
.
0

%
4
8
—
0
8

  M

2
0
I
2
H
I
4
H
C

D
V
C
P

z
t
r
a
u
q

m
n
4

0
0
2

2
2

0
0
1

0
0
2

m
p
7
.
0

t
a

:
)
2
4
4
(
=

c
i
n
o
s
a
r
t
l
u

]
8
9
9
1

s
u
h
B
[

:
)
9
.
7
5
-
9
.
5
9
(

]
.
0

g
n
i
h
s
i
l
o
p

]
0
0
0
2

n
e
h
C
[

0
0
1

m
p
7
.
0

t
a

:
)
0
2
-
4
(
=

c
i
n
o
s
a
r
t
l
u

]
8
9
9
1

s
u
h
B
[

:
)
9
.
9
7
-
9
.
5
9
(

]
.
0

g
n
i
h
s
i
l
o
p

]
0
0
0
2

n
e
h
C
[

a
-
4

w
e
f

0
0
0
1

6
-
0
9
5

6
.
0
-
5
.
0

%
4
8
-
0
8

  M

D
V
C
P

m
.
p
1
.
0

z
t
r
a
u
q

2
0
2
2
H
I
4
H
C

0
0
2

2
2

0
0
1

0
0
2

m
p
7
.
0

t
a

:
)
2
4
-
0
2
(
=

c
i
n
o
s
a
r
t
l
u

]
8
9
9
1

s
u
h
B
[

:
)
9
.
7
5
-
9
.
9
7
(

l

.
0

g
n
i
h
s
i
l
o
p

]
0
0
0
2

n
e
h
C
[

43

1

0
1

3
.
1

5
6

2
1

a

a

4
3

w
e
f

0
0
7

0
0
5

)
m
u

-
w
e
f

)
m
m

0
.
l
~

0
0
0
1

0
0
5
1

 

2
-
2
.
0
(

 

 

 

 

 

%
1
9
-
0
6

%
6
3
-
0
a

5
.
2
-
7
6
.
0
(

%
5
6
=
.
e
v
a

)
m
u
9
7
.
0
-
7
6
.
0
(

  
0  M
0  

2
.
0
:
8
.
6
9
:
3
=

D
V
C
W
M

.
d
e
c
n
a
h
n
E

g
n
i
h
s
i
l
o
p

c
i
n
o
s
a
r
t
l
u

2
0
Z
2
H
Z
4
H
C

)
0
0
1
(
i
S

 

z
t
r
a
u
q

m
p
5
.
0

D
V
C
P

 

V
0
6
2
-

V
0
6
2
-

d
e
s
a
i
B

g
n
a
Y
[

]
3
0
0
2

]
1
0
0
2

]
2
0
0
2

2
H
I
4
H
C

r
a
h
S
[

r
a
h
S
[

0
0
2
~

 

 

 

A
N

0
0
1

]
3
0
0
2

r
a
h
S
[

:
5
=

0
3

0
3

t

  

e 

u
n
i
t
n
o
C

e
l
b
a
T

4
.
2

l
a
t
s
y
r
C

e
z
i
S

)
m
n
(

)
m
n
(

.
R
S

e
c
n
a
t
t
i
m
s
n
a
r
T

m
a
r
t

)
m
u
(

r
e
w
o
P

)
W
(

)
n
o
T
(

w
o
l
F

e
t
a
R

)
m
c
c
s
(

s
a
G

y
r
t
s
i
m
e
h
C

)
r
e
d
w
o
p

t
n
e
m
t
a
e
r
t
e
r
P

d
n
o
m
a
i
d
(

e
t
a
r
t
s
b
u
S

r
o
t
c
a
e
R

A
N

A
N

%
6
6
-
4
5

5
.
0

0
0
5
2

0
3

A
N

2
H
1
4
H
C

c
i
n
o
s
a
r
t
l
u

i
S
-
n

D
V
C
E
P
M

 

  

=

)
m
u
3
3
.
8
-
8
7
.
2
(

)
m
u
7
6
.
6
-
4
4
.
4
(

%
6
6
=
.
e
v
a

0
0
1
:
1
=

g
n
i
h
s
i
l
o
p

]
2
0
0
2

i
L
[

7
2

A
N

0
5
6

0
0
5
2

&
5
.
0

%
5
8
-
6
6

  M

D
V
C
E
P

c
i
n
o
s
a
r
t
l
u

i
S
-
n

A
N

0
3

2
H
I
4
H
C

)
m
u
7
6
.
6
-
3
3
.
3
(

%
0
8
=
.
e
v
a

)
m
u
3
3
.
8
-
8
7
.
2
(

3
.
0

0
0
1
:
1
=

g
n
i
h
s
i
l
o
p

]
2
0
0
2

i
L
[

3
.
1

0
5
6

0
0
5
2

A
N

A
N

%
3
5
-
8
4

  M

D
V
C
E
P

c
i
n
o
s
a
r
t
l
u

2
H
Z
4
H
C

i
S
-
n

A
N

0
3

&

3
.
1

A
N

A
N

0
0
5
2

%
0
8
-
8
5

0
5
6

  M

D
V
C
E
P

c
i
n
o
s
a
r
t
l
u

2
H
I
4
H
C

A
N

i
S
-
n

0
3

)
m
u
3
.
4
1
-
6
7
.
4
(

0
0
1
:
1
=

g
n
i
h
s
i
l
o
p

]
2
0
0
2

i
L
[

)
m
.
p
5
.
2
1
-
7
6
.
6
(

%
5
7
=
.
e
v
a

)
m
.
p
3
.
4
1
-
6
7
.
4
(

9
.
0

0
0
1
:
1
=

g
n
i
h
s
i
l
o
p

]
2
0
0
2

i
L
[

0
2

 

 

 

 

0
0
8

0
5
5

4
-
0
2

)
m
u

%
6
6
-
5

 

 

8
8
.
0
-
5
8
1
.
0
(

  
  R

 

 

 

D
V
C
M
T

d
e
h
c
t
a
r
c
s

 

d
n
o
m
a
i
d

 

r
e
d
w
o
p

2
H
1
2
0
2
4
H
C

s
u
o
e
u
q
a

 

z
t
r
a
u
q

d
e
t
u
l
i
d

,
a
i
d
e
m

]
3
9
9
1

5
.
7
3

z
r
E
[

1
0
.
0

7
9
:
1

0
0
1

m
u

:
2
=

n
i

y
a
w
a

d
e
h
c
t
e

s
i

e
t
a
r
t
s
b
u
s

e
h
t

r
e
t
f
a

d
e
r
u
s
a
e
m

e
c
n
a
t
t
i
m
s
n
a
r
t

,
m
l
ﬁ

g
n
i
d
n
a
t
s
-
e
e
r
f

:
a

n
o
i
t
i
s
o
p
e
d

r
o
p
a
v

l
a
c
i
m
e
h
c

e
v
a
w
o
r
c
i
m

r
a
l
u
b
u
t

e
t
o
m
e
r

:
D
V
C
M
T
R

s
t
n
e
m
i
r
e
p
x
e
f
o

s
r
u
o
h

2

r
e
t
f
a

d
e
d
d
a

s
a
w

2
0

:
*
2
0
:
2
H
:
4
H
C

2.4.2 Tribological Application-Seal Coatings

Since the ﬁrst success of chemical vapor deposition diamond synthesis, there has

been great interest in producing high-quality, smooth diamond ﬁlms on various substrates

(for example, steel, WC-Co, SiC, etc.) for use in a wide range of engineering applications.

This is because diamond, the hardest among all materials, has high thermal conductivity

and is impervious to most chemicals. However, microcrystalline diamond ﬁlms are usually

rough due to the large faceted crystallinity, which makes post processing, like polishing, a

necessity to improve its friction and wear characteristics. Thick polycrystalline diamond

ﬁlms grown by the hot ﬁlament CVD method was coated on cemented carbide. However,

it only reduced the wear of cutting edge if the cutting path was short [Mori 1998], [Endl

1999]. When the polycrystalline diamond ﬁlms were used in sliding wear applications and

machining, their rough surfaces cause severe wear damage on the mating surfaces and

high frictional loss [Hayw 1992].

The importance of smooth diamond surface in tribological contacts has been theo-

retically studied and experimentally explored [Case 1973], [Miyo 1993], [H01] 1994],

[Kohz 1994]. Miyoshi et al. tested diamond coated pins sliding against diamond coatings

in both dry nitrogen and humid air and found that the initial friction coefﬁcients were 0.14

and 0.6 with the ﬁlm surface roughness of 15 nm and 160 nm, respectively [Miyo 1993].

When the roughness of diamond ﬁlms was reduced from 530 nm to 75 nm, a 50-75%

decrease in friction for sliding against aluminum and about 50% for carbon steel were

Observed by Hollman et a1. [H01] 1994]. As a result, the development of smooth, as-depos-

ited nanocrystalline diamond coatings is desirable. Table 2.5 shows the results of a number

45

of different investigations concerned with diamond ﬁlm for tribological application. Both

the growth methods and testing techniques are described.

46

m
(

)
m
n
(

)
m
u
(

)
m
.
N
/
3
m
m
(

) 

e
u
q
i
n
h
c
e
T

d
o
h
t
e
M

N
E
I
C
I
F
F
E
O
C
N
O
I
T
C
I
R
F
E
H
T
D
N
A
Y
G
O
L
O
D
O
H
T
E
M
N
O
I
T
I
S
O
P
E
D
M
L
I
F
D
N
O
M
A
I
D
:
5
.
2
E
L
B
A
T

T 

.
D
.
S

r
a
e
W

n
o
i
t
c
i
r
F

.
R
.
S

m
l
i
F
t

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

n
o
i
t
i
s
o
p
e
D

e
l
p
m
a
S

r
o
t
c
a
e
R

3
'
0
1
x
2
.
4
3

5
5

A
N

A
N

A
N

  N

2
M
]
S
I
A

A

A
N

e
v
o
b
a

e
m
a
S

s
a

A
N

A
N

0
9
9

5
1
2
1

,
N
0
6

4
8
.
0
:
r
a
l
p
o
p

  R

D
V
C
A
P
F

,
k
s
i
d
-
n
o
-
n
i
p

,
n
i
p
d
o
o
w

C
C
/
D
C
N

0
0
1
-
0
5

,
m
c
c
s

,
4
H
C

3
1
1
1
1
1
1

l
e
e
t
S

]
3
0
0
2

v
l
i
S
[

8
8
.
0
:
k
a
O

c
e
s
/
m

l
=
d
e
e
p
s

g
n
i
d
i
l
s

,
n
i
m

0
0
3

,
W
k

3

,
a
P
0
5

]
1
0
0
2

d
e
i
N
[

 

5
5

3
'
0
1
x
5
.
2

3
m
m

A
N

0
8
2
6
1

5
.
3
1

:
D
C
P

,
y
r
t
e
m
o
e
g

t
a
ﬂ
-
n
o
-
e
r
e
h
p
s

C
/
i
N
/
D
C
P

d
n
o
m
a
i
d
m
p
1

D
V
C
P
M

4
-
3

:
i
N

y
b
s
u
t
a
r
a
p
p
a

n
o
i
s
a
r
b
a
-
o
r
c
i
n
r

l
e
e
t
S
/
i
T
/
u

c
i
n
o
s
a
r
t
l
u

r
e
d
w
o
p

]
3
0
0
2

v
l
i
S
[

5
.
2
-
5
.
0

:
i
T

-
a
i
d

d
e
z
i
s
-
m
i
r

0
2
-
0
1
(

y
r
r
u
l
s

6
3
-
2
3

:
u
C

d
n
o
m
a
i
d

h
t
i
w

l
l
a
b

g
n
i
t
a
t
o
r

a

g
n
i
h
s
i
l
o
p

.
)
r
e
t
a
w

n
i

s
e
l
c
i
t
r
a
p

d
n
o
m

,
l
l
a
b

l
e
e
t
s

0
0
1
2
5

]
S
I
A

n
o
i
t
a
t
o
r

,
m
m

5
2
=
.
a
i
d

8
3
.
8
(

m
p
r

0
8
=
d
e
e
p
s

d
a
o
l

l
a
m
r
o
n
N
5
2
.
0

,
)
c
e
s
/
d
a
r

3
'
0
1
x
5
2
.
6
2

5
5

:
i
N

4
-
3

A
N

A
N

  N

/
i
‘
I
/
u
C
/
i
N

-
o
r
t
c
e
l
e

u
C
&
i
N

e
m
a
S

y
b

e
v
o
b
a

A

s
a

3
m
m

6
3
-
2
3

:
u
C

5
.
2
-
5
.
0

:
i
T

d
o
h
t
e
m
D
V
P
y
b

i
T

l
e
e
t
S

,
d
o
h
t
e
m

g
n
i
t
a
l
p

]
3
0
0
2

v
l
i
S
[

47

l

.
1

A
N

A
N

A
N

A
N

0
9
9

0
9
9

0
3
9

 

5
7
5
1

 

:
r
a
l
p
o
p

 

3
8
.
0
:
k
a
O

 

 

6
8
.
0
:
k
a
O

 

8
8
.
0
:
r
a
l
p
o
p

  R
  N
  

 

 

C
°
0
0
9
=
s
T

 

D
V
C
A
P
F

C
C
/
D
C
N

0
0
1
-
0
5

]
1
0
0
2

]
1
0
0
2

,
m
c
c
s

d
e
i
N
[

d
e
i
N
[

,
a
P
0
5

e
v
o
b
a

,
4
H
C

e
m
a
S

e
v
o
b
a

,
n
i
m

e
m
a
S

,
W
k

A
N

C
C

0
3

s
a

s
a

A

3

C
°
0
0
9
=
s
T

 

.
D
8

r
a
e
W

)
n
I
(

)
m
.
N
/
3
m
m
(

n
o
i
t
c
i
r
F

.
2
1
.
8

)
m
n
(

m
l
i
F
t

)
m
u
(

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

e
u
q
i
n
h
c
e
T

e
l
p
m
a
S

n
o
i
t
i
s
o
p
e
D

d
o
h
t
e
M

r
o
t
c
a
e
R

 

e
u
n
i
i
n
o
C

5
.
2

e
l
b
a
T

7
'
0
1
x
6
.
1

:
C
i
S

0
2
7

o
~

:
D
C
N

 4
-  1

7
0
.
0
-
5
0
.
0

0

2
.
2

,
r
e
t
e
m
o
b
i
r
t

k
s
i
d
-
n
o
-
n
i
p

i
S
/
D
C
N

,
r
r
o
T
4
2

,
2
H
-
4
H
C

%
0
1

D
V
C
P
W
M

y
r
d
,
m
m
6
3
.
6
=
.
a
i
d

,
l
l
a
b

C
i
S

c
e
s
/
m
c

5
=
d
e
e
p
s

,
N

5
.
0

,
r
i
a

,
C
°
0
8
6
-
0
4
6
=
S
T

]
2
0
0
2

g
n
o
H
[

 

d
n
o
m
a
i
d
m
n
5
&
m
u

1

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

,
W
k

2
.
2
-
8
.
1

0
1

2
.
2

A
N

0
2
7

1
1
.
0

  M

D
V
C
P
W

,
2
H
-
4
H
C

k
s
i
d
-
n
o
-
n
i
p

,
r
e
t
e
m
o
b
i
r
t

i
S
/
D
C
N

,
r
r
o
T
4
2

%
0
1

y
r
d

,
m
m
6
3
.
6
=
.
a
i
d

,
l
l
a
b

C
i
S

c
e
s
/
m
c

0
1
=
d
e
e
p
s

,
N

1

,
r
i
a

,
C
°
0
8
6
-
0
4
6
=
s
T

]
2
0
0
2

g
n
o
H
[

d
n
o
m
a
i
d
m
n
5
&
m
u
l

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

.
W
k

2
.
2
-
8
.
1

:
a

0
3

,
1
.
0

0
2
1

A
N

A
N

,
N
2

,
l
l
a
b

  B

,
k
s
i
d
-
n
o
-
l
l
a
b

,
1
2
9
.
0
=
2
H
2
C
2
2
0

C
C
/
D
C
N

d
e
s
a
i

S
B
B

t
o
H

48

,
3
.
0

:
l
a
i
t
i
n
i
(

m
0
1
—
5

r
o
f

)
r
i
a

n
i

,
8
0
.
0

:
w

,
7
0
.
0

:
o

,
)
C
°
0
2

,
H
R

%
0
5
-
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o
"

,
r
e
t
a
w
A

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

d
n
o
m
a
i
d
m
u
l

,
C
°
0
3
i

0
5
8
=
s
T

]
8
9
9
1

1
1
0
H
[

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

,
m
c
c
s

0
6
9
1

e
m
a
l
F

,
8
.
0

:
a

0
2
1

A
N

A
N

,
N
2

,
l
l
a
b

A
N

  N

,
k
s
i
d
-
n
o
-
l
l
a
b

S
B
B

S
B
B

A
N

A

 

 

 

,
4
.
0

:
w

3
1
.
0

:
o

 

 

 

,
)
C
°
0
2

,
H
R

%
0
5
-
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o

,
r
e
t
a
w

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

 

 

 

]
8
9
9
1

l
l
o
H
[

]
8
9
9
1

1
1
0
H
[

 

8
:
0
1
1
1
1
.
2

0
4
6
8

A
N

A
N

A
N

,
N
2

,
l
l
a
b

  N

,
k
s
i
d
-
n
o
-
l
l
a
b

S
B
B

S
B
B

A
N

A

 

~

(
I
D
-
u
h
n
h
n
n
.

.
u
‘
n
‘
t
n

N

n
o
i
t
c
i
r
F

.
0
.
8

r
a
e
W

.
R
.
S

m
l
i
F
t

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

n
o
i
t
i
s
o
p
e
D

e
l
p
m
a
S

r
o
t
c
a
e
R

 

e
u
n
i
t
n
o
C

5
.
2

e
l
b
a
T

)
m
n
(

m
(

)
m
u
(

)
m
.
N
/
3
m
m
(

) 

d
o
h
t
e
M

e
u
q
i
n
h
c
e
T

 

 

z

0
2
1

A
N

,
8
0
.
0

:
a

0
3

A
N

,
N
2

,
l
l
a
b
C
C
,
k
s
i
d
-
n
o
-
l
l
a
b

C
C
/
D
C
N

,

I
2
9
0
=
2
H
2
C
2
2
0

t
o
H

d
e
s
a
i
B

.
3
.
0

:
l
a
i
t
i
n
i
(

m
0
1
-
5

r
o
f

)
r
i
a

n
i

,
7
0
.
0

:
w

,
8
0
.
0

:
o

,
)
C
°
0
2

,
H
R

%
0
5
-
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o
"

,
r
e
t
a
w
"

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

d
n
o
m
a
i
d
m
u
l

,
C
°
0
3
1

0
5
8
=
s
T

]
8
9
9
1

l
l
o
H
[

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

,
m
c
c
s

0
6
9
1

e
m
a
l
F

:
a

A
N

0
2
1

A
N

A
N

,
N
2

,
5
7
.
0

  N

,
l
l
a
b
C
C
,
k
s
i
d
-
n
o
-
l
l
a
b

A
N

S
B
B

A

 

0
4
6
8

O
"
0
1
x
9
.
1

A
N

0
3

A
N

,
N
2

,
l
l
a
b

S
S

,
k
s
i
d
-
n
o
-
l
l
a
b

C
C
/
D
C
N

e
v
o
b
a

s
a

e
m
a
S

t
o
H

d
e
s
a
i
B

,
3
.
0

:
l
a
i
t
i
n
i
(

m
0
1
-
5

r
o
f

)
r
i
a

n
i

,
6
0
.
0

:
w

,
8
0
.
0

:
o

,
)
C
°
0
2

,
H
R

%
0
5
-
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o
"

,
r
e
t
a
w
"

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

d
n
o
m
a
i
d

m
1
1
1

,
c
°
0
3
:

0
5
8
=
s
T

,
m
c
c
s

0
6
9
1

e
m
a
l
F

]
8
9
9
1

]
1
0
H
[

 

0
2
1

A
N

,
7
0
.
0

:
a

0
3

A
N

,
N
2

,
l
l
a
b

S
S

,
k
s
i
d
-
n
o
-
l
l
a
b

C
C
/
D
C
N

,

I

Z
9
.
0
=
2
H
2
C
2
2
0

t
o
H

d
e
s
a
i
B

49

,
2
.
0

:
w

1
.
0

:
o

,
)
C
°
0
2

,
H
R

%
0
5
-
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o

,
r
e
t
a
w

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

]
8
9
9
1

l
l
e
H
[

 

 

 

 

 

 

0
2
1

A
N

A
N

A
N

,
3
.
0

:
w

1
1
.
0

:
o

,
6
.
0

:
a

,
)
C
°
0
2

  
 

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

,
k
s
i
d
-
n
o
-
l
l
a
b

1
.
0
=
d
e
e
p
s

 

%
0
5
-
0
3
(

%
0
5
-
0
3
(

,
)
C
°
0
2

,
c
e
s
/
m

,
r
e
t
a
w

]
8
9
9
1

l
l
e
H
[

]
8
9
9
1

]
1
0
H
[

S
B
B

.
H
R

,
N
2

,
l
l
a
b

S
S

,
H
R

r
i
a

l
i
o

A
N

A
N

)
l
i
o

r
i
a

 

 

,
c
e
s
/
m

1
.
0
=
d
e
e
p
s

e
m
a
l
F

u
l
n
a
-
o
i
n
u
u
v

\

.
D
S

r
a
e
W

)
m
(

)
m
.
N
/
3
m
m
(

n
o
i
t
c
i
r
F

.
R
.
S

)
m
n
(

n
a
‘

)
m
u
(

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

e
u
q
i
n
h
c
e
T

e
l
p
m
a
S

n
o
i
t
i
s
o
p
e
D

d
o
h
t
e
M

r
o
t
c
a
e
R

 

e
u
n
i
t
n
o
C

5
.
2

e
l
b
a
T

 

 

r

 

 

 
 

n
=
l

:
a

0
3

A
N

0
2
1

A
N

,
N
2

,
3
5
.
0

  B

,
k
s
i
d
-
n
o
-
l
l
a
b

C
C
/
D
C
N

2
9
.
0
=
2
H
2
C
2
2
0

d
e
s
a
i

,
l
l
a
b

t
o
H

l
A

I

,

0
2
1

A
N

,
5
.
0

:
a

0
3

A
N

,
N
2

,
l
l
a
b

i
T

,
k
s
i
0
-
f
r
0
-
l
l
a
b

C
C
/
D
C
N

,
1
I
9
.
0
=
2
H
2
C
1
2
0

t
o
H

d
e
s
a
i
B

,
5
3
.
0

:
w

2
3
.
0

:
o

,
)
C
°
0
2

.
H
R

%
0
5
—
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o

,
r
e
t
a
w

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

d
n
o
m
a
i
d
m
p

1

,
c
°
o
3
:

0
5
8
=
s
T

]
8
9
9
1

1
1
0
H
[

,
c
e
s
l
m

1
.
0
=
d
e
e
p
s

,
m
c
c
s

0
6
9
1

e
m
a
l
F

:
a

5
.
2

,
1
.
0

A
N

A
N

0
2
1

  B

d
e
s
a
i

e
v
o
b
a

e
v
o
b
a

e
m
a
S

e
m
a
S

t
o
H

s
a

s
a

c
c
/
D
C
N
‘

:
a

A
N

0
2
1

A
N

A
N

,
5
4
.
0

  N

e
v
o
b
a

S
B
B

e
m
a
S

A
N

A

s
a

,
6
0
.
0

:
w

,
4
0
.
0

:
o

]
8
9
9
1

1
1
0
H
[

e
m
a
l
F

,
1
4
.
0

:
w

4
.
0

:
o

]
8
9
9
1

1
1
0
H
[

50

,
]
5
.
0

:
w

1
1
.
0

:
o

,
)
C
°
0
2

,
H
R

%
0
5
-
0
3
(

r
i
a

)
l
i
o

n
ﬁ
e
l
o
a
f
l
a
y
l
o
p
(

l
i
o

,
r
e
t
a
w

,
c
e
s
l
m

1
.
0
=
d
e
e
p
s

d
e
h
c
t
a
r
c
s

r
e
d
w
o
p

d
n
o
m
a
i
d
m
p
1

,
C
0
0
3
i

0
5
8
=
s
T

,
m
c
c
s

0
6
9
1

e
m
a
l
F

]
8
9
9
1

1
1
0
H
[

:
a

,
1

.
O

:
a

:
o

:
0

:
w

:
w

5
.
2

A
N

A
N

0
2
1

A
N

A
N

0
2
1

A
N

5
1
.
0

4
0
.
0

,
3
0
.
0

,
2
5
.
0

,
5
5
.
0

 

 

 

 

 

  
  B
  N

 

 

c
c
/
D
C
N
’

 

e
m
a
l
F

]
8
9
9
1

1
1
0
H
[

d
e
s
a
i

]
8
9
9
1

e
v
o
b
a

e
v
o
b
a

1
1
0
H
[

e
v
o
b
a

S
B
B

e
m
a
S

e
m
a
S

e
m
a
S

A
N

t
o
H

 

A

s
a

s
a

s
a

\

O
o
I
I
Q
E
U
U
I
u

v

r
l
‘
t
‘
I
I

m
(

)
m
n
(

)
m
u
(

)
m
.
N
/
3
m
m
(

) 

d
o
h
t
e
M

e
u
q
i
n
h
c
e
T

n
o
i
t
c
i
r
F

.
D
.
S

r
a
e
W

.
R
.
S

m
l
i
F
[

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

n
o
i
t
i
s
o
p
e
D

e
l
p
m
a
S

r
o
t
c
a
e
R

 

e
u
n
i
t
n
o
C

5
.
2

e
l
b
a
T

 

 

.
V
C
'
I

-
°
'
0
1
x
5
.
6

4
'
0
1
x
7
.
5

6
5
.
0

:
l
a
i
t
i
n
i

c
e
s
/
m

2
.
0
-
5
0
.
0
=
d
e
e
p
s

r
o

i
S

n
o

,
r
r
o
T
0
6

,
n
i
m

0
6
-
0
3

]
6
9
9
1

e
d
r
E
[

,
5
2
.
0

5
1
3

A
N

,
r
e
t
e
m
o
b
i
r
t

k
s
i
d
-
n
o
-
n
i
p

)
1
1
1
(
D
C
P

r
o
f

V
0
5
1
-

t
a
N
E
B

D
V
C
P
M

,
m
m
5
9
=
.
.
i
a
d
,
N
2
,
l
l
a
b
4
N
3
]
S

c
r
s
‘

,
C
°

0
5
8
-
0
0
8
=
s
T

)
H
R

%
0
2
(

r
i
a

4
N
3
i
S
*

r
o

,
m
c
c
s

2
/
8
9

=
4
H
C
/
2
H

W
0
0
5
1

s
a

,
1
.
0

A
N

8
0
2

e
m
a
S

  M

)
0
0
1
(
D
C
P

k
s
i
d
-
n
o
-
n
i
p

,
r
e
t
e
m
o
b
i
r
t

D
V
C
P

e
v
o
b
a

e
m
a
S

s
a

.
V
C
I

e
v
o
b
a

3
2
.
0

:
l
a
i
t
i
n
i

c
e
s
/
m

2
.
0
-
5
0
.
0
=
d
e
e
p
s

r
o

i
S

n
o

]
6
9
9
1

e
d
r
E
[

,
m
m
5
.
9
=
.
i
a
d

,
N
2

,
l
l
a
b

4
N
3
i
S

c
r
s
’

)
H
R

%
0
2
(

r
i
a

4
N
3
i
S
‘

r
o

s
a

5
2
1

A
N

e
m
a
S

0
0
0
4

7
.
0
-
2
.
0

  M

k
s
i
d
-
n
o
-
n
i
p

,
r
e
t
e
m
o
b
i
r
t

D
V
C
P

n
o
D
C
P

e
v
o
b
a

e
m
a
S

s
a

.
v
e
r

e
v
o
b
a

)
H
R

%
0
5
-
0
3
(

r
i
a

,
2
N

y
r
d

,
m
m
5
.
9
=
.
i
a
d

,
N
2

,
l
l
a
b

4
N
3
i
S

c
e
s
/
n
r

2
.
0
-
5
0
.
0
=
d
e
e
p
s

c
i
S
‘

r
o

i
S

4
N
3
i
S
*

r
o

]
6
9
9
1

e
d
r
E
[

.
2
-
7
'
0
1
x
8
.
1

:
2
N

A
N

A
N

9
0
.
0

0
0
0
5

  M

k
s
i
d
-
n
o
-
n
i
p

,
r
e
t
e
m
o
b
i
r
t

D
V
C
P

n
o
D
C
N

t
a
N
E
B

V
0
5
1
-

r
o
f

.
v
e
r

7
0
1
x
5

5
1
.
0

:
r
i
a

,
m
m
5
.
9
=
.
i
a
d

,
N
2

,
l
l
a
b

4
N
3
i
S

c
e
s
/
m

2
.
0
—
5
0
.
0
=
d
e
e
p
s

c
i
S
’

r
o

i
S

4
N
3
i
S
'

r
o

,
C
°

0
5
8
-
0
0
8
=
s
T

,
r
r
o
T

0
0
1

,
n
i
m

0
6
-
0
3

]
6
9
9
1

e
d
r
E
[

51

.
v
e
r

s
a

,
m
c

0
4

A
N

:
r
i
a

e
m
a
S

:
2
N

e
v
o
b
a

0
0
0
8

9
1
.
0

8
0
.
0

 

 

 

 

 

  
  

2
.
0
-
5
0
.
0
=
d
e
e
p
s

k
s
i
d
-
n
o
-
n
i
p

 

,
r
e
t
e
m
o
b
i
r
t

0
5
8
-
0
0
8
=
s
T

 

 

D
V
C
P
M

%
0
5
(
r
i
a

4
N
3
i
S
"

=
0
6
C
/
z
I
'
I
/
r
A

n
o
D
C
N

t
a
N
E
B

7
.
2
1
=
r

=
4
1
-
l
C
/
2
1
-
I
/
r
A

W
0
0
8

W
0
0
8

c
e
s
/
m

V
0
5
1
-

e
d
r
E
[

]
6
9
9
1

1
/
2
/
8
9

1
1
2
/
7
9

0
6
-
0
3

c
r
s
’

,
m
c
c
s

,
m
c
c
s

%
0
4
(

)
H
R

,
r
r
o
T

0
0
1

,
2
N

y
r
d

,
n
i
m

,
N
5

,
n
i
p

C
i
S

r
o

i
S

)
H
R

r
i
a

r
o

,
C
°

,
2
N

y
r
d

r
o
f

0
0
0

1

.
v
e
r

 

 

m
(

)
m
u
(

)
m
u
(

)
m
.
N
/
3
m
m
(

) 

e
u
q
i
n
h
c
e
T

d
o
h
t
e
M

n
o
i
t
c
i
r
F

.
D
.
S

r
a
e
W

.
R
S

m
l
i
F
(

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

n
o
i
t
i
s
o
p
e
D

e
l
p
m
a
S

r
o
t
c
a
e
R

 

e
u
n
i
t
n
o
C

5
.
2

e
l
b
a
T

x
7
-
7
'
0
1
x
8
.
1

,
4
0
.
0

5
4
—
0
2

A
N

,
r
e
t
e
m
o
b
i
r
t

k
s
i
d
-
n
o
-
n
i
p

n
o
D
C
N

r
o
f

V
0
5
1
-

D
V
C
P
M

t
a
N
E

7  B

—
'
”
-
=
=

7
'
0
1

)
4
.
0

:
l
a
i
t
i
n
i
(

4
.
0
=
p
m
u
j

.
m
m
5
.
9
=
.
i
a
d

,
N
5

,
l
l
a
b

4
N
3
i
S

c
e
s
/
m

4
0
.
0
=
d
e
e
p
s

c
i
S
"

r
o

i
S

4
N
3
i
S
’

r
o

,
C
°

0
5
8
-
0
0
8
=
s
T

,
r
r
o
T

0
0
1

,
n
i
m

0
6
-
0
3

]
6
9
9
1

e
d
r
E
[

,
2
N

y
r
d

=
0
6
C
/
2
1
-
I
/
r
A

W
0
0
8

,
m
c
c
s

1
/
2
/
7
9

s
a

A
N

A
N

,
4
0
.
0

e
m
a
S

0
0
0
1

  M

D
V
C
P

t
a
N
E
B

n
o
D
C
N

V
0
5
1
-

e
v
o
b
a

e
m
a
S

r
o
f

s
a

0
0
0

1

.
v
e
r

e
v
o
b
a

5
3
.
0

:
l
a
i
t
i
n
i

c
e
s
/
m

2
.
0
-
5
0
.
0
=
d
e
e
p
s

c
i
S
’

r
o

i
S

,
r
r
o
T

0
0
1

,
n
i
m

0
6
-
0
3

]
6
9
9
1

e
d
r
E
[

,
m
m
5
.
9
=
.
i
a
d

,
N
2

,
l
l
a
b

4
N
3
i
S

4
N
3
i
S
*

r
o

W
0
0
8
,
C
°
0
5
8
-
0
0
8
=
s
T

)
H
R

%
0
5
-
0
3
(

r
i
a

r
0

,
m
c
c
s

1
/
2
/
8
9

=
4
H
C
I
Z
H
/
r
A

:
0
6
C
/
z
H
M

m
c
c
s

1
/
2
/
7
9

.
V
C
I

e
v
o
b
a

)
4
.
0

:
l
a
i
t
i
n
i
(

4
.
0
=
p
m
u
j

c
i
S
*

r
o

i
S

4
N
3
i
S
*

r
o

W
0
0
8

,
m
c
c
s

1
/
2
/
8
9

,
C
°

0
5
8
-
0
0
8
=
s
T

=
4
H
C
/
2
H
/
r
A

,
r
r
o
T

0
0
1

,
n
i
m

0
6
-
0
3

]
6
9
9
1

e
d
r
E
[

a

s
a

A
N

w
e
f

A
N

e
m
a
S

,
5
1
.
0
-
3
1
.
0

  M

n
o
D
C
N

,
r
e
t
e
m
o
b
i
r
t

k
s
i
d
-
n
o
-
n
i
p

D
V
C
P

t
a
N
E
B

V
0
5
1
-

r
o
f

52

 

 

A
N

7
.
0

8
2

8
.
0

,
l
l
a
b

8
2
.
0

0
5
2

r
a
c
s

5
4
.
0

:
l
a
n
ﬁ

r
a
c
s

r
a
e
w

3
2
.
0

:
l
a
n
ﬁ

0
0
0
1

r
a
e
w

e
l
c
y
c

s
e
l
c
y
c

:
l
a
i
t
i
n
i

:
l
a
i
t
i
n
i

 

 

 

4
.
0
=
a
i
d

2
3
.
0
=
a
i
d

s  
  M
  

e
t
s
a
p
d
n
o
m
a
i
d
n
r
u

 

W
k
2
,
r
r
o
T
0
6

 

,
k
s
i
d
-
n
o
-
l
l
a
b

D
V
C
E
P
M

,
4
8
1
6
1
=
2
H
2
2
H
2
C

,
d
e
h
c
t
a
r
c
s

5
.
2
=
d
e
e
p
s

i
S
/
D
C
N
a

D
V
C
E
P

 

i
S
/
D
C
N

,
c
e
s
/
m
m

%
0
5
-
0
4

]
1
0
0
2

]
1
0
0
2

3
0
2
1
A

e
v
o
b
a

3
=
a
i
d

e
m
a
S

e
v
o
b
a

s
a

e
m
a
S

,
C
°

0
5
7

,
N
3

,
m
m

u
X
[

u
X
[

5
.
0

H
R

s
a

 

 

 

 

 

 

8
0
.
0

:
l
a
n
ﬁ

e
l
c
y
c

5
0
.
0
=
a
i
d

s 

]
1
0
0
2

u
X
[

 

 

 

 

)
m
n
(

)
m
u
(

e
u
q
i
n
h
c
e
T

e
l
p
m
a
S

)
m
.
N
/
3
m
m
(

m
(

) 

d
o
h
t
e
M

r
o
t
c
a
e
R

5
1
.
0

A
N

0
0
0
1

r
a
c
s

8
.
O
~

r
a
e
w

:
l
a
i
t
i
n
i

  M

i
S
/
D
C
N
b

D
V
C
E
P

e
v
o
b
a

e
v
o
b
a

e
m
a
S

e
m
a
S

s
a

s
a

.
D
.
S

r
a
e
W

.

.

n
o
i
t
c
i
r
F

.
R
.
S

m
l
i
p
t

n
o
i
t
a
z
i
r
e
t
c
a
r
a
h
C

n
o
i
t
i
s
o
p
e
D

 

e
u
n
i
t
n
o
C

5
.
2

e
l
b
a
T

 

.
e
u
l
a
v

e
t
a
t
s
-
y
d
a
e
t
s

a

d
e
h
c
a
e
r

d
a
h

n
o
i
t
c
i
r
f

e
h
t

n
e
h
w

g
n
i
d
i
l
s

y
r
d

f
o
m
0
1
-
5

r
e
t
f
a

d
e
c
u
d
o
r
t
n
i

s
a
w

r
e
t
a
W

:
r
e
t
a
w
"

.
e
u
l
a
v

e
t
a
t
s
-
y
d
a
e
t
s

a

d
e
h
c
a
e
r

d
a
h

n
o
i
t
c
i
r
f

e
h
t

n
e
h
w

g
n
i
d
i
l
s

y
r
d
f
o
m
0
1
-
5

r
e
t
f
a

d
e
c
u
d
o
r
t
n
i

s
a
w

l
i
O

:
l
i
o
"

C
°

0
8

t
a

2
m
c
/
s
n
o
i

°
1
0
1
x
1

t
a

n
o
i
t
a
t
n
e
m
e
l
p
m
i

2
N
:
i
S
/
D
C
N
a

d
n
o
m
a
i
d

e
n
i
l
l
a
t
s
y
r
c
o
n
a
n

d
e
h
s
i
l
o
p

y
l
t
h
g
i
l

:
D
C
N
*

C
°
0
8

t
a

2
m
c
/
s
n
o
i

7
1
0
1
x
1

t
a

n
o
i
t
a
t
n
e
m
e
l
p
m
i

2
N

:
i
S
/
D
C
N
°

)
e
g
a
r
e
v
a

e
n
i
l
r
e
t
n
e
c
(
A
L
C
m
u
2
0
.
0
t
:
1
.
0

,
d
e
h
s
i
l
o
p

:
4
N
3
i
S
*

)
e
g
a
r
e
v
a

e
n
i
l
r
e
t
n
e
c
(
A
L
C
m
u
2
0
.
0
i
1
.
0

,
d
e
h
s
i
l
o
p

:
C
i
S
"
a

m
u
n
i
m
u
l
a

:
l
A

,
m
u
i
n
a
t
i
t

:
i
T

,
r
e
p
p
o
c

:
u
C

l
e
e
t
s

g
n
i
r
a
e
b
-
l
l
a
b

:
S
B
B

e
d
i
b
r
a
c

d
e
t
n
e
m
e
c

:
C
C

l
e
e
t
s

s
s
e
l
n
i
a
t
s

:
S
S

n
o
i
t
u
l
o
v
e
r

:
.
v
e
r

e
r
u
t
a
r
e
p
m
e
t

e
t
a
r
t
s
b
u
s

:
s
T

s
s
e
n
h
g
u
o
r

e
c
a
f
r
u
s

:
.
R
.
S

s
s
e
n
k
c
i
h
t

m
l
ﬁ

:
m
m
t

e
c
n
a
t
s
i
d

g
n
i
d
i
l
s

:
.
D
.
S

y
t
i
d
i
m
u
h

e
v
i
t
a
l
e
r

:
H
R

n
o
i
t
a
e
l
c
u
n

d
e
c
n
a
h
n
e

d
e
s
a
i
b

:
N
E
B

d
n
o
m
a
i
d

e
n
i
l
l
a
t
s
y
r
c
y
l
o
p

:
D
C
P

d
n
o
m
a
i
d

e
n
i
l
l
a
t
s
y
r
c
o
n
a
n

:
D
C
N

53

2.4.3 SAW Devices Based on Nanocrystalline Diamond

Surface acoustic wave (SAW) devices are largely used as frequency ﬁlters or reso-

nators in the range of 100 MHz - 1 GHz and are applied to such applications as digital

communications, mobile phones and video systems. SAW devices have the advantage in

that their size is smaller than the equivalent electromagnetic device. Due to the rapid

development of the high speed communication systems, the demand for large volume data

transmission and mobile communication is increasing and requires devices to operate at

high frequency with high power durability and low loss. Multi-layered diamond substrates

are of interest because of their high SAW velocities which satisfy the need for increasing

the operational frequency range without reducing the electrode spacing into sub-micron

region. Diamond is not piezoelectric. Thus, it is necessary to incorporate piezoelectric

materials, like ZnO, into the composite layered structure. Higaki et a1. [Higa 1997] com-

pared the performance of ZnO/Diamond/Si SAW devices with the conventional SAW

devices composed of LiTaO3 (X-112°Y) and found that at 2.9 GHz, up to 36 dBm of the

input power, the input-output relationship was maintained for ZnO/Diamond/Si SAW ﬁl-

ters, whereas the LiTaO3 SAW ﬁlters that operated at 822 MHz, degraded signiﬁcantly

with only 27.7 dBm of input power, and the input—output relationship became nonlinear

starting from the input power of only 23 dBm. Clearly, compared to LiTaO3 SAW ﬁlters,

ZnO/Diamond/Si SAW ﬁlters were durable for 8 dB higher input power even at 3.5 times

higher frequency.

Table 2.6 lists the typical SAW velocities of layered structure with a diamond layer

and those of conventional SAW materials [Higa 1997], [Naka 2003], [Chen 2003], [Jian

54

1991], [Begh 2002], [Ferr 1999]. The surfaces of the diamond ﬁlms, which were used as

SAW device substrates were mechanically polished until a less than 3 nm surface rough-

ness was obtained.

 

TABLE 2.6: SAW VELOCITIES

Sample

Velocity

(m/sec)

 

SiOz/ZnO/IDT/Diamond/Si [Naka 2003]

9000

 

SiOZ/IDT/ZnO/Diamond/Si [Naka 2003]

10000

 

ZnO/IDT/Diamond/Si [Naka 2003]

10500

 

IDT/ZnO/Diamond/Si [Higa 1997], [Naka 2003]

11600

 

IDT/LiNbO3/Diamond/Si [N aka 2003]

11900

 

IDT/LiTaO3/Diamond/Si [Naka 2003]

10600

 

IDT/ZnO/R-sapphire [Chen 2004]

10500

 

ZnO/sapphire [Naka 2003]

5500

 

LiNbO3 (128Y cut X prop.) [Naka 2003]

3980

 

LiNbO3 (64Y cut X prop.) [Naka 2003]

4742

 

LiTaO3 (X cut 112Y prop.) [Higa 1997], [Naka 2003]

3290

 

LiTaO3 (36Y cut X prop.) [Naka 2003]

4160

 

ST-cut quartz [Naka 2003]

3160

 

 

36°Y-cut quartz [Naka 2003]

5100

 

a-H:C [Jian 1991]

4583

 

ta-C (8 nm thick) [Begh 2002]

ta-C (76 nm thick) [Ferr 1999]

 

 

DLC/Quartz (1 pm thick diamond-like carbon)

[Zhan 2001]

 

ta-C/Quartz (120 nm thick ta-C) [Zhan 2001]

3278

 

 

5138

6250

3243

 

55

A comprehensive theoretical study was carried out by Nakahata et a1. [naka 1995]

and they concluded that when diamond was combined with a piezoelectric thin ﬁlm, the

SAW velocities up to 12000 m/sec could be reached as propagating in ZnO/Diamond/Si,

LiNbO3/Diamond/Si, and LiTaO3/Diamond/Si structures. Table 2.7 shows the calculated

SAW velocities of layered Structure with a diamond layer and those of conventional SAW

materials [Naka 1995].

TABLE 2.7: CALCULATED SAW VELOCITIES [Naka 1995]

 

Sample

Velocity

(mlsec)

 

SiOZIZnO/Diamond/Si (kHZnO=0.3, kHSi02=0.45)

10710

 

Si02/ZnO/Diamond/Si (kHZnO=0.5, kHSi02=0.72)

8050

 

ZnO/Diamond/Si (kHZnO=0-5)

10520

 

ZnO/Diamond/Si (kHZnO=l .0)

7180

 

LiNbO3/Diamond/Si

11890

 

LiTaO3/Diamond/Si

10610

 

ZnO/glass

2600

 

LiNbO3 (128Y cut X prop.)

3980

 

LiNbO3 (64Y cut X prop.)

 

 

LiTaO3 (X cut 112Y prop.)

  
 

LiTaO3 (36Y cut X prop.)

ST-cut quartz

4742

3290

4160

3158

 

 

Smooth diamond surface is an important fabrication requirement for utilizing dia-

mond ﬁlms as SAW materials. Smooth diamond ﬁlms: (1) reduce the propagation loss

which affects the insertion loss of SAW devices and (2) ensure the correct generation and

propagation of surface acoustic wave. Polishing the surface of the diamond ﬁlms is a difﬁ-

56

cult task especially for large wafers. Thick nanocrystalline diamond ﬁlms, with good

insulating properties can offer considerable advantages over traditionally grown polycrys-

talline diamond in SAW applications. Therefore, the application of nanocrystalline

diamond ﬁlms to SAW substrate are of interest for the fabrication of high frequency SAW

devices.

57

3

Description of Experimental Systems and

Procedures

3.1

Introduction

This chapter, in Section 3.2, ﬁrst describes the two experimental systems, reactor

geometry, and reactor conditions that were utilized in this thesis research, including

microwave power supply, waveguide/transmission system, vacuum pump and the gas ﬂow

control system, computer control system, and reactor operating ﬁeld map. The experimen-

tal procedure, including seeding procedures, start-up and shut-down procedures are

described in Section 3.3. The experimental parameter space that was employed in the

experiments for this thesis is described in Section 3.4. and the measurement methodolo—

gies for reactor performance and ﬁlm characterization are described in Section 3.5.

3.2

Experimental systems

As shown in Figure 3.1, two MSU-MPACVD systems that were used in this thesis

research consisted of (1) a microwave cavity plasma reactor, (2) a microwave power sup-

ply and waveguide/transmission system, (3) a vacuum pump and the gas ﬂow control

system, and (4) a computer control system. These systems are described in detail in the

subsections below.

58

 

gas tanks

monitoring computer

 

 

 

 

  

   

 

 

 

MKS

gas ﬂow controllers

 

l %

 

transition unit

powerrneters

+—

.

rectangular wavegurde

 

 

 

 

 

 

 

4—4

coaxial waveguide

 

 

 

-..@ <- ressure

J

l

 

p
control unit

sliding-short

dummy load

 

 
     

  
    

  

_ «@3.

pressure

gauge

 

  

circulator—>

FE

  
LL

 

 

magnetrorbl

I

  

 

 
 

microwave powersupply

 

 

  
 

 
  

process chamber

throttle valve

pump

L—D—S—
 
 L
1

1

system vent

valve

exhaust

FIGURE 3.1 : MSU-MPACVD SYSTEM

59

 

3.2.1 Microwave Power Supply and Waveguide/Transmission System

Figure

3.2 shows a

schematic drawing of the microwave power and

waveguide/transmission system. The microwave power supply (1) consists of a magnetron

(2), a circulator (3) and a dummy load (4). The waveguide/transmission system consists of

the rectangular waveguides (5), a dual-directional coupler (6), incident and reﬂected

power meters (7) and (8), a ﬂexible waveguide (9), and a waveguide to coaxial transition

unit ( 10).

The microwave power supplied by the magnetron (2) is transmitted to a coaxial

waveguide (11) by the transition unit (10) after it propagates through waveguides (5) and

(9). Microwave power is coupled into the cavity applicator (14) through a mechanically

tunable coaxial excitation probe (12). The excitation probe (12) is located at the center of

the sliding short (13). Both the excitation probe depth, LP’ and the sliding short position,

Ls, can be manually adjusted. That is, LS and Lp can be moved up and down along the

reactor axis and can be adjusted independently from each other. Whenever there is a tun-

ing mismatch between the impedance of the cavity applicator and the waveguides, some of

the incident power is reﬂected back from the cavity applicator and propagates in the oppo-

site direction of the

incident power. This reﬂected power passes through the

dual-directional coupler (6), and is directed by the circulator (3) into a matched dummy

load (4), where it is absorbed and dissipated as thermal energy. The circulator and the

matched dummy load protect the power source from being damaged by preventing the

propagation of the reﬂected power back into the power supply. The microwave power

source is a Cober (Model No. S6F/4503), 2.45 GHz, 6 kW power source. The dual-direc-

60

tional power coupler attenuation factors for incident and reﬂected power are 55.6 dB and

55.7 dB for system I, and for system I] are 55.88 dB and 55.96 dB respectively. Pabs

(absorbed power) = Pin (input power) - Pref (reﬂected power).

(15) Monitoring computer

\

 

 

 
   

 

 

 

(7) Incident Power Meter

    

(8) Reﬂected Power Meter

(5) Rectangular Waveguide

 

_= (6) Dual-Direction Power Coupler

(4) Dummy Load===
 

  

 

 

(I

 

 

(3)Circulator

(11) Coaxial Waveguide

  

(12) Excitation Probe

  

(l3) Sliding Short4-

:

._-_ '. I I .- 2

, T

‘__> j _______ E .61"

‘

or.

[J (2) Magnetron

 
   
(14) CaVity Applicator a @F

1 _______

 

T

(1) Microwave Power Supply

 

Process Chamber  

 

FIGURE 3.2 : MICROWAVE POWER SUPPLY AND WAVEGUIDE/TRANSMISSION SYSTEM

FOR MSU-MPACVD REACTOR

61

  

 

3.2.2 Vacuum Pump and Gas Flow Control System

3.2.2.1 MSU-MPACVD System I

Figure 3.3 displays a schematic drawing of the vacuum pump and gas ﬂow control

systems for the MSU-MPACVD reactor 1. The source gases consist of Ar (1), H2 (2), H2

(3), and CH4 (4) with a respective purity of 99.999%, 99.999%, 99.999%, and 99.99%.

The gas ﬂow control is monitored by four MKS type 1159 mass ﬂow controllers (5) along

with a 4-channel MKS type 247C ﬂow controller (6). The system operator sets the ﬂow

rates of the source gases and the values are displayed on the 4-channel MKS ﬂow control-

ler (6). The source gases with desired ﬂow rates are mixed before they enter the baseplate

(7). An ALCATEL 2033 type mechanical roughing pump (16) is used to pump down the

chamber pressure to ~ 0.01 Torr, which is measured by a MKS type 286 TC-l (Thermal

Conductivity) vacuum gauge (11). The system operator sets the pressure on the MKS type

651C pressure controller (14) and its value is displayed on the Baratron 1000 Torr gauge

(13). An automatic throttle valve (10) controls the pressure of the process chamber (8)

with the value set in the pressure controller. A manual valve (9), normally open, is

installed and can be used to manually control the pressure of the chamber if the automatic

throttle valve malfunctions. During the experiment, the manual valve is open and the

chamber pressure is controlled by the throttle valve. A pirani gauge (12) is installed for

routine leakage check. Nitrogen gas (17) is released through the system vent valve (15) to

bring the chamber pressure up to atmospheric pressure. It is also used to dilute the exhaust

gases (19) through an exhaust valve (18) with the N2 ﬂow rate equal to or greater than

twenty times of the total ﬂow rate. The mixing of nitrogen into the exhaust gases ensures

that the exhaust gases are not in a ﬂammable condition.

62

(6) 4-channel MKS Flow Controller

[00057

 

(5) Gas Flow Controllers

 

(4)

(3)

H4

H2

 

(7) Base Plate

\EE

 

 

 

4— (8) Process Chamber

 

 
 

 

 

 

(12) Pirani
 

(11) TC-l
 

Gauge

 
  

(13) 1000 Torr

Gauge

Gauge 
 

 

  

  

 

 

(9) Manual Valve

 
  

 

[T

 

 

L—

 

(10) Automatic Throttle

{TIZjO—r
 
  

  
 

Valve

(14) Pressure Controller

 

N605) System Vent
 

Valve

 

 

 

(16) Mechanical Roughing

Pump

 
  

   
 

OD 
 
 

 
  

 

 

(18) Exhaust Valve

(19) Exhaust

FIGURE 3.3 : VACUUM PUMP AND THE GAS FLOW CONTROLLER SYSTEM

FOR MSU-MPACVD REACTOR I

63

 

3.2.2.2 MSU-MPACVD System I]

Figure 3.4 displays a schematic drawing of the vacuum pump and gas ﬂow control

systems for MSU-MPACVD reactor II. The source gases consist of N2 impurity pre-mixed

tanks, Nz-Hz 1% (1), Nz-Hz 2% (2), NZ-Ar 5% (3), Ar (4), H2 (5), CH4 (6), and He (7)

with a respective purity of 99.995%, 99.995%, 99.995%, 99.999%, 99.9995%, 99.99%,

and 99.995%. The gas ﬂow control is monitored by six MKS type 1179 mass ﬂow control-

lers (8) along with a 8-channel MKS type 247C ﬂow controller (9). The system operator

sets the ﬂow rates of source gases and the values are displayed on this 8-channel MKS

ﬂow controller. A process gas ﬂow valve (labeled as gas on the panel) (10) is installed to

have manual control on when the input gases are following into the reactor and isolates the

gas lines from the main chamber. The source gases with desired ﬂow rates are mixed

before they enter the baseplate (11). A mechanical roughing pump (24) is used to pump

down the chamber pressure to the low pressure required for the ALCATEL PTM 5150

turbo molecular pump (15) to work efﬁciently, and to remove gases from the outlet of the

turbo pump. A turbo isolation valve (23) is connected between the turbo pump and the

mechanic roughing pump to protect the turbo pump. It’s only open when the system pres-

sure is below 10'3 Torr. A SensaVac 919 hot cathode ion gauge (13) is used to measure the

pressure of this high vacuum system. A PPT Quadrupole RGA (Residual Gas Analyzer)

(14) is available to determine the gas composition of the remaining gases at low pressure,

i.e. 10'7 ~ 10'°I'orr, before starting an experiment. An automatic throttle valve (20) adjusts

the pressure of the process chamber (12) to the value set by the system operator in the

MKS 651C pressure controller (21). A roughing valve (labeled as process on the panel)

(22) is installed between the automatic throttle valve and the mechanic roughing pump.

’9-

    

Three temperature controlled Baratron gauges with capacitance manometer technology:

10 Torr gauge (17), 100 Torr gauge (18), and 1000 Torr gauge (19) are installed to func-

tion at different pressure regimes. There are two isolation valves (16). One is connected

between process chamber and the lO-Torr head and it closes when the pressure is greater

than 9.5 Torr to protect the lO-Torr head. The other one is located between process cham-

ber and the lOO-Torr head and it closes when the pressure is greater than 99.5 Torr to

protect the lOO-Torr head. Either nitrogen (26) or Ar with 99.999% purity (27) is released

through the system vent valve (25) when raising the chamber pressure to the atmospheric

pressure. Nitrogen gas (29) is used to dilute the exhaust gases (30) through a 5000 sccm

gas ﬂow controller (28) with a N2 ﬂow rate equal to 2000 sccm. The mixing of nitrogen

into the exhaust gases ensures that the exhaust gases are not in a ﬂammable condition.

Compared to MSU-MPACVD system I, the vacuum pump and gas ﬂow control

systems of MSU-MPACVD system I] are greatly improved, providing a better high purity

controlled environment for deposition experiments. (System I has a leak rate of 12.6

mTorr/hr, which yields a purity of 99.987%. System H has a leak rate of 4 mTorr/hr, which

yields a purity of 99.995%.) The low-pressure-pump-down capability of system I] ensures

that there is a low leak rate before each experiment. In addition, it removes the impurities

that come off the chamber walls at very low pressures. The use of argon as a vent gas in

system I] also reduces N2 impurities in the system and the low leak rate of system I]

ensures negligible introduction of impurities from the atmosphere. In system H, the source

gases have higher purity to minimize the introduction of impurities, such as nitrogen, oxy-

gen, etc., into the system. The high vacuum capability of system I] lowers the combined

65

N2 impurities from the input gases to 5-8 ppm level, as compared to system I where N2

impurities are 30-50 ppm (Nitrogen—impurity determination was measured by optical

emission spectroscopy technique and was performed by Dr. Mossbrucker.) The inﬂuence

of nitrogen on the growth, morphology, and crystalline quality of MPACVD diamond

ﬁlms has been studied by Asmussen et al. [Asmu 1998]. It was found nitrogen impurities

as low as 20 ppm could affect the morphology and quality of the ﬁlms. With limited impu-

rities in the system I] and the precise gas ﬂow controllers, the operator can (if desired)

carefully control the nitrogen impurities and then can investigate the inﬂuence of nitrogen

impurities on ﬁlm deposition quality and properties.

66

(9) 8-Channel@;l (8) Gas Flow Controllers

MKS Flow Controller

,, o

(6)

(10) Process Valve

(11) Base Plate

‘

    
 

2393
.

(7) He

3.0.01

' . . Q

0.0.0.4
5.0.0;

(13) Ion

 

Gaugeg

(12) Process Chamber

(14) RGA l7

 

(15) Turbo

Pump

 
   

Tl

 

 

(20) Automatic Throttle

l

L

(17) 10 Torr Gauge

(16) Isolation Valves

   

(18) 100 torr Gauge

   

(

)

19 1000 T

G

°” auge

 
  

Valve

(21) Pressure

Controller

(22) RoughingE

Valve

M (25) System Vent Valve

 

 

_

  

(23) Turbo

Valve

(24) Mechanical Roughing

Pump

  
 

  

  

 

 

(28) Gas Flow Controller

  

(30) Exhaust

FIGURE 3.4 : VACUUM PUMP AND THE GAS FLOW CONTROLLER SYSTEM

FOR MSU-MPACVD REACTOR I]

67

3.2.3 Computer Control System

3.2.3.1 MSU-MPACVD System I

A computer is used to monitor and regulate system I operating conditions and to

control the experimental running time and shut down procedures. The preset running time

and system operating pressure along with the reﬂected power readings are the input sig-

nals to the computer. Figure 3.5 shows the ﬂow chart of the computer monitoring program

[Zhan 1993]. As shown, the experimental running time, reﬂected power upper limit, and

the operating pressure threshold are ﬁrst set. The experimental system is then enabled to

allow the feed gases to ﬂow and the microwave power to turn on. After the system is

enabled, the automatic throttle valve operates in a remote mode controlled by the com-

puter control system to adjust the pressure of the chamber. During the experiment, a

checking loop compares the pressure, reﬂected microwave power, and time with the preset

values to determine the state of the experiment. An emergency Shut down of the micro-

wave power and feed gas is performed if the reﬂected microwave power, and/or operating

pressure exceed the preset values, and/or power supply is shut down, and/or cooling water

ﬂow is too low at any time during the experiment. In an emergency, the automatic throttle

valve controls the pressure of the system with the preset value. At normal conditions,

when the experiment is complete (i.e., the timer expires), the computer program directs the

system into a normal shut-down sequence (see Section 3.3.2.1) when the timer expires.

68

 

1. Set experiment running time

2. Set threshold pressure

3. Enable the system

4. Experiment start-up sequence, see Section 3.3.2.]

 

5. Start timer

 

 

 

1. Record the incident and reﬂected power and

operating pressure. Compare these values

wrth preset values.

2. Check timer.

 

 

 

Operating pressure over threshold

Reﬂected power over upper limit

Time expires

Microwave power shut down

 

y

  Water cooling ﬂow too low

Y

 

 

Normal shut down procedure

Emergency shut down procedure

See Section 3.3.2.1

 

 

 

1. Turn off microwave power

2. Turn off all gases ﬂow

3. Close automatic throttle valve

 

 

 

FIGURE 3.5 : COMPUTER CONTROL SYSTEM FOR MSU-MPACVD REACTOR I

69

3.2.3.2 MSU-MPACVD System 11

System I] is also equipped with a computer to monitor and regulate the system

operating conditions and to control the experimental running time and shut down proce-

dure of the experiments. The preset system operating pressure and input microwave power

along with running time are the input signals to the computer. Figure 3.5 shows the ﬂow

chart of the computer monitoring program written by Michael J. Ulczynski. As shown, the

operating pressure, input microwave power, and the experimental running time are ﬁrst set

in the computer monitoring program. The experimental system is then enabled to allow

the feed gases to ﬂow and the microwave power to turn on. After the system is enabled, the

automatic throttle valve operates in a remote mode controlled by the program to adjust the

pressure of the chamber. During the experiment, a checking loop compares the pressure,

incident microwave power, and time with the preset values to determine the state of the

experiment. An emergency shut down of the microwave power and feed gas is performed

if the reﬂected microwave power is more than 25% of incident microwave power value,

and/or operating pressure exceeds i 5 Torr of the preset values at any time during the

experiment. In an emergency, the automatic throttle valve controls the pressure of the sys-

tem with the preset value. It closes if the chamber pressure drops below 10 Torr of preset

value. It opens if there is a leak in the system and the gases are pushed into the system

from atmosphere. At normal conditions, when the experiment is complete (i.e., the timer

expires), the computer program directs the system into a normal shut-down sequence (see

Section 3.3.2.2) when the timer expires.

70

 

1. Set experiment pressure

2. Set input microwave power

3. Set Experiment running time

4. Enable the system

5. Experiment start-up sequence, see Section 3.3.2.2

 
  

6. Start timer as pressure reaches its set point

 

1. Record the incident and reﬂected power and

operating pressure. Compare these values

with preset values.

2. Check timer.

 
  

Operating pressure over/below 5 Torr

Time 6"131“”

 

 

Reﬂected power is over 25% of

incident power (disabled)

 

 y

I

Normal shut down procedure

Emergency shut down procedure

See Section 33.2.2

  
 

1. Turn off microwave power

2. Turn off all gases ﬂow

3. Automatic throttle valve

controlled by pressure controller  
  

FIGURE 3.6 : COMPUTER CONTROL SYSTEM FOR MSU-MPACVD REACTOR II

71

3.2.4 Reactor Geometry

3.2.4.1 Microwave Cavity Plasma Reactor

Figure 3.7 shows a cross sectional drawing of the microwave cavity plasma reactor

conﬁguration. As shown, the cavity applicator side wall (1) is made of a 17.78 cm inside

diameter cylindrical brass tube. This brass tube which forms the conducting shell of the

cavity applicator is electrically shorted to a water—cooled baseplate assembly (2 - 4) and a

water cooled (21) sliding Short (8) via a ﬁnger stock (9). Thus the cylindrical volume

bounded by the sliding short, the cavity applicator side wall and the baseplate forms the

cylindrical cavity applicator electromagnetic excitation region. With a ﬁxed excitation fre-

quency and cavity radius, a cavity length was determined by .1. Zhang et. a1. [Zhan 1994]

to have TM013 mode for exciting a ellipsoid-like discharge (12) in good contact with the

substrate (7). The TM013 mode was used so that the probe was located far enough away

from the discharges to eliminate the near ﬁeld effect caused by coaxial excitation probe

(11). 2.45 GHz (CW) microwave power is coupled into the cylindrical cavity applicator

through a mechanically tunable coaxial excitation probe which is inside a coaxial

waveguide (10) and is located in the center of the sliding short. The sliding short controls

the applicator height, Ls and the excitation probe controls the depth of the coaxial excita-

tion probe, LP. Both L5 and Lp can be moved up and down along the longitudinal axis of

the applicator cavity wall and can be adjusted independently to excite the desired electro-

magnetic mode and optimally match the resonance. The applicator height L8 is adjusted

approximately to 21.7 cm and the probe depth, Lp is about 3.2 cm [Zhan 1993]. The base-

plate assembly consists of a water-cooled (22) and air—cooled (19) baseplate (2), an

annular input gas feed plate (3), and a gas distribution plate (4). A 12.5cm inside diameter

72

quartz dome (5) is sealed by O-ring (20) in contact with baseplate assembly. The thermally

ﬂoating substrate holder setup assembly includes a ﬂow pattern regulator (15), a metal

tube (16), a quartz tube with I.D=95 mm, O.D.=100 mm, and height=50 mm, (17), and a

holder-baseplate (6). The premixed input gases are fed into the gas inlet (23) in the base-

plate assembly. The substrate (7) is placed on top of a molybdenum substrate holder (15),

also called a ﬂow pattern regulator, which is supported by a quartz tube (17). Quartz tubes

of different heights may be used to change the position of the substrate with respect to the

plasma to optimize the ﬁlm deposition. A metal tube (16) which serves as an electromag-

netic ﬁeld resonance breaker is placed inside the quartz tube. The metal tube prevents the

plasma discharge from forming underneath the substrate by reducing the electric ﬁeld

underneath the substrate. The metal tube (16) and quartz tube (17) are placed on a sub-

strate holder-baseplate (6) which has 3 cm diameter hole at its center to pass the hot gases

from within the quartz dome (5) to the exhaust roughing pump. The baseplate, the annular

input gas feed plate, and the gas distribution plate introduce an uniform ring of input gases

into the quartz dome where the electromagnetic ﬁelds produce a microwave discharge.

The plasma consists of a mixture of neutral gases, electrons, and ions, i.e. dissociated spe-

cies. A screened view window (13) is cut into the cavity wall for viewing the discharge.

By focusing an optical pyrometer onto the substrate through the view window, the sub-

strate temperature can be determined. A air blower with 60 CFM (cubic foot per minute)

blows the cooling air stream into the air blower.inlet (14), onto the quartz dome (5) and

cavity side wall (1), and then ﬁnally ﬂows out of the cavity through the screened window

(13) and optical access ports(l9) in the baseplate (2). The air blower existing inside the

Cober microwave power supply adds another air cooling stream into the microwave cavity

73

r.-.

.4

LI

.

_

1

it.

1' r I

i- ..

.

T ..

._

rl 0’

I

3.. .C

\

3.!»

Q I, 1.1

mi.-

5;; .n

.

p

a

1r, ,1

It! i

n.

.r

{I

r

.p...

s

L.

n

it

A

v

t

I...

l

v

(I

o

t

I

.0

.

I

’t

.

I

.4

o....l...v

.

.

Q

_

Id

of .

 

 

plasma reactor. Two Teﬂon pieces (18) were drilled with four of 1/8” diameter holes. This

allows the cooling air from the air blower in the microwave power supply to ﬂow through

the coaxial waveguide, onto quartz dome and cavity side walls, and then ﬂows out of the

air blower outlet and the optical access ports. On the system I, one fan is used to cool the

cavity applicator Side wall. On system 1], two fans are used to cool the cavity applicator

side wall and the air cooling provided by the fans also extends the lifetime of the quartz

dome. The microwave cavity plasma reactor is mounted on a process chamber with the

chamber outlet leading to vacuum pumps. On system I], water cooling: (21) and (22) is

replaced by a chiller which controls the temperature of the input coolant liquid. The cool-

ant temperature can be set by the system operator.

The thermally ﬂoating substrate holder setup utilizes a ﬂow pattern regulator for

the molybdenum substrate holder (I.D.=3.03125” and O.D.=4.03125”) as shown in Fig.

3.9. The ﬂow pattern regulator is a plate with a series of holes arranged in a circle right

inside the big circumference. The gas ﬂow coming out the gas inlet, into the quartz dome,

is directed by the ﬂow pattern regulator as it ﬂows through the plasma. The conﬁguration

is designed to increase the uniformity of the ﬁlm deposition by changing the ﬂow pattern

in the plasma discharge and inﬂuencing the shape of the plasma discharge [Zhan 1993].

The top-side substrate temperature measurement was carried out by focusing the

pyrometer onto the substrate through the screen view window for system 1. Strong argon

emissions interfere with top-side substrate temperature measurements using the pyrometer

positioned to look through the plasma discharge. A silicon wafer back-side substrate tem-

perature measurement was made in system H by aiming the pyrometer at the back side of

74

the wafer image that was reﬂected off the mirror for system I]. In system I], the

holder-baseplate has a 1” diameter hole cut in the center to allow the optical access to the

backside of the wafer. The pyrometer defects 1 pm (near IR) wavelength light. Since the

mirror is not a perfect reﬂector, the bottom temperature measurement may be somewhat

low. Also, there is a slight reﬂection loss at the window. However, these effects probably

not signiﬁcant because blackbody radiation is proportional to T4. Figure 3.8 displays how

the substrate temperature was measured for systems I and II.

75

18

 

”1| ‘

. l .

2

:°

     

'<E

z“

8
0

"a“
10». U

  
  

to'

g

\

\

B

§

 

 

21

.

l

 

e e

,

  

5

l

8

11—

l

Lp

L

9%

1_9

   

 

 

.

C 1

5

V

13

  

Ls

VCooling Air \  

.. Aﬂomg

 

      

;__.___I_l

  

 

O'O’O'O'C
9'. o o o 0

“5.0.0.0.... 0..

   

  

4

Legend

(1)

Cavity Side Wall

(2)

Baseplate

(3)

Annular Plate

(4)

Distribution Plate

(5)

Quartz Dome

(6)

Holder-baseplate

(7)

Substrate

(8)

Sliding Short

(9)

Finger Stock

(10)

Coaxial Waveguide

(11) Excitation Probe

(12) Plasma discharge

(13) View VVrndow (grid)

(13) Air Blower Outlet

(14) Air Blower Inlet

(15) Flow Pattern Regulator

(15) Substrate Holder

(16) Metal Tube

(17)

Quartz Tube

(18) Teﬂon Pieces

(19) Optical Access Port

(20)

Seal O-ring

(2]) Cooling Sliding Short (22) Cooling Baseplate

(23) Gas Inlet

(24) Gas Outlet

FIGURE 3.7 : THE CROSS SECTIONAL VIEW OF THE MCPR

 

 

 

76

 

  

ljﬂ I

q

x

i

N

N

 
 
  

9.;

 

LP

  

 

.

Arr Blower —> 1:

Substrate

‘

x-’ ,

--__‘IJ

 

:

 

 

i]

  
. a. on

, 4,: I ’

."::_::-

.9 . 9-L-.- ,1
    

 

.... .1109...

ﬂ

I L—‘JI

ll

:

|

Mirrok----------

(b)

Window

  
  

FIGURE 3.8 : SUBSTRATE TEMPERATURE MEASUREMENT METHOD

(a) Top Side substrate temperature measurement through view window for system I.

(b) Back side substrate temperature measurement by the mirror reﬂection for system H.

77

 

Top view of the substrate holder (ﬂow pattern regulator) [Zhan 1993].

  

Substrate

   

  Gas Inlet

    

 

ll

ll

 

     

,

,

Holder Baseplate

 

W V

Quartz Tube

Metal Tube

‘1

l
a
i

FIGURE 3.9 : THERMALLY FLOATING SUBSTRATE HOLDER SET UP

Optimized substrate holder design that achieves deposition uniformity.

 

78

  

 

3.2.4.2 Substrate Heating Stage

Unlike conventional H2/CH4 CVD diamond growth, ultra-nanocrystalline dia-

mond is not grown in a hydrogen rich environment. A noble gas, Ar is the dominant input

gas. Argon changes the plasma chemistry, including the gas temperature and thus has an

impact on substrate temperature. Gas temperatures and the power densities are lower in

the “hydrogen poor” carbon containing argon plasmas than in the more conventional

H2/CH4 discharge. In order to increase the deposition rate and provide independent con-

trol of the substrate temperatures, a substrate heating stage may be desirable. Figure 3.10

shows the cross sectional view of the substrate heating stage that was developed for

ultra-nanocrystalline diamond ﬁlm deposition.

Substrate

Gaslnlet

 

 

 

O!

 

 
 
 

 

  

Y........._JU

 

'

'

 

Quartz Dome

Plasma

 

 

 

. r ’ 79" ‘

r

“%;~ .
ill-1'” 4‘

-..'

A

'

.

i

1

Substrate Holder

Baseplate

l

 

,_l________________

 

 

 
  

 

 

   

 
 
    

    

'1‘r—’ :r'

'1: ;_ 1T-

’7

 

m

1d B

l

H0 er

3391’ ate

 

V Thermal l

V

Couples Heater Support

Quartz Tube

 

I-Ieater

FIGURE 3.10 : SUBSTRATE HEATING STAGE SET UP

Substrate heater stage design for independent temperature control.

79

..v

_

.._

Hv.vc; l.

u .‘.ur

,

.trl

OI..

"

[’2' an

.I. A

.

4!
0.:

.

.

Ir

_

.7.

I 'lo

'2- ..Ir

 
 

Figure 3.11 illuminates the construction of the prototype substrate heating stage.

The substrate heating stage consists of a can body (1), which is machined out of a solid

piece of stainless steel, and a bottom plate (2). These two parts were welded together to

prevent the plasma discharge from forming beneath the substrate. A Boralectric PBN/PG

(Pyrolytic Boron Nitride/Pyrolytic Graphite) type HT-63, diagonal pattern heater (3) is

placed inside the top part of the can body as shown in Fig. 3.11. A pair of ceramic hats

served as insulators (5) to prevent the electrical connectors (4) from being electrically

shorted by the can body. The electrical connector consists of a molybdenum threaded rod,

tantalum nuts, tantalum washers, and a graphite washer for good electrical contact in the

hazardous environment. A pair of molybdenum wires are used as the heater wire (6).

These wires are connected via vacuum feedthrough to a DC power supply located outside

the vacuum chamber. A heater support (7) and 4 height adjusters (8) are designed to adjust

the heater longitudinal position. Two K-type thermal couples (9), insulated by ceramic

beads, are used to measure the temperature. One is located at the center of the heater. The

other one is located at the edge.

80

'-
l1
I.

3.340”
3....”
..

4
j
j

  

  

  

5

.\

3

4

/._—T

 

1

  

 

A

 

  

 

     
 
   

ii

ii

n

WI i

 

l

I

I

n

u

n

I

7

Legend

(1)

(4)

(7)

Can Body

(2)

Bottom Plate

(3)

Heater

Electrical Connector

(5)

Insulator

(6)

Heater Wire

Heater Support

(8)

Height Adjuster

(9)

Thermal Couples

FIGURE 3.11 : THE SUBSTRATE HEATING STAGE

81

3.2.5 Microwave Cavity Plasma Reactor General Operating Field Map

for Ultra-nanocrystalline Diamond Synthesis

The experimental measurements of the

general

operating ﬁeld map for

ultra-nanocrystalline diamond synthesis was canied out with the microwave cavity plasma

reactor operating under the thermally ﬂoating substrate holder set up that was utilized

throughout this thesis. Under the thermally ﬂoating substrate holder set up, the deposition

pressure, p, the microwave power, Pabs, and the substrate temperature, T8, are interrelated

and interdependent. Figure 3.12 displays the repeatable experimental non-linear relation-

ships between these variables. As shown, with ﬁxed Ar/H2/CH4 gas chemistry, the

substrate temperature, Ts, and the discharge volume, Vd, are affected both by the deposi-

tion pressure, p, and the microwave power, Pabs. The substrate temperature, TS, increases

with either increases in the deposition pressure, p, and/or the microwave power, Pabs. For a

ﬁxed deposition pressure, p, the discharge volume, Vd, increases with increases in the

microwave power, Pabs. For a ﬁxed microwave power, Pabs, the discharge volume, Vd,

decreases with increases in the deposition pressure, p. By the observation through the view

window in the cavity wall, the lower absorbed microwave power limit is determined by the

minimum power required to maintained a discharge volume (Vd)min that covers a 3” diam-

eter substrate. Below (Vd)min’ the discharge volume is too small to fully cover the 3”

substrate surface, and also in higher deposition pressure region, an orange color starts to

appear in the outer edge of the plasma, which is usually undesirable. Specially, with too

small a plasma and above 220 Torr, the discharge becomes ﬂashing. The upper limit of the

micrOwave power is determined by the maximum power that can be used to generate a dis-

Charge volume (Vd)max that is not too big to touch the quartz dome walls. The upper limit

0f the microwave power is the maximum microwave power that can be used to operate the

82

reactor safely for ultra-nanocrystalline diamond deposition without over heating the quartz

dome. It is very critical for the higher pressure region. The measurements were performed

using system H.

 

      

120 Torr

/

i

-

 

800 '

750 '

Q 700 -

PM

650 r

 

600—

./0/

 

700

800

900

1000 1100 1200 1300 1400 1500 1600 1700

Pabs (W)

FIGURE 3.12 : THE MPACVD OPERATING FIELD MAP UNDER

THERMALLY FLOATING SUBSTRATE HOLDER SET UP

For 5” quartz dome and 3” substrate reactor geometry conﬁguration.

The gas chemistry is Ar/I-12/CH4=100/4/1 sccm.

83

The inﬂuence of the total gas ﬂow rate has also been studied. The results indicate

that the total gas ﬂow rates have no signiﬁcant inﬂuence in P-PabS-Ts relationship, as

shown in Figure 3.13(a) and (b). In Figure 3.13(a), the substrate temperature, Ts increases

as the pressure increases. The corresponding microwave power, Pabs is shown in Figure

3.13(b). The experimental results displayed in Figure 3.13 shows that the ﬁeld map curves

of Figure 3.12 vary as the gas chemistry changes. Thus, one needs a separate operating

ﬁeld map for each gas chemistry ratio.

84

 

740p

720

l

700 r

680

I

660 -

640

I

)
C
°
(

S
T

620 -  600 r 
 

 

58060

80

1 00

120

140

1 60

1 80

200

220

240

(a)

Pressure (Torr)

 

140° 4- Ar/H2/CH4=100/2/1 sccm

‘

'

'

‘

+ Ar/H2/CH4=300/6/3 sccrn

1350

a

Ar/H2/CH4=100/4/1sccm

1300 + Ar/H2/CH4=600/24/6 sccm

 

 

7
/K/ll

‘

)
W
(

s
b
a
P

1 250

1 200

l

1150

I

-.

1100

1 050 

1

l

I

l

I

1

I

1

I

l

l

4

 

 

00060

80

1 00

120

140

1 60

1 80

200

220

240

(b)

Pressure (Torr)

FIGURE 3.13 : P-PabS-TS RELATIONSHIP UNDER THERMALLY FLOATING

SUBSTRATE HOLDER SET UP

For 5” quartz dome and 3” substrate reactor geometry conﬁguration.

The gas chemistry ratios are ArzH2:CH4=100:4:1 and 100:2:1.

 

85

w—'-

lll

3.3

Experimental Procedures

3.3.1 Seeding Procedures

Without wafer pre-treatment, CVD ultra-nanocrystalline diamond ﬁlms could not

be synthesized during 8-hour experimental runs. Therefore, a nucleation enhancement

step was required for the deposition of CVD ultra—nanocrystalline diamond from gas

phase onto non-diamond substrates, like mirror-polished silicon wafers. Two substrate

seeding methods were investigated in this thesis. The comparison of photoresist seeding

and scratch seeding is presented in Section 3.3.1.3.

3.3.1.1

Photoresist Seeding

The process for the photoresist seeding is as followed.

86

1) Put 310mg of Amplex 0.1 um sized natural diamond powder into a clean brown bottle.

2) Dehydrate diamond powder on a hot plate at 150°C for 2 hours with bottle open partly

covered with aluminum foil to let the moisture vapor out.

3) Add 12 ml of type P Shipley photoresist thinner into the bottle.

4) Put a magnetic stirrer in the bottle.

5) Place the bottle on a stirring plate with setting at 4.5 for lSminutes.

6) Place the bottle in an ultrasonic bath for one hour.

7) Add 30 ml of Shipley 1813 photoresist into the bottle.

8) Stir the solution on a stirring plate with the setting at 4.5 for one hour.

9) Place the bottle in an ultrasonic bath for two hours.

10) Drop the diamond powder/photoresist solution on the substrate surface with full cover-

age. Spin the substrate at 2500 rpm for 30 seconds.

11) Bake the seeded substrate in an oven at 150°C for 15 minutes.

12) Check the substrate seeding result under the optical microscopy. If there are big dia-

mond chunks, repeat step 8 and step 9 to break the big diamond chunks in the solution.

Then continue to perform step 10 through 12.

87

Figure 3.14 shows the result of a photoresist seeding followed by the above steps.

The picture was taken by a digital camera under an optical microscope. In the image, no

chunks of diamond are found.

 

FIGURE 3.14 : PHOIORESIST SEEDING

Image size: 13 um x10 um.

88

3.3.1.2 Scratch Seeding

The procedure for mechanical scratch seeding is as followed.

1) Place the substrate on the seeding stage.

2) Connect the seeding stage to a pump and turn on the power. The vacuum sucks the

substrate and keeps the substrate from moving.

3) Sprinkle some quantity of Amplex 0.1 pm sized natural diamond powder onto the sur-

face of the substrate. If the humidity in the room is high, bake the diamond powder at

150°C for 2 hours before usage.

4) Use a ﬁnger wrapped in a Kim WipeTM to polish the substrate with the combination of

several different angles of straight line motion and several different diameters of circu-

lar motion. Make sure the substrate surface is scratched everywhere with a median

force.

5) Put a piece of curved glass inside a container. Pick up the substrate from the seeding

stage and put it in the container, on top of the curved glass, with the scratched surface

facing down.

6) Fill the container with methanol to the liquid line above the substrate about 1 cm.

7) Put the container in an ultrasonic bath for 30 minutes for cleaning and agitation pur-

pose.

8) Take the substrate out and put it in another container with the scratched surface facing

up. Fill the container with enough acetone to cover the substrate.

9) Use Q-tip gently wipe the substrate surface to remove any dirt or diamond powder.

10) Put the substrate on a wafer holder.

11) Rinse the substrate with acetone and methanol for 2 minutes each step.

12) Rinse the substrate with de-ionized water for 15 minutes.

13) Blow dry with a clean room nitrogen gun.

14) Check the substrate surface cleanness under optical microscopy. If there is dirt or dia-

mond powder left on the substrate surface, repeat step 8 to step 14.

89

Figure 3.15 shows the result of a mechanical scratch seeding followed by the

above steps. The picture was taken by a digital camera under an optical microscope. In the

image, there is no dirt or diamond powder residue but the different width and depth lines

left on the scratched silicon wafer are clearly shown. The different width and depth of

scratching lines were caused by different sized diamond particles and the applied scratch-

ing forces.

 

FIGURE 3.15 : SCRATCH SEEDING

Image size: 13 um x10 um.

90

3.3.1.3 Comparison of Photoresist Seeding and Scratch Seeding

Two set of experiments were performed to compare the growth results of photore-

sist seeding and mechanical scratch seeding in this thesis research. The experiments were

all done at 120 Torr with argon gas ﬂow of 100 sccm, hydrogen gas ﬂow of 4 seem, and

methane gas ﬂow of 1 sccm, but the deposition times were varied.

Figures 3.16 (a), (b), and (c) Show the deposition results for photoresist seeding

after 10 minutes, l-hour, and 8-hour runs respectively. Under these conditions, the photo-

resist seeding usually (1) produced a non continuous ﬁlm, (2) larger crystals were

nucleated, (3) seeding results were more uniform and repeatable.

91

(z!)

Digital Instruments NanoScope

Scan

1'

5. 000

Scan rate

1. 001 Hz

Number of samples

256

Image Data

Height

Data scale

600.0 nm

arn077.001

(0)

x

z

1 000 um/div

600 000 nm/div

Digital Instruments Nanoscope

Scan size

Scan rat 2

5.000 um

1. 001 Hz

Nuuber or sauples

258

Image Data

Height

Data scale

1.000 un

  

X

1.000unldiu

2 1000.mnu/div

arol3$.000

FIGURE 3.16 : PHOI'ORESIST SEEDING FILM MORPHOLOGY AND ROUGHNESS

92

INSPECTED BY AFM

   

(C)

Digital Instruments NanoScooe

Scan size

Scan rate

5.000 pm

1.00] Hz

Number of samples

256

Image Data

Height

Data scale

400.0 nm

arn068.000

x

Z

1 000 pm/div

400 000 nm/div

(a)

t=10 minutes

Tst°P=700°C

rsback=659°c

(b)

t=1 hrs

(e)

t=8 hrs

Tst°p=700°C

Tst°P=660°C

stack= 638 0C

stack= 692 0C

Pabs=977W

Pabs=880W

Pabs=977W

Ave. Growth Rate

Ave. Growth Rate

Ave. Growth Rate

=0 11 m/hr

=0.239 u thr

=0.488 u m/hr

Img. RMS=112.78 mm

Img. RMS=134.86 nm

Img. RMS=81.748 nm

FIGURE 3.16 (cont’d)

 

93

Figure 3.17 (a), (b), and (c) Show the deposition results for scratching seeding after

10 minutes, l-hour, and 8-hour runs respectively. The scratch seeding method (1) pro-

duced a continuous ﬁlm, (2) smaller crystals were nucleated, (3) higher nucleation density

occurred due to higher seeding density, (4) some big crystals developed at deep scratches,

(5) lower RMS surface roughness was achieved.

94

(a)

Digital Instruments Hanoscope

Scan srze

Scan rate

5.000 um

1.00] Hz

Number of samples

256

Image Data

Height

Data scale

400.0 nm

arn082.000

(b)

X

2

1.000 um/div

400.000 nm/div

Digital Instruments Hanoscope

Scan size

Scan rate

5.000 um

1.001 H2

Number of samples

58

Image Data

Height

Data scale

500.0 nm

arn143.000

X

2

1.000

um/diu

500.000 nm/div

FIGURE 3.17 : SCRATCH SEEDING FILM MORPHOLOGY AND ROUGI-INESS

INSPECTED BY AFM

 

 

95

(C)

Digital Instruments NanoScope

Scan size

Scan rate

S

.0

00 um

1.001 H2

Number of samples

256

Image Data

Height

Data scale

300 0 nm

arn076.000

X

2

1.000 um/div

300.000 nm/div

(a)

t=10 minutes

(b)

t=l hrs

(C)

t=8 hrs

Tst°P=720°C

Ts‘°p=715°C

Tst°p=695°C

stack=6640C

T3”: 630 °C

stack= 666 0C

Pabs=898W

Pabs=886W

Pabs=981W

Ave. Growth Rate

Ave. Growth Rate

Ave. Growth Rate

=0 p. m/hr

=0.295 um/hr

=O.347 u m/hr

Img. RMS=17.827 nm

Img. RMS=13.086 nm

Img. RMS=23.860 nm

FIGURE 3.17 (cont’d)

 

96

When there is a higher nucleation density, competition between neighboring grains

prevents individual grains from growing at the speed of ﬁlm growth rate. 80, at the same

thickness, 3 higher nucleation rate results in smoother ﬁlms [Wild 1991]. Therefore, for

thin ﬁlms, the inﬂuence of seeding is very important. It may not be as important for thicker

ﬁlms. From Fig. 3.16 and Fig. 3.17, the results clearly show the seeding procedure plays

an important role on thin ﬁlm growth and its inﬂuence on thick ﬁlms is less observable.

Because the mechanical scratch seeding (R.M.S. surface roughness ranges from 13

to 24 nm) produces better ﬁlm quality than photoresist seeding (R.M.S. surface roughness

ranges from 82 to 135 nm), the scratch seeding method was chosen to be utilized through

out this thesis.

3.3.2 Start-up and Shut-down Procedures

3.3.2.1 System I

After loading the sample, the system is pumped down to approximately lOmTorr.

Now, one can start the experiment.

0

Turn on the water to microwave power supply.

0

Turn on the microwave power supply. The power level control knob should be zero.

0

Turn on the nitrogen purge to the pump.

0

Turn on the water for MPCR water cooling system.

97

Set the experimental running time and threshold pressure in the plasma software. Set

the gas ﬂow rates on the MKS type 247C ﬂow controller.

Set the experimental pressure on the pressure controller of the throttle valve.

Adjust the cavity length to 21.7 cm by moving the sliding short position and set the

probe depth at Lp=3.2 cm.

Set the throttle valve at auto position and enable the system to let the input gases come

in.

Enable the microwave power supply when the system pressure reaches 5~10 Torr.

Turn on the cooling fans.

Slowly increase the input microwave power as pressure increases so that the ellip-

soid-like discharge covers the entire substrate surface.

Fine tune the cavity length, Ls and Lp to obtain the maximum absorption power.

The experiment starts to run on its own under time control when the system pressure

reaches the valued set in the plasma software.

When the experiment is completed, the system will perform the automatic shut

down procedure as follows.

Shut off the H2 and CH4 mass ﬂow controllers.

3 minutes self-cleaning process with argon plasma.

98

Turn off microwave power.

Turn off the water to microwave power supply when the system is cooled.

Turn off the argon mass ﬂow controller.

The automatic throttle valve will open to evacuate the chamber.

3.3.2.2 System 11

After loading the sample, the system is pumped down to less than mTorr range.

Close the roughing valve and open the turbo isolation valve. Turn on the turbo molecular

pump. Let the system pump for a few hours until the base pressure is about 1~2x10'6 Torr.

When the system reaches high vacuum status, turn off the turbo molecular pump and close

the turbo isolation valve. Now, the experimental run is initiated by the follow steps.

Turn on the water to microwave power supply.

0

Turn on the microwave power supply. The power level control knob should be zero.

0

Turn on the nitrogen purge to the pump.

0

Turn on the Neslab chiller and set the temperature at 15°C.

0

Set the experimental running time, input microwave power, pressure, and gas ﬂows

(During the start-up procedure, set the gas ﬂows six times higher than the planned gas

ﬂows. Keep the same gas chemistry ratio and remember to set the value back to the

desired gas ﬂows.) for each channel in the plasma software.

0

Set the experimental pressure on the pressure controller of the throttle valve.

99

Adjust the cavity length to 21.7 cm by moving the sliding short position and set the

probe depth at Lp=3.2 cm.

Open the process valve and turn on the mass ﬂow controllers to let the input gases ﬂow

in. (Optional, ﬂushing the chamber with Argon gas. So far, there is no quantitatively

evidence that it changes the ultra—nanocrystalline diamond deposition result.)

Open the roughing valve. Now the chamber pressure is controlled by the automatic

throttle valve.

Enable the microwave power supply when the system pressure reaches 10 Torr.

Turn on the cooling fans.

Slowly increase the input microwave power as pressure increases so that the ellip-

soid-like discharge covers the entire substrate surface.

Fine tune the cavity length, LS and LI) to obtain the minimum reﬂected power.

The experiment starts to run on its own under time control when the system pressure

reaches the valued set in the pressure controller.

When the experiment is completed, the system will perform the shut down proce-

dure as followed.

No self-cleaning process.

Turn off microwave power.

Turn off the water to microwave power supply when the system is cooled.

100

 

0

Turn off all the mass ﬂow controllers.

-

The automatic throttle valve will close at 10 torr below set point. If there is a leak in

the system and the atmospheric air is forced into the system, the throttle will reopen at

the set point pressure to let the system pressure remain at the set point, below atmo—

sphere.

3.4

Experimental Parameter Space

The performance of the microwave plasma reactor is closely related to the input,

internal, and output variables (Fig. 3.18). The input variables include the controllable

input variables (U1), reactor design variables (U2), and deposition process variables (U 3).

As the input parameters (U) vary, internal variables (X) also change. The output variables

which can be subdivided into reactor performance variables (Y1) and ﬁlm characteristics

variables (Y2) are the outcome of input and internal variables.

101

)
1
Y
(

e
c
n
a
m
r
o
f
r
e
P

r
o
t
c
a
e
R

)
1
U
(

s
e
l
b
a
i
r
a
V

t
u
p
n
I

e
l
b
a
l
l
o
r
t
n
o
C

 

V

)
2
Y
(

s
m
s
'
i
‘
m
f
i
r
g
a
1
‘
3
“

.

.

.

h
t
:
)
2
1
1
2
1
3
2
(

$
1
6
3

s
s
e
i
i
h
g
u
o
R

m
l
i
F

)
c

y
t
r
e
p
o
r
P

l
a
c
i
t
p
O

)
f

e
r
u
t
x
e
T

m
l
i
F

)
1
(

e
z
i
S

l
a
t
s
y
r
C

)
6

)
m
u

-
-
-
-
-
-

.
z
u
.
1
U
(
=
2
U

)
“
U
.
.
.
.
.
.
,
2
u
,
1
u
(

:
3
U

)
"
X
.
.
.
.
.
,
2
x
,
1
x
(

=
X

)
n
Y
-
w
2
Y

.
r
y
(

=
Y

)
l
u
.
.
.
.
.
.
,
z
u
h
u
(

:
]
U

e
r
e
h
w

2
”

C
I
R
E
N
E
G
A
N
I

S
R
E
T
E
M
A
R
A
P
T
U
P
T
U
O
D
N
A
,
L
A
N
R
E
T
N
I

.
T
U
P
N
I
N
E
E
W
T
E
B
P
I
H
S
N
O
I
T
A
L
E
R

:

8
1
.
3
E
R
U
G
I
F

.
M
E
T
S
Y
S
-
D
V
C
P
M

d
l
e
i
Y

3 0

c
ﬁ
i
c
e
p
S

)

.
.
.
m
m

y
c
n
e
i
c
ﬁ
f
e

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

)
1
(

‘

 

 a

t
u
p
t
u
O

e
r
u
t
a
r
e
p
m
e
T

s
a
G

,

:
)
X
(

s
e
l
b
a
i
r
a
V

l
a
n
r
e
t
n
I

e
m
u
l
o
V

e
g
r
a
h
c
s
i
D

y
t
i
s
n
e
D

r
e
w
o
P

e
g
r
a
h
c
s
i
D

e
r
u
t
a
r
e
p
m
e
T

e
t
a
r
t
s
b
u
S

s
e
l
b
a
i
r
a
V

e
r
u
t
a
r
e
p
m
e
T

n
o
r
t
c
e
l
E

)
Y
(

e
m
i
T

e
c
n
e
d
i
s
e
R

s
n
o
i
t
a
r
t
n
e
c
n
o
C

s
e
i
c
e
p
S

s
n
o
i
t
u
b
i
r
t
s
i
D

   

s
e
i
c
e
p
S

l
a
i
d
a
R

.
C
I
C

A

)
x
,
3
U

,
2
U
,
]
m
g

=
Y

)
3
U

,
2
U

,
1
U
(
f

=
X

 

 

_

)
2
U
(

s
e
l
b
a
i
r
a
V

n
g
i
s
e
D

r
o
t
c
a
e
R

p
u
-
t
e
S

r
e
d
l
o
H

e
t
a
r
t
s
b
u
S

)
d

l
a
i
r
e
t
a
M

l
l
a
W

r
o
t
c
a
e
R

)
C

e
c
r
u
o
S

r
e
w
o
P
f
o

e
p
y
T

)
a

n
o
i
t
a
r
u
g
ﬁ
n
o
C

r
o
t
c
a
e
R

)
b

s
m
e
t
s
y
S

m
u
u
é
a
V

)
e

s
e
s
a
G
f
0

y
m
u
P
9

e
r
u
s
s
e
r
P

n
o
i
t
i
s
o
p
e
D

)
c

”
‘
i

r
e
w
o
P

t
u
p
n
I

)
1
(

y
t
i
r
u
p
m
I

2
N

)
e

l
—

)
3
U
(

s
e
l
b
a
i
r
a
V

s
s
e
c
o
r
P

n
o
i
t
i
s
o
p
e
D

 

s
e
r
u
d
e
c
o
r
P

n
w
o
d
-
t
u
h
S

d
n
a

p
u
-
t
r
a
t
S

)
d

e
z
i
S

d
n
a

l
a
i
r
e
t
a
M

e
t
a
r
t
s
b
u
S

e
m
i
T

n
o
i
t
i
s
o
p
e
D

)
a

)
b

e
r
u
d
e
c
o
r
P

g
n
i
d
e
e
S

)
c

102

Because of the complex relations between the large number of experimental vari-

ables, some of the variables are ﬁxed to limit the experimental parameter space. Various

operating conditions including pressure, gas chemistry, temperature, powers, and ﬂow rate

will be investigated and mapped to the attributes of ﬁlm quality, ﬁlm uniformity, and ﬁlm

growth rate (determined by weight gain). The relationships between macroscopically con—

trollable input variables (U1), reactor design variables (U2), deposition process variables

(U3), vs. internal variables (X), and output variables (Y) are investigated in this thesis.

Table 3.1 and 3.2 specify the experimental variables chosen for systems I and II in this the-

sis research.

103

TABLE 3.1: PARAMETER SPACE FOR EXPERIMENTS PERFORMED AT SYSTEM I

 

Controllable

a) Gas Chemistry, Ar/HZ/CH4=99-91/ 1-7/0-2 (%)

1111’“t

b) Total Gas Flow Rate, Ar=102.5-642 sccm

Variables

c) Deposition Pressure, p=120-160 Torr

 

U1

d) Absorbed Microwave Power, Pabs=691-1221W

a) 2.45 GHz Microwave Power, ﬁxed

b) Reactor Conﬁguration, MPCR, ﬁxed

c) End Feed Excitation, ﬁxed

d) Electromagnetic Mode and Cavity Tuning, Fixed

at TMon

Input

Reactor Design

e) Reactor Wall Material, quartz dome, ﬁxed, dia.=5”

Variables

Variables

f) Substrate Holder Set-up, molybdenum holder,

U

U2

thermally ﬂoating stage

g) Vacuum System, mechanical pump, l~2x10'2 Torr,

leak rate: 12.6 mTorr/hr

h) Purity of gases, ﬁxed, See Section 3.2.2.1

Process

 

 

Deposition

a) Deposition Time, 2-48 hours

  
 

  d) Start-up and Shut-down, ﬁxed

b) Substrate Material and Size, 3” Si (100) wafers

Variables

c) Seeding Procedure, mechanically scratched by

U3

0.1 pm natural diamond powder

104

TABLE 3.2: PARAMETER SPACE FOR EXPERIMENTS PERFORMED AT SYSTEM II

 

a) Gas Chemistry, Ar/li2/CH4=99-9O/0-9/1 (%)

Controllable

b) Argon Flow Rate, Ar=lOl-63O sccm

Input

c) Deposition Pressure, p=60-240 Torr

Variables

d) Absorbed Microwave Power, Pabs=732-1518W

U1

e) N2 impurity, 5-2500ppm

 

a) 2.45 GHz Microwave Power, ﬁxed

b) Reactor Conﬁguration, MPCR, ﬁxed

c) End Feed Excitation, ﬁxed

d) Electromagnetic Mode and Cavity Tuning, Fixed

Input

Reactor Design

at mm 3

Variables

Variables

e) Reactor Wall Material, quartz dome, ﬁxed, dia.=5”

U

U2

f) Substrate Holder Set-up, molybdenum holder,

thermally ﬂoating stage

g) Vacuum System, turbo pump, 10‘7~10‘6 Torr, leak

rate*: 4mTorr/hr

h) Purity of gases, ﬁxed, See Section 3.2.2.2

Deposition

a) Deposition Time, 8-25 hours

 
  

b) Substrate Material and Size, 3” Si (100) wafers

c) Seeding Procedure, mechanically scratched by

U3

0.1 pm natural diamond powder

Process

Variables

 

(1) Start-up and Shut-down, ﬁxed

  

 

* Measurement of leak rate was done by Brian Wright.

105

3.5 Measurement Methodologies

3.5.1 Reactor Performance

3.5.1.1

Total Growth Rate

The total growth rate of the ﬁlm, GT (mg/h) is determined by the weight gain, W

(mg) of the substrate during the deposition process divided by the deposition time, t

W
(hour). GT = T

3.5.1.2 Average Growth Rate

The average growth rate, Gawe (um/h) is deﬁned as the total growth rate divided by

the deposition area A (A=45.60 cm2) and the diamond density (Dz3.515 g/cm3) with the

assumption that the whole wafer is covered with ultra-nanocrystalline diamond uniformly.

This method includes a certain inaccuracy caused by the scale. Also, the CVD

ultra-nanocrystalline diamond may not have the same density as natural diamond type 11

because of the numerous amount of grain boundaries. Gave = A—D

GT

3.5.1.3 Speciﬁc Yield

The speciﬁc yield, S (kw - h /g) is given by the microwave power, Pabs (kw) versus

the ﬁlm total growth rate in the units of “g/h” to describe the energy efﬁciency of the dia-

mond deposition process. S =  

GT x

g
3

 

10 mg

3.5.1.4 Carbon Conversion Efﬁciency

The carbon conversion efﬁciency, C (%) is the percentage of the total carbon atoms

provided by the input gases that are actually converted into the diamond ﬁlm on the

substrate.

106

For example, the number of carbon atoms which are converted into the diamond

ﬁlm for a total growth rate of 1 mg/h is:

mg

g

1h

mole

6.02 x 10

atom: 8, 316 X10l7atoms

    

h

3

x 60min x 12g x

mole

mrn

23

mg

For a ﬂow rate of 1 sccm methane, the total number of carbon atoms provided is:

cm3

L

mole

6.02 x 1023atom

l9atoms

l

. x

x

x

 

   

= 2, 6875 x 10

.

mrn

103cm3

22.4L

mole

mrn

In this case, the carbon conversion efﬁciency would be:

8.316x1017
X 100% = 3.09% 

2.6875 x 10

19

So, for a total growth rate of GT and fM sccm of methane, the carbon conversion

efﬁciency is:

GTx 8.316 x 1017

- C =

 

19

x 2.6875 x10

fM

x100%

3.5.2 Film Characteristics

3.5.2.1 Film Texture

The direction of the largest dimension (the direction of the most rapid growth) of

crystals often determines the texture of the ﬁlm. Usually the texture is determined by

means of x-ray diffraction crystallography. With the information of the structure of crys-

talline phases from x-ray diffraction, x-ray diffraction crystallography was used to identify

the presence of diamond. At the same time, x-ray diffraction crystallography becomes a

107

substitute tool for Raman spectroscopy. A two—theta scan of a ﬁlm provides an indication

of growth orientation/texture in the crystalline ﬁlms by comparison the intensity of the

peaks. However, a quantitative determination of the ﬁlm texture can be better pursued by

pole ﬁgures and inverse pole ﬁgures, which were not used in this thesis research. The

x-ray diffractometer at Chemistry Department of MSU is a Rigaku 200 rotating anode

powder diffractometer with maximum power: 60 kV and maximum current: 200 mA.

3.5.2.2 Film Roughness and Morphology

The roughness and surface proﬁle of the ﬁlms were measured by AFM (Atomic

Force Microscopy) with the technique of tapping mode. TappingMode AFM has the

advantage of higher lateral resolution and virtually eliminates lateral forces (no scraping

on sample). There is one minor disadvantage: slightly Slower scan speed than Contact-

Mode AFM. The AFM system at KMF (W. M. Keck Microfabrication Facility at

Michigan State University, a NSF MRSEC facility) at PHY Department of MSU is a SPM

(Scanning Probe Microscopy) model DI D3100.

3.5.2.3 Film Quality

Raman Spectroscopy can identify the presence of various forms of carbon such as

diamond, amorphous carbon, and graphite. It is a common method to determine the qual-

ity of CVD diamond ﬁlms. The method of Raman spectroscopy is described in chapter

4.]. Different Raman systems were used in this thesis research. In the early stages of this

research, when the ﬁlms had larger crystal sizes, the Raman system at ECE Department of

MSU with 30 pm spot size could detect the diamond peak. The Raman system at ECE

Department of MSU is: (1) an argon ion CW laser operates at 514.5 nm at multimode with

108

a spectral width of 10 GHz, (2) output intensity is automatically regulated within 1%, (3) a

scanning 0.75 m double monochromator which is coupled with a photomultiplier tube

(PMT) is used as the discriminating instrument, (4) a peltier element in the housing of the

PMT assures cooling down to -23 °C for reduction of dark count, (5) the double mono-

chromator has a spectral resolution of 1.08 nm/mmslitwidth, (6) slits are set accordingly to

allow high discrimination and reasonable throughput for measuring CVD diamond ﬁlms,

(7) a 180° geometry, backscattering arrangement, is chosen to allow superior use of micro-

scopic optics [Moss 1997]. In this research, the spectra resolution of the acquired spectra

is 6.6 cm‘l. The spectra are recorded in steps of 1 cm]. The Raman spectra were taken

from 1200-1600 cm'l. With a 60X objective lens of a optical microscope, the illuminated

spot Size is 30 pm in diameter.

As the research continued, the crystal size became smaller. Thus, the broadening

effect of the small crystal size required the use of the UV Raman at ANL (Argonne

National Laboratory). Unfortunately, the UV Raman became unavailable (since year

2000). Therefore, the measurements could not be performed for most of the ﬁlms. The UV

Raman system at Materials Science Division of Argonne National Lab is: (1) beginning

wavelength is 266 x 4 nm (It is Nd YAG), (2) the doubling part is commercially available

from Coherent (The Verdi laser), (3) the quadrupliing part is an external cavity made by

Laser Analytical Systems (LAS) that uses a BBC crystal, (4) spot Size with present lens is

about 100 mm wide by 200 to 300 um wide.

109

3.5.2.4

Crystal Size

Because of the small crystal Size of ultra-nanocrystalline diamond ﬁlms, determin-

ing the crystal size is a difﬁcult task with the available instruments. There are a few

SEM-analysis equipments at Michigan State University. Because of the small crystal size

of ultra-nanocrystalline diamond ﬁlms, a high resolution SEM is required. The Hitachi

S-47OOII FESEM (Field Emission SEM) at W. M. Keck Microfabrication Facility in the

Physics Department of MSU has the resolution in the order of tens of nanometers for dia-

mond ﬁlms. An Everhart-Thomley detector to acquire secondary electron images (SE

detector) was used to take the SEM pictures. With an ultra-high resolution setting on a

Hitachi S—4700II FESEM, a Optimum resolution on the order of ten nanometers can be

obtained for diamond ﬁlms. Due to charging problems non-conducting of the diamond

ﬁlms and the beam damage caused by the bombardment of the energized electron beam,

the images were often blurry and darkened during the scan.

Another approach was to use AFM technique. Using a tip with a nominal 5—10 nm

radius of curvature, TappingMode AFM can achieve 1-5 nm of lateral resolution on most

samples. However, in reality, the tip radius of curvature is more than 5-10 am. In addition

to that, the silicon probe wears out quickly during the scan on diamond samples. So, the

radius of the curvature of the tip is constantly changing. As a result, it changes the lateral

resolution of the tip as it scans. This makes the measurements of the crystal Size inaccu-

rate. Therefore, the AFM approach for determination of the crystal size was only applied

to the ﬁlms produced by the system I.

110

With the crystal size on the order of nanometers, like those synthesized in

MSU-MPACVD system 11, the Hitachi H-800 TEM (Transmission Electron Microscopy)

with 200 kV in the CHEMS (Chemical Engineering and Materials Science) Department of

MSU was used. Bright ﬁeld, dark ﬁeld imaging techniques, and SAD (Selected Area Dif-

fraction) patterns were used in this study. The growth mechanism of ultra-nanocrystalline

diamond was also studied with cross-section TEM images. Plan-view and cross-section

TEM specimens were prepared for the study of crystal sizes and growth mechanism of

ultra-nanocrystalline diamond ﬁlms.

To prepare plan-view TEM specimens, a 3 mm diameter piece of the diamond

sample was cut by a mechanical disc cutter, and a 3 mm diameter molybdenum ring was

glued to the diamond ﬁlm side using G-l epoxy (cured at 130 °C for 10 minutes). The Si]-

icon substrate was mechanically polished to ~80 pm. The center of the Specimen was

further thinned to ~10 um using a dimpler. Then the specimen was milled until perforation

with 5 kV argon ions. Therefore, the specimen could be imaged in the TEM. The resulting

specimen had a hole at the center of the electron transparent diamond ﬁlm.

To prepare cross-section TEM specimens, two (4 mm x 8 mm) pieces of the sam-

ple were cut by a wafer saw and glued face-to-face with Gl-epoxy. TWO pieces of silicon

were glued to the backs of the sample pieces if the Silicon substrates were too thin to have

a good mechanical support. A ~1.5 mm thick slice of this diamond sandwich was cut by a

wafer saw. A 3 mm diameter disk was cut out from the slice with a mechanical disc cutter.

One Side of the cross-section specimen was polished with sand paper and followed with

diamond lapping ﬁlms of progressively ﬁner grit size until a mirror ﬁnish was obtained. A

111

3 mm diameter molybdenum ring was glued to the polished side using G-l epoxy (cured

at 130 °C for 10 minutes). Then, the specimen was turned over and ground down to ~1OO

pm mechanically. The center of the specimen was dimpled to ~10 um thickness. Then the

specimen was ion milled until perforation with 5 kV argon ions. Then, the specimen was

ready to be used for imaging in the TEM. The resulting specimen had a hole located at the

boundary of Silicon substrate and diamond ﬁlm.

3.5.2.5

Optical Property

Transmission measurements were performed using a Beckman Spectrophotometer

model IR4220 in the ECE Department of MSU. Ultra-nanocrystalline diamond ﬁlms of

different thickness were grown on both side polished silicon wafers. The diamond ﬁlm

was ﬁrst deposited on one side, then on the backside. For the two layers, the silicon wafer

went through an oxidation process in a furnace, then grew the ultra-nanocrystalline dia-

mond ﬁlm on the SiOz ﬁlm. The measured transmission data were compared to the

theoretical calculations. The method of IR transmission measurements and theoretical cal-

culations are described in Section 7.2.2 and 7.2.3.

112

1 ..b1

Icy

Ola

..

e. to

e

fr. r

v

"
f

1
1
‘
(

..’

I...

 

4

MSU-MPACVD Reactor Experimental Output

Variables-Reactor Performance (Y1)

4.1

Introduction

The output variables (Y) of the experimental parameter space can be subdivided

into reactor performance variables (Y 1) and ﬁlm characteristics (Y2). The results of the

experimental investigation of the reactor performance (Y1) of the two MSU-MPACVD

systems for ultra-nanocrystalline diamond synthesis are presented in this chapter. The ﬁlm

characteristics (Y2) will be presented in next chapter. The theories that describe and pre-

dict the reactor behavior are still under development. Therefore, it’s necessary to explore

the experimental behavior of the MSU-MPACVD systems.

The output variable (Y) is a function of many input (U=[U1,U2,U3) and internal

(X) variables, i.e., Y=g(U, X), and the internal variables are dependent upon the input

variables, i.e., X=f(U). Because of the large number of variables and the intricacy between

the variables, the microwave plasma deposition process is very complicated and the reac-

tor behavior is difﬁcult to analyze. Thus in this chapter, a subset of the parameter space

shown in Figure 3.18 and their relationships are investigated. Figure 4.1 illustrates this

subset and Tables 4.1 and 4.2 display the input (U), internal (X), and output (Y) variables

considered and the associated experimental ranges that are explored in this chapter.

As shown in Figure 4.1, the input variables U are divided into three groups, (1)

controllable input variables U1 which include (a) gas chemistry Ar/HZICH4, (b) total gas

ﬂow rate ft, (0) deposition pressure p, and (d) microwave power Pabs; (2) reactor design

113

1an u,

{$.17 “(r

“can? A

I. ..f

.

Wﬂr.rhﬁ.rJu

1.... r by

.».r_ .«vrlf

with... Tn.

F5" ..r

J ..u

.

.32:

:31 Jay,
Kr: f-f

.

1.1. 0‘ o

it, .v. ‘r.

2. -4.

.1651.

0’

r17.
.rCr'L

variables U2 which are held constant and (a) include a 2.45 GHz microwave excitation fre-

quency, (b) a cavity inner diameter of 17.78 cm, (c) a cavity tuned at electromagnetic

excitation mode of TM013, (d) a quartz dome with inner base diameter of 5 inches, (e) a

molybdenum holder with thermally ﬂoating substrate holder set up (i.e., no external sub-

strate heating or cooling), (f) a discharge diameter of approximately 10 cm, (g) a vacuum

system (see Table 4.1 and 4.2), (h) purity of gases (see Table 4.1 and 4.2); and (3) deposi-

tion process variables U3 which include (a) a variable deposition time t, (b) Si (100)

substrates with a diameter of 3 inches, (c) a substrate seeding procedure described in Sec-

tion 3.3.1.2, and (d) a start up and shut down procedure described in Section 3.3.2. AS

described in Section 3.2.4.1, the reactor is operated with a thermally ﬂoating substrate

holder set up. Therefore, the substrate temperature is not independently controlled. The

gas chemistry, deposition time, total gas ﬂow rate, deposition pressure, and microwave

power are the input variables for the experiments presented in this chapter. One of the

important internal variables X is the substrate temperature, Ts. Output variables (Y) con-

sist of (1) reactor performance variables Y1 such as growth rate, speciﬁc yield, and carbon

conversion efﬁciency; and (2) ﬁlm characteristics Y2 such as ﬁlm roughness, morphology,

structure quality, ﬁlm texture and crystal size.

This chapter presents an overview of the MSU-MPACVD reactor performance Y1

versus a selected number of input variables: Ar/HZ/CH4 gas chemistry, total gas ﬂow rate

ft deposition pressure, p, microwave power, Pabs, and deposition time, t. It was experi-

mentally determined that the relationships between input, intemal, and output variables

114

were complex and non-linear,

i.e., Y1=g(Ar/H2/CH4, ft,

p,

Pabs,

TS,

t), and

Ts=f(Ar/H2/CH4. p. Pabs)'

In Section 4.2 and Section 4.3, the inﬂuences of gas chemistry, deposition time,

total gas ﬂow rate, deposition pressure and microwave power on the reactor performance

of system I and II respectively are investigated. Section 4.4 presents the experimental

results of the inﬂuence of additional nitrogen impurities on reactor performance of system

II. At the end, a summary of the reactor performance is presented in Section 4.5.

115

Y
_
_

y
c
n
e
i
c
i
f
f
E

1
Y

e
t
a
R

h
t
w
o
r
G

r
a
e
n
i
L

e
t
a
R

h
t
w
o
r
G

l
a
t
o
T

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

I

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

3

'

J

;
,

,
e
r
u
t
a
r
e
p
m
e
T

e
t
a
r
t
s
b
u
S

l
a
n
r
e
t
n
I

,
s
e
l
b
a
n
a
V

'

_

'

>
—

)
e
l
b
a
i
r
a
V
(

t
f

,
e
t
a
R

w
o
l
F

s
a
G

l
a
t
o
T

)
e
l
b
a
i
r
a
/
X
(
4
H
C
/
2
H
/
r
A

r
t
s
i
n
r
e
h
C

s
a
G

 

I

|

|

I

I

d
n
a
m
e
t
s

S
m
u
u
c
a
V

e
t
a
r
t
s
b
u
S

g
n
i
t
a
o
l
F

y
l
l
a
m
r
e
h
T

e
m
o
D
z
t
r
a
u
Q

r
e
w
o
P

e
v
a
w
o
r
c
i
M
z
H
G
5
4
.
2

r
o
f

s
e
s
a

f
o

y
t
i
r
u
P

l
o
r
t
n
o
C

y
t
i
r
u
p
m
I

p
U
t
e
S

r
e
d
l
o
H

"
5
=
r
e
t
e
m
a
i
D

o
t
e
d
o
M
3
1
0
M
T

o
t
n
i

d
e
l
p
u
o
C

e
g
r
a
h
c
s
i
D
m
c
9

a

e
t
i
c
x
E

 

 

—

—
—

y
g
o
l
o
h
p
r
o
M

m
l
i
F

e
z
i
s

l
a
t
s
y
r
C

  
—

.
_

s
s
e
n
h
g
u
o
R

m
l
i
F

e
r
u
t
x
e
T

m
l
i
F

2
Y

—

y
t
i
l
a
u
Q

e
r
u
t
c
u
r
t
S

m
l
i
F

 

s
b
a
P

,
r
e
w
o
P
e
v
a
w
o
r
c
i
M

)
e
l
b
a
i
r
a
V
(

)
e
l
b
a
i
r
a
V
(

p

,
e
r
u
s
s
e
r
P

 

—
—

 

116

n
w
o
d
-
t
u
h
S

d
n
a

g
n
i
t
r
a
t
S

e
r
u
d
e
c
o
r
P

g
n
i
d
e
e
S

e
t
a
r
t
s
b
u
S

n
o
c
i
l
i
S

t

,
e
m
i
T

n
o
i
t
i
s
o
p
e
D

s
e
r
u
d
e
c
o
r
P

”
3
=
r
e
t
e
m
a
i
D

)
e
l
b
a
i
r
a
V
(

1

L
L

1

L

3
U

S
T
N
E
M
I
R
E
P
X
E
E
H
T
R
O
F
M
A
R
G
A
I
D
K
C
O
L
B
R
P
C
M

:

1
.
4
E
R
U
G
I
F

TABLE 4.1: PARAMETER SPACE FOR EXPERIMENTS PERFORMED ON SYSTEM I

 

Controllable

a) Gas Chemistry, Ar/Hz/CH4=(99-9l)/(1-7)/(0-2) (%)

Input

b) Total Gas Flow Rate, ft=102.5-642 sccm

Variables

c) Deposition Pressure, p=120-160 Torr

U1

d) Absorbed Microwave Power, Pabs=691-1221W

 

a) 2.45 GHz Microwave Power, ﬁxed

b) Reactor Conﬁguration, MPCR, ﬁxed

c) End Feed Excitation, ﬁxed

(1) Electromagnetic Mode and Cavity Tuning, Fixed

at TMon

Input

Reactor Design

e) Reactor Wall Material, quartz dome, ﬁxed, dia.=5”

Variables

Variables

f) Substrate Holder Set-up, molybdenum holder,

U

U2

thermally ﬂoating stage

g) Vacuum System, mechanical pump, 1~2x10'2 Torr,

leak rate: 12.6 mTorr/hr

h) Purity of gases, ﬁxed, See Section 3.2.2.1

 

Deposition

a) Deposition Time, 2-48 hours

Process

b) Substrate Material and Size, 3” Si (100) wafers

Variables

c) Seeding Procedure, mechanically scratched by

U3

0.1 pm natural diamond powder

(1) Start-up and Shut-down, ﬁxed

 

Internal

Variables

X

top Side)

Substrate Temperature, TS=750-950 °C (by pyrometer,

 
  
  (1) Carbon Conversion Efﬁciency: 0.504-33.382%
 

a) Total Growth Rate: 0163-12963 mg/hr

c) Speciﬁc Yield: 4411-69 kW*hr/g

Variable

Performance

b) Linear Growth Rate: 0.0100809 rim/hr

Output

Reactor

Y

Y1

 

117

TABLE 4.2: PARAMETER SPACE FOR EXPERIMENTS PERFORMED ON SYSTEM 11

 

a) Gas Chemistry, Ar/H2/CH4=(99-90)/(0-9)/1 (%)

Controllable

b) Total Gas Flow Rate, f,=101-630 sccm

Input

c) Deposition Pressure, p=60~240 Torr

vanables

d) Absorbed Microwave Power, Pabs=732-1518W

 

U1

e) Controlled N2 impurity, 5-2500ppm

a) 2.45 GHz Microwave Power, ﬁxed

b) Reactor Conﬁguration, MPCR, ﬁxed

c) End Feed Excitation, ﬁxed

(1) EM Mode and Cavity Tuning, Fixed at TM013

Input

Reactor Design

e) Reactor Wall Material, quartz dome, ﬁxed, dia.=5”

Variables

Variables

f) Substrate Holder Set-up, molybdenum holder,

U

U2

thermally ﬂoating stage

g) Vacuum System, turbo pump, 10'7~10'6 Torr, leak

rate*: 4mTorr/hr

h) Purity of gases, ﬁxed, See Section 3.2.2.2

 

Deposition

a) Deposition Time, 8-25 hours

Process

b) Substrate Material and Size, 3” Si (100) wafers

Variables

c) Seeding Procedure, mechanically scratched by

U3

0.1 um natural diamond powder

 

Internal

Variables

X

ter, back side)

(1) Start-up and Shut-down, ﬁxed

a) Substrate Temperature, TS=UFL-77O °C (by pyrome-

  
 
 

(1) Carbon Conversion Efﬁciency: 0.542-44.094%

3) Total Growth Rate: 0.175-14.25 mg/hr

0) Speciﬁc Yield: 4229-99 kW*hr/g

Variable

Performance

b) Linear Growth Rate: 0.0110889 rim/hr

Output

Reactor

 

Y1

 

*

LW Y

easurement of leak rate was done by Brran Wright.

UFL: Temperature is too low to be within the pyrometer measurement range.

118

4.2

System I

4.2.1

Introduction

The initial ultra-nanocrystalline diamond ﬁlm synthesis experiments were per-

formed using system I with argon-hydrogen-methane plasma chemistries. According to

the results of Gruen et a1. [Zuik 1995], the gas chemistry required for ultra-nanocrystalline

diamond growth is a high concentration of argon (90 vol% and above), some methane, and

no or little hydrogen. After many trial experiments, it was discovered that in order to Stabi-

lize the plasma for the entire experiment run and remove the ﬂames inside the plasma (the

ﬂames often produce a graphitic by-product), the hydrogen to methane ratio could not be

less than 3 at any gas ﬂow rate, i.e. Hz/CH4 Z 3.

The thermally ﬂoating substrate holder setup utilized in this thesis research is

shown in Fig. 3.9. With the setup displayed in Figure 3.9, quartz tubes of different heights

can be used to change the position of the substrate with respect to the plasma to optimize

the ﬁlm deposition process. It was found that quartz tubes with about 50 mm height gave

the best results. The quartz tube with 35 mm height had a reduced growth rate around

15%. The quartz tube with 57 mm height had not only reduced the growth rate by at least

47% but also made the plasma unstable which sometimes reduced the growth rate by 76%.

The substrate temperature measurements of the experiments conducted in system I

wel‘e performed by using a pyrometer and aiming it at the surface of the substrate (see Fig-

Ute 3.8). The top-side substrate temperature measurement was about 200 °C higher (see

Tables 4.1 and 4.2) than the back—side substrate temperature measurement (see Figure

119

3.8). The intense emission from the plasma discharges inﬂuences the top—side temperature

measurements which are thus inaccurate. The experimental data measured in system I is

listed in Appendix 9.1. All data points presented in Section 4.3 are listed in Appendix 9.2.

4.2.2

Total Growth Rate = g(Ar/Hz/CH4, t, ft, p, Pabs)

Figure 4.2 displays the total growth rate plotted versus hydrogen concentration

(the percentage of hydrogen ﬂow rate to the total gas ﬂow rate). These experiments were

performed at a constant pressure of 120 Torr and deposition time of 8 hours. The argon

concentration was varied from 90.91-98.68%, the H2 concentration varied from

0.99—7.27%, and the CH4 concentration varied from 0.33—1.85% with the total gas ﬂow

rate varied from 102.5-642 sccm. The performance of the reactor’s total growth rate versus

hydrogen concentration is shown in Figure 4.2 and thus deﬁnes the experimental perfor-

mance of the system I at the pressure of 120 Torr. Figure 4.2 Shows (i) given a constant

CH4 concentration, the growth rate increases as H2 concentration increases (if provided

with enough methane concentration to supply carbon atoms for the diamond growth), (ii)

given a constant H2 concentration, growth rate increases as CH4 concentration increases.

 

120

15

I

I

I

I

I

 

I

I

0.95

 

CH4~0.5°/o

*

0

 

CH4~0.95°/o

CH4~1.85°/o

 

 

O

8

O

i?

I
'
1
u
o
e
r
g
r
o
w
q
a
e
r
a

)
1
r

r
u
/
q
J
(

~0317

 

0.00

5

DD

5

*3
0.1

S

E

e

O

8 5»

I“

 

0

i

a

g

31'

$1

r

r

r

1

l

1:,

a

r

1

 

0

1

2

3

4

5

6

7

8

FIGURE 4.2 : TOTAL GROWTH RATE VS. HYDROGEN CONCENTRATION

Pressure=120 Torr, deposition time=8 hours,

Ar:H2:CH4=(9O.91-98.68):(O.99-7.27):(0.33-l .85)%

and total gas ﬂow rate varied from 102.5-642 sccm

by system I.

121

Figure 4.3 displays the total growth rate versus the deposition time, t, with a con-

stant pressure of 120 Torr, ArszzCH4 =100:2:0.5, and f,=615 sccm. As shown, the growth

rate increases about 25% from an 8 hour run to a 24 hour run. The lower growth rate at the

shorter deposition time region may be due to the presence of impurities at the beginning of

the experiments.

 

T

I

I

I

0.075

O

e

O 8

1_

A

g 0.8-

é

a)

‘5

9‘ 06—

o

0.4—

e

0
3g

r9

o ~0.0625

—0.05

E

8
8:

o

8

2

-0.0375 9

8

9
-0025 3

:3.

0.2-

  
 

L

r

~0.0125

0

r

1

000

0

10

2o

30

40

50

Time (hours)

FIGURE 4.3 : TOTAL GROWTH RATE VS. DEPOSITION TIME

Pressure=120 Torr and Ar:H2:CH4=100:2:0.5, and ft=615 sccm by system I

122

)
I
I
I
/
s
u
n
(

v
l
u
R

l
'
l
’
\
\
)
(
l
;

(

”
.
t
I

:
'
1
'

Figure 4.4 displays the total growth rate plotted as a function of total gas ﬂow rate,

ft, with pressure at 140 Torr, deposition time of 8 hours, and three different gas chemistry

ratios. Figure 4.4 indicates that the total gas ﬂow rate does not inﬂuence the growth rate.

With a ﬁxed total gas ﬂow rate, the growth rate is mainly affected by gas chemistry. In Fig-

ure 4.4, the gas chemistry is expressed as ratios.

9

r

.

.

.

r

1

0.560

 

O

‘1‘

8‘

7~

g 6*

E”

I 5-
a

°‘
5
g 4—

o

g 3-

 

O

a

Ar:H2:CH4=100:4:1

Ar:H2:CH4=100:3:1

 
  

Ar:H2:CH4=100:2:0.5 10‘498

<2

o

O

O

-O.436

r

,

0.373 g

w

:3,

«0.311 a
2
9
7::
40.249 g.

4S

-0.l87 g

40.124

«0.062

 

2*

1-

 

0

1

1

1

1

1

1

0000

 

O

100

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 4.4 : TOTAL GROWTH RATE VS. TOTAL GAS FLOW RATE

Pressure=140 Torr and deposition time=8 hours by system I

123

Figures 4.5 and 4.6 display the total growth rate versus deposition pressure, p, and

absorbed microwave power, Pabs, respectively, with deposition time of 8 hours and fl:

102.5-630 sccm. Because at a constant input power, the discharge area Ad decreased with

the deposition pressure, p. During the ultra-nanocrystalline diamond ﬁlm synthesis with

system I, the microwave power input was increased as the pressure increased to maintain a

discharge area Ad that covered the three inch substrates.

As shown in Figure 4.5, the growth rate increases with increases in deposition

pressure. With a ﬁxed pressure, the higher hydrogen or methane concentration resulted in

higher growth. This result is consistent with the data in Figure 4.2. In Figure 4.6, the

growth rates of the experiments with gas chemistry ratio of Ar:H2:CH4=10022:0.5 approx-

imately doubles as the absorbed microwave power increases from 702W to 903 W. The

growth rates of those experiments with gas chemistry ratios of ArszzCH4=100:3:l or

100:4:1 increase rapidly from 3.450 mg/hr to 8.675 mg/hr for ArzH22CH4=100:3:1 and

from 4.625 mg/hr to 10.788 mg/hr for ArszzCH4=10024zl as the absorbed microwave

power increases from 795W to 1017 W. But at the same time, the deposition pressure is

increased, too. The contribution can come from either deposition pressure or absorbed

microwave power.

124

 12

I

I

1

l

l

0.78

 

O

a

Ar:H2:CH4=100:4:1

Ar:H2:CH4=100:3:1

 
  

Ar:H2:CH4=100:2:0.5

0

o

a

70.65

‘

~0.52

‘1?

i

Q

@

“0.39

a0.26

10-

8

6-

4-

)
r
h
/
g
m
(

e
t
a
R

h
t
w
o
r
G

l
a
t
o
T

n
a
o
m
)
f
r
o
m
r
q

3
2
1
9

)
1
1

r
u
/
q
r
(

2.

I“:

*

  
 

9

L

l

l

013

9

l

1

0.00

10

120

130

140

150

160

170

Pressure (Torr)

FIGURE 4.5 : TOTAL GROWTH RATE VS. DEPOSITION PRESSURE

Deposition time=8 hours and total gas ﬂow rate=100-600 sccm by system I

125

 

12

I

l

l

l

l

I

l

0.75

 

Ar:H2:CH4=100:4:1

Ar:H2:CH4=100:3:1

O

O

a

0

 

Ar:H2:CH4=100:2:O.5

 

 

O

080

0

0

0

if?!

9

a

s

‘1

10-

8-

6-

4-

)
r
h
/
g
m
(

e
t
a
R

h
t
w
o
r
G

l
a
t
o
T

2*

 

~ 0.625

I?

- 0.5

- 0.375

~ 0.25

]
'
I
U
Q
B
J

)
r
r
o
m
i
q

a
e
r
a

)
1
7

r
u
/
q
J
(

- 0.125

 

 

0.0

850

700

750

800

850

900

950

1000

1050

Absorbed Microwave Power (W)

FIGURE 4.6 : TOTAL GROWTH RATE VS. ABSORBED POWER

Deposition time=8 hours and total gas ﬂow rate=100-600 sccm by system I

126

4.2.3

Speciﬁc Yield = g(Ar/Hz/CH4, t, ft, p, Pabs)

Speciﬁc yield is a measurement of the energy efﬁciency of the deposition process.

The lower the Speciﬁc yield, the less microwave power required to deposit a given weight

of ultra-nanocrystalline diamond.

Figure 4.7 displays the speciﬁc yield versus the hydrogen concentration (the per-

centage of hydrogen ﬂow rate to the total gas ﬂow rate). These experiments were

performed at a constant pressure of 120 Torr and deposition time of 8 hours. The argon

concentration was varied from 90.91-98.68%, the H2 concentration varied from

O.99-7.27%, and the CH4 concentration varied from O.33-1.85% with the total gas ﬂow

rate varied from 102.5-642 sccm. Figure 4.7 thus deﬁnes the microwave energy efﬁciency

of the system I at the pressure of 120 Torr. In particular, Figure 4.7 shows (i) given a con-

stant CH4 concentration, Speciﬁc yield decreases as H2 concentration increases, (ii) given

a constant H2 concentration, speciﬁc yield decreases as CH4 concentration increases (if

H2/CH4 Z 3 to prevent the present of ﬂames inside the plasma, which reduces the energy

efﬁciency for diamond growth).

127

5000

 

4500~

4000 -

3500 -

3000 ~

2500 -

2000 -

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

1500 b

1V,

 

0

CH4~0.5°/o

.
o

CH4~0.95°/o .
CH4~1.85°/o
 
  
-

-

_

_

-

T

1000 -

500 —

 
  

w

1,

1 O

l

(y

l

_

-

3

0

1

l

1

*1

o

1

2

4

5

6

7

H2 (%)

FIGURE 4.7 : SPECIFIC YIELD VS. HYDROGEN CONCENTRATION

Pressure=120 Torr, deposition time=8 hours,

Ar:H2:CH4=(90.91—98.68):(0.99-7.27):(0.33-l .85)%

and total gas ﬂow rate varied from 102.5-642 sccm

by system I.

128

Figure 4.8 displays the speciﬁc yield versus the deposition time, t, with pressure at

120 Torr, Ar:H2:CH4 =100/2/0.5, and ft=615 sccm. As shown, the speciﬁc yield becomes

lower as deposition time increases. Since speciﬁc yield is proportional to the reverse of

total growth rate, the higher speciﬁc yield in the shorter deposition time region is a reﬂec-

tion of the lower growth rate.

1500

 

.

,

.

.

O

O

0 e

o

3, 1000-

.5

i

E

2
'>~

s3

5%

500.

O

0

-

-

 
  

l

l

l

l

0

0

10

20

30

4o

50

Time (hours)

FIGURE 4.8 : SPECIFIC YIELD VS. DEPOSITION TIME

Pressure=120 Torr and Ar:H2:CH4=100:2:0.5, ft=615 sccm by system I

129

Figure 4.9 displays the speciﬁc yield versus total gas ﬂow rate, ft, with pressure at

140 Torr and deposition time of 8 hours. The data indicate that the total gas ﬂow rate does

not inﬂuence the speciﬁc yield when the plasma is stable and no ﬂames are inside the

plasma. With a ﬁxed total gas ﬂow rate, the speciﬁc yield is mainly inﬂuenced by gas

chemistry.

1 000

I

I

I

I

 

 

900

I

800

I

700

I

600

I

500

I

400

I

300

I

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

<3

a

Ar:H2:CH4=100:2:0.5

Ar:H2:CH4=100:3:1

_

 
  

Ar:H2:CH4=100:4:1

0

I

1

l

l

J

l

200

T

 

100*

1

- 

a

s
a

»
3
:

}
1

l

l

l

 

0

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 4.9 : SPECIFIC YIELD VS. TOTAL GAS FLOW RATE

Pressure=l40 Torr and deposition time=8 hours by system I

130

Figure 4.10 and 4.11 display the speciﬁc yield versus deposition pressure, p, and

absorbed microwave power, Pabs, respectively, with deposition time of 8 hours. AS shown

in Figure 4.10, the speciﬁc yield decreases with increases in deposition pressure. With a

ﬁxed pressure, higher hydrogen and methane concentration result in lower speciﬁc yield.

This result is consistent with the data in Figure 4.7.

131

1
'
I
.

 1 500

 

0

a

Ar:H2:CH4=100:2:0.5

Ar:H2:CH4=100:3:1

 
  

Ar:H2:CH4=100:4:1

o

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

1 000

I

1 O O I

0

}
1
1

3

o

n
o
n
a
c

  
 

X?
0

$10

120

130

140

150

160

170

Pressure (Torr)

FIGURE 4.10 : SPECIFIC YIELD VS. DEPOSITION PRESSURE

Deposition time=8 hours by system I

132

 1 500

 

\

{2

a

Ar:H2:CH4=100:2:0.5

Ar:H2:CH4=100:3:1

 
  

Ar:H2:CH4=100:4:1

o

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

1 000

I

1 O O I

0

0

tr

Q) 0

g:

at?

000 o

a

O

 

 

 

850

700

750

800

850

900

950

1000

1050

Absorbed Microwave Power (W)

FIGURE 4.11 : SPECIFIC YIELD VS. MICROWAVE POWER

Deposition time=8 hours by system I

133

4.2.4 Carbon Conversion Efﬁciency = g(Ar/H2/CH4, t, ft, p, Pabs)

The carbon conversion efﬁciency is a measure of how many of the input carbon

atoms that are actually converted to diamond in the ultra-nanocrystalline diamond ﬁlms.

Figure 4.12 displays the carbon conversion efﬁciency versus the percentage of

hydrogen input gas. These experiments were performed at a constant pressure of 120 Torr

and deposition time of 8 hours. The argon concentration was varied from 90.91-98.68%,

the H2 concentration varied from 0.99-7.27%, and the CH4 concentration varied from

0.33-1.85% with the total gas ﬂow rate varied from 102.5-642 sccm. The performance of

the reactor’s carbon conversion efﬁciency versus hydrogen concentration is shown in Fig-

ure 4.12 and thus deﬁnes the carbon conversion efﬁciency of the system 1 at the pressure

of 120 Torr. In particular, Figure 4.12 Shows (i) given a constant CH4 concentration, car-

bon conversion efﬁciency increases as H2 concentration increases, (ii) given a constant 1H2

concentration, carbon conversion efﬁciency increases as CH4 concentration increases.

134

25

 

 

O

CH4~O.5°/o

‘1

CH4~O.95°/o

  
 

CH4~1.85°/o

O

20

"

15

I

10

)
%
(

y
c
n
e
i
c
ﬁ
f
E

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

O

a
a
it

a

O

O

O

5..

  
 

is

t

I

[All

1

1

a

l

l

1

r

a

a

O

0

1

2

3

4

5

6

7

H2 (%)

FIGURE 4.12 : CARBON CONVERSION EFFICIENCY VS. HYDROGEN

CONCENTRATION

Pressure=120 Torr, deposition time=8 hours,

Ar:H2:CH4=(9O.91-98.68):(0.99-7.27):(0.33-1.85)%

and total gas ﬂow rate varied from 102.5-642 sccm

by system I.

135

Figure 4.13 displays the carbon conversion efﬁciency versus the deposition time, t,

with pressure at 120 Torr, Ar:H2:CH4 =100:2:0.5, and ft=615 sccm. As shown, the carbon

conversion efﬁciency increased about 20% from an 8 hour run to a 24 hour run. Since car-

bon conversion efﬁciency is proportional to total growth rate, the lower carbon conversion

efﬁciency in the shorter deposition time region is a reﬂection of the lower growth rate.

 

0.8 r

g

0.6 ~

>
<

r

~

 

L

l

1

1

)
%
(

y
c
n
e
i
c
ﬁ
f
E

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

 

o

 

0

1 0

20

30

4O

50

Time (hours)

FIGURE 4.13 : CARBON CONVERSION EFFICIENCY VS.

DEPOSITION TIME

Pressure=120 Torr, Ar:H2:CH4=100:2:0.5, and ft=615 sccm by system I

136

Figure 4.14 displays the carbon conversion efﬁciency versus total gas ﬂow rate, f ,

with pressure at 140 Torr and deposition time of 8 hours and indicates that carbon conver-

sion efﬁciency increases as total gas ﬂow rate decreases. The carbon conversion efﬁciency

increases dramatically at low-total-gas-ﬂow-rates. The carbon conversion efﬁciency of

100 sccm total gas ﬂow rate is over 5 times than 600 sccm total gas ﬂow rate. The carbon

conversion efﬁciency is affected by both gas chemistry and total gas ﬂow rate.

 

25

I

I

I

I

I

 

I

O

a

O

Ar:H2:CH4=100:4:1

Ar:H2:CH4=100:3:1

Ar:H2:CH4=100:2:0.5

 

 

 

O

a

-

4

a

_

 

O

f:

20-

e3

s

'3 15-

E
LL]

.5

:2

2
g 10—

U

8

.D

:3

U 5-

 

0

O

a

0

a

0

ﬁ’

l

l

1

l

1

l

 

0

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 4.14 : CARBON CONVERSION EFFICIENCY VS.

TOTAL GAS FLOW RATE

Pressure=140 Torr and deposition time=8 hours by system I

137

Figure 4.15 and 4.16 display the carbon conversion efﬁciency versus deposition

pressure, p, and absorbed microwave power, Pabs, respectively, with deposition time of 8

hours. As Shown in Figure 4.15, the carbon conversion efﬁciency increases with increases

in deposition pressure. With a ﬁxed pressure, higher hydrogen and methane concentration

usually result in higher carbon conversion efﬁciency. This result is consistent with the data

in Figure 4.12. The ﬂuctuation of the data points is caused by different total gas ﬂow rates.

In Figure 4.16, the carbon conversion efﬁciency of the experiments with gas chemistry

ratio of Ar:H2:CH4=100:2:0.5 increase as the absorbed microwave power increases from

702W to 903 W. The carbon conversion efﬁciency of those experiments with gas chemis-

try ratios of Ar:H2:CH4=100:3:1 and 100:4:1 increase rapidly as the absorbed microwave

power increases from 795W to 1017 W.

138

35

I

I

 

I

j

I

 

O

a

<>

30 -

 

Ar:H2:CH4=100:4:1

0

Ar:H2:CH4=100:3:1

Ar:H2:CH4=100:2:0.5

 

 

g 25'

g

if 20—

In

.8.

2 15-

a

5
-°
5 10-

5_

 

0

tr

0

i“:

i?

{i

l

o

I

6

5

0

e

E

l

.

l

l

~

_

-

—

_

 

 

910

120

130

140

150

160

170

Pressure (Torr)

FIGURE 4.15 : CARBON CONVERSION EFFICIENCY VS.

DEPOSITION PRESSURE

Deposition time=8 hours by system I

139

 

 

 

O

a

Ar:H2:CH4=100:4:1

o

Ar:H2:CH4=100:3:1

 
  

Ar:H2:CH4=100:2:0.5

O

30—

g3 25-

g

E3 20—

1.1.1

.5

g

E 15’

8

I:

é
5 10-

o

a

0

o

a

5-

*0

t?

O

  
 

it *0.

0 O

,

O

C"

850

700

750

800

850

900

950

Absorbed Microwave Power (W)

FIGURE 4.16 : CARBON CONVERSION EFFICIENCY VS.

MICROWAVE POWER

Deposition time=8 hours by system I

140

4.3

System II

4.3.1

Introduction

Based upon the “optimized” experimental results from system I, the thermally

ﬂoating substrate holder setup with quartz tube height of about 50 mm was used for the

synthesis of ultra-nanocrystalline diamond ﬁlms in system H. The temperature of the

Neslab chiller of system II was set at 15°C; i.e. it was set (1) to be close to the temperature

of the water cooling in system I in order to provide a similar growth environment as sys-

tem I, and (2) to provide a safe condition for the higher pressure regime experiments. The

substrate temperature measurements of the experiments conducted in system 11 were per-

formed by using a pyrometer directed at the mirror (see Figure 3.8). According to the

results from system I, the higher total gas ﬂow rate (1) does not increase the grow rate, and

(2) has lower carbon conversion efﬁciency. Therefore, most of the experiments performed

with system H had a total gas ﬂow rate of ~100 sccm. The experimental data for Section

4.3 is given in detail in Appendix 9.2. Using the system II and operating under 240 Torr,

the discharge is ﬂame free.

4.3.2

Total Growth Rate = g(Ar/Hz/CH4, t, ft, p, Pabs, Ts)

Figure 4.17 displays the total growth rate versus hydrogen ﬂow rate with all other

variables held constant, i.e. argon ﬂow rate is ﬁxed at 100 sccm, methane ﬂow rate is ﬁxed

at 1 sccm, the pressure is held constant at 120 Torr and deposition time is 8 hours. As

shown, the total growth rate increases from 2.06 mg/hr to 5.4 mg/hr when the hydrogen

ﬂow rate is varied from O to 10 sccm.

 

141

 

6

m

I

I

I

I

0.375

5 ~

4 _

3 -

2<2

)
r
h
/
g
m
(

e
t
a
R

h
t
w
o
r
G

l
a
t
o
T

0

O

<>

0

- 0.3125

0

o

r 0.25

. 0.1875

-0.125

]
r
u
c
e
r

9
1
0
1
1
/
I
I
I

;
)
e
r
a

)
1
1

w
/
I
I
I
(

1 ~

 

~ 0.0625

 

0

1

1

1

1

1

0.0

 

O

2

4

6

8

10

H2 (sccm)

FIGURE 4.17 : TOTAL GROWTH RATE VS. H2 FLOW RATE

Pressure=120 Torr, Ar=100 sccm, CH4=1 sccm and deposition time=8 hours

by system 11

142

In order to understand why the hydrogen concentration had such a contribution to

total growth rate of ultra-nanocrystalline diamond synthesis, the study of the inﬂuence of

the hydrogen concentration on substrate temperature and absorbed microwave power was

performed as Shown in Figure 4.18. Figure 4.18(a) Shows that the substrate temperature

increases as hydrogen concentration increases. Figure 4.18(b) shows the absorbed micro-

wave power increases as hydrogen ﬂow rate increases. Overall, higher hydrogen

concentration has increased the substrate temperature and absorbed microwave power

directly and indirectly. The covered area by the plasma was about the same.

143

 

640'

A 620'

<3

“I 600-

580t

 

.

‘

‘

-

 

‘

 

0

2

4

6

8

10

(a)

H2 (sccm)

950

I

I

I

I

I

 

:9...

 

850

-

 

1

1

1

1

1

 

0

2

4

6

8

10

(b)

H2 (sccm)

FIGURE 4.18 : THE INFLUENCE OF H2 FLOW RATE ON Ts AND Pabs

Arr-100 sccm, CH4=1 sccm and pressure=120 Torr by system 11

144

Figure 4.19 displays the total growth rate versus the deposition time, t, with pres-

sure at 160 Torr and Ar/H2/CH4 =100/4/l sccm. As shown, the growth rate is

approximately constant between 8 hour and 25 hour experiments. The experiments here

were not done with the same quartz dome and that could affect the growth rate slightly.

Therefore, the data points ﬂuctuate a little. With these exploratory experiments, one can

notice that system II does not behave the same as system I (See Figure 4.3).

 

10

I

I

I

I

I

I

I

I

I

0.625

9 _

-0.563

8 _

1’}

if;

:31

‘0.500

A 7 ‘

“

s

a:

4.35

5 .

e
o 4

_

'3

a 3 _

#0438 I:

:1

2

-O.313 E;—

.

E
a
0.250 E‘

~0.188 %

2 _

1 _

  
 

1

1

1

1

1

1

1

0

1

1

—O.125

~0.063

0000

8

10

12

14

16

18

2O

22

24

26

Time (hours)

FIGURE 4.19 : TOTAL GROWTH RATE VS. DEPOSITION TIME

Pressure=160 Torr and Ar/H2/CH4=100/4/1 sccm by system 11

145

Figure 4.20 displays the total growth rate versus total gas ﬂow rate, ft, with pres-

sure at 120 Torr and deposition time of 8 hours and it indicates that the total gas ﬂow rate

does not inﬂuence the growth rate. This result iS similar to system I but the

high-total-gas-ﬂow-rate experiments run at system II did not have ﬂames in the plasma

discharge as it sometimes happened in system I. Also, the plasma with this gas chemistry

ratio, Ar:H2:CH4 =100:2:1 was unable to be Stabilized in system I.

5

 

r

r

r

r

r

1

0,310

45 l'

*3

  
 

AIIIHZCH4=1 001321 _ 0.279

 

O

Ar:H2:CH4=100:4:1

-

4

O

O

o

O

-

0.248

O

c 3.5

En

V 3"

o
.5

a: 25-

5

3

e

0

To

13

_

2

1.5—

0.217 1:

i

"0.186 O

8

é

«0.156 S—

7O
e.

(b

_

0.124 5:?

0093 E

1 -

  
 

1

1

1

1

1

1

0

0.5-

~0.062

-0.031

0000

0

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 4.20 : TOTAL GROWTH RATE VS. TOTAL GAS FLOW RATE

Pressure=120 Torr and deposition time=8 hours by system 11.

146

The study result of the inﬂuence of the total gas ﬂow rate on substrate temperature

and absorbed microwave power is shown in Figure 4.21. Figure 4.21(a) indicates that the

total gas ﬂow rate has little inﬂuence on substrate temperature. Figure 4.21(b) indicates

that the total gas ﬂow rate has no inﬂuence on absorbed microwave power. Total gas ﬂow

rate has no effect on absorbed microwave power and very little inﬂuence on substrate

temperature.

147

 

 

100

200

300

400

500

600

700

(a)

Total Gas Flow Rate (sccm)

950

 

 

)
W
(

s
b
a
P

900-

 

 

 

 

850

 

0

100

200

300

400

500

600

700

(b)

Total Gas Flow Rate (sccm)

FIGURE 4.21 : THE INFLUENCE OF TOTAL GAS FLOW RATE ON

Ar:H2:CH4=100:4:1 and pressure=120 Torr by system H.

Ts and Pabs

148

Figure 4.22 and 4.23 display the total growth rate versus deposition pressure, p,

and absorbed microwave power, Pabs, respectively, with argon ﬂow equal to 100 sccm,

methane ﬂow equal to l sccm and deposition time of 8 hours. Because at a constant input

power, the discharge area Ad decreased with the deposition pressure, p. Same methodol-

ogy as in system I, while the ultra-nanocrystalline diamond ﬁlms were synthesized with

system H, the microwave power input was increased as the pressure increased to maintain

a discharge area Ad that covered the three inch substrates.

As shown in Figure 4.22, the total growth rate increases as pressure increases and

the higher hydrogen concentrations generally produce higher growth rate at the same pres-

sure. A small ﬂame starts to show at the center of the plasma at different pressure for each

different hydrogen gas ﬂow rate. For the experiments of H2=4 sccm, it’s around 240 Torr.

For the experiments of H2=2 sccm, it’s around 220 Torr. For the experiments of H2=l and

0 sccm, it’s around 200 Torr. It has been investigated that adding more power can reduce

the ﬂame but the operator should be careful not to overheat the reactor at higher pressure

region. In Figure 4.23, the growth rate increases with increases in absorbed microwave

power. But at the same time, the deposition pressure is increased, too. The contribution

can come from either deposition pressure or absorbed microwave power. At a ﬁxed pres-

sure, those experiments with higher hydrogen concentration have higher absorbed

microwave power.

As we have learned from Figure 3.12, the substrate temperature increases with

increases in pressure and absorbed microwave power. This explains why the growth rate

149

increases with increases in pressure and absorbed microwave power. For the values of data

points, please see appendix.

15

 

u

r

.

0.942

 

0

H2=4 sccm

* H2=2 sccm

<> H2=1sccm

 
  

v H2=0 sccm

1: 10-

go

5:,

g

‘3‘
a
g
a
{é

5-

0

0

o

o

v:

V

V

8

x
v

o
*

o

O

ﬁ’

-O.628 E

8

i

n

2
e

5
a
9
03143

<7

0

it

{'7‘

  
 

§

0

0

0

Q

6

1

1

I

0.000

50

100

150

200

250

Pressure (Torr)

FIGURE 4.22 : TOTAL GROWTH RATE VS. PRESSURE

Arr-100 sccm, CH4=l sccm and deposition time=8 hours by system 11.

150

 

15

I

I

I

I

 

0

9:

O

H2=4 sccm

H2=2 sccrn

H2=1 sccm

v H2=O sccm

 

 

 

I

0.942

O

I

0

o

*

a

7: 10

O

-0.628

q
t
u
a
m
Q
J
O
M
J
[
[

3
9
1
9

)

l
r

w
/
I
I
I
(

~0.314

 

0000

f.

E

*3

m

a

e

2

{3

5—

*

w

o

I’\

_

"V

V

 

rt

0

V

..

it

Q.

0 ﬁg 1

l

1

l

1

l

1

1

 

700

800

900

1000

1100

1200

1300

1400

1500

1600

Absorbed Microwave Power (W)

FIGURE 4.23 : TOTAL GROWTH RATE VS. MICROWAVE POWER

Ar=lOO sccm, CH4=l sccm and deposition time=8 hours by system II.

15]

4.3.3

Speciﬁc Yield=g(Ar/H2/CH4, t, ft, p, Pabs)

Figure 4.24 displays the speciﬁc yield versus hydrogen ﬂow rate, with argon ﬂow

ﬁxed at 100 sccm, methane ﬂow ﬁxed at 1 sccm, pressure at 120 Torr and deposition time

of 8 hours by system 11. As shown, the speciﬁc yield decreases with the increases in the

hydrogen concentration.

 500

.

.

ﬂ

,

r

450

I

0

400 -

350

I

300 -

250 -

200 _

150

l

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

<>

I

—

l

-

_

_

l

100—

  O
— 

l

l

l

l

l

0

2

4

6

8

10

H2 (sccm)

FIGURE 4.24 : SPECIFIC YIELD VS. H2 FLOW RATE

Pressure=120 Torr, AFIOO sccm, CH4=1 sccm and deposition time=8 hours

by system H

152

Figure 4.25 displays the speciﬁc yield versus the deposition time, t, with pressure

at 160 Torr and Ar/HZ/CH4 =100/4/l sccm. As shown, the speciﬁc yield was constant

between 8 hour and 25 hour experiments.

 200

I

r

I

I

I

I

r

l

I

180*

160‘

r:

140*

120~

100*

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

-

‘

-

_

'1

b OT

N O I

 

1

1

 

 O

_

L

I

l

l

l

l

I

l

10

12

14

16

18

2O

22

24

26

Time (hours)

FIGURE 4.25 : SPECIFIC YIELD VS. DEPOSITION TIME

Pressure=160 Torr and Ar/HZICH4=100/4/ 1 sccm by system H

153

Figure 4.26 displays the speciﬁc yield versus total gas ﬂow rate, ft, with pressure at

120 Torr and deposition time of 8 hours and it indicates that the total gas ﬂow rate does not

inﬂuence the speciﬁc yield. The ﬂuctuation of the data points is because the experiments

are not performed with the same quartz dome and they may not have the same energy

efﬁciency.

 400

I 

I

I

I

350

I

300-

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

250

I

O

O

200 -

150-

<>

Ar:H2:CH4=100:2:1

Ar:H2:CH4=100:4:1  
  

O

l

_

1

~

_

100*

50-

 

‘

‘ 

O

4

I

I

l

l

l

 

0

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 4.26 : SPECIFIC YIELD VS. TOTAL GAS FLOW RATE

Pressure=120 Torr and deposition time=8 hours by system 11.

154

 

Figure 4.27 and 4.28 display the speciﬁc yield versus deposition pressure, p, and

absorbed microwave power, Pabs, respectively, with argon ﬂow equal to 100 sccm, meth-

ane ﬂow equal to l sccm and deposition time of 8 hours. As shown in Figure 4.27, the

speciﬁc yield decreases as pressure increases. The higher hydrogen concentrations have

lower speciﬁc yield at a ﬁxed pressure. As a result, the microwave energy efﬁciency is

higher. This result is consistent with the data in Figure 4.24. In Figure 4.28, the speciﬁc

yield decreases with increases in absorbed microwave power. But meanwhile, the deposi-

tion pressure is increased, too. The contribution can come from either deposition pressure

or absorbed microwave power.

155

5000 

 

4500 _

v

4000 -

3500 r

3000-

2500* *

v

2000 -

O

a

1500 -

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

v

0

w

H2=0 sccm

H2=1 SCCI’TI

H2=2 sccm

 
  

H2=4 sccm

O

_

-

-

‘

_

o

v

a

0

.

g

‘1?

.

I

I

.

o

-

F

 

1000 *

500 '

 

0

 

50

100

150

200

250

Pressure (Torr)

FIGURE 4.27 : SPECIFIC YIELD VS. PRESSURE

A1=100 sccm, CH4=1 sccm and deposition time=8 hours by system 11.

156

5000

I

I

T

r

I

I

T

I

 

 

V H2=0 sccm

0 H2=1sccm _

a H2=2 sccm

 

 
o H2=4 socm _

 

4500L

V

4000-

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

3500 -

3000 -

2500 -

*

2000

9.

1500 -

1000-

500-

 

O

d

-

A

d

d

 

ﬂ

"

8,

*

0

17
.ﬁ 0

'3: V0

p

,

It

w

:10"

l

l

l

I

1w 0

p 0

1O

 

700

800

900

1000

1100

1200

1300

1400

1500

1600

Absorbed Microwave Power (W)

FIGURE 4.28 : SPECIFIC YIELD VS. MICROWAVE POWER

Ar=100 sccm, CH4=1 sccm and deposition time=8 hours by system II.

157

4.3.4 Carbon Conversion Efﬁciency = g(Ar/Hz/CH4, t, ft, p, Pabs)

Figure 4.29 displays the carbon conversion efﬁciency versus hydrogen ﬂow rate,

with argon ﬂow ﬁxed at 100 sccm, methane ﬂow ﬁxed at 1 sccm, pressure at 120 Torr and

deposition time of 8 hours by system II. As shown, the carbon conversion efﬁciency

increases with the increases in the hydrogen concentration.

18

T

I

 

F

I

I

)
%
(

y
c
n
e
i
c
ﬁ
f
E

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

16 I

14

I

12 I

10 I

>
<

O

O

l

1

l

1

  
 

00

2

4

6

8

10

H2 (sccm)

FIGURE 4.29 : CARBON CONVERSION EFFICIENCY VS. H2 FLOW RATE

Pressure=120 Torr, Ar=100 sccm, CH4=l sccm and deposition time=8 hours

by system 11

158

Figure 4.30 displays the carbon conversion efﬁciency versus the deposition time, t,

with pressure at 160 Torr and Ar/H2/CH4 =100/4/1 sccm. As shown, the carbon conver-

sion efﬁciency remains constant between 8 hour and 25 hour experiments. The

experiments here were not done with the same quartz dome and that could affect the car-

bon conversion efﬁciency slightly. Therefore, the data points ﬂuctuate a little.

)
%
(

y
c
n
e
i
c
ﬁ
f
E

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

30

I

I

I

I

I

I

I

I

I

 

25—,

20-

15—

g-

-

-

 
  

I

I

I

I

I

O

l

l

l

l

8

1 0

1 2

14

1 6

1 8

20

22

24

26

Time (hours)

FIGURE 4.30 : CARBON CONVERSION EFFICIENCY VS. DEPOSITION

Pressure=l60 Torr and Ar/HZ/CH4=100/4/l sccm by system 11

TIME

159

Figure 4.31 displays the carbon conversion efﬁciency versus total gas ﬂow rate, f ,

with pressure at 120 Torr and deposition time of 8 hours and indicates that carbon conver-

sion efﬁciency increases as total gas ﬂow rate decreases. This result is similar to system I

but the plasma with this gas chemistry, Ar:H2:CH4 =100:2:1 was unable to be stabilized in

system I.

 

12

b

I

I

I

I

I

 

O

Ar:H2:CH4=100:4:1

 
  

Ar:H2:CH4=100:2:1

10-

8

>.

8—

E
In
a
E

O)

E

8
c:

43

8

O

6—

4*

2-

 

a

a

-

_

‘

 

O

‘

O

0

O

l

l

I

l

l

1

 

0

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 4.31 : CARBON CONVERSION EFFICIENCY VS. TOTAL GAS

Pressure=120 Torr and deposition time=8 hours by system 11.

FLOW RATE

160

Figure 4.32 and 4.33 display the carbon conversion efﬁciency versus deposition

pressure, p, and absorbed microwave power, Pabs, respectively, with argon ﬂow equal to

100 sccm, methane ﬂow equal to l sccm and deposition time of 8 hours. As shown in Fig-

ure 4.32, the carbon conversion efﬁciency increases as pressure increases. Also higher

hydrogen concentrations generally produce higher carbon conversion efﬁciency at the

same pressure. In Figure 4.33, the carbon conversion efﬁciency increases with increases in

absorbed microwave power. But at the same time, the deposition pressure is increased, too.

The contribution can come from either deposition pressure or absorbed microwave power.

161

45 

I

 

O

*

0

a?

V

4

.

-

-

-

-

-

-

 

o

a

<7

o

*

__

H2=4 sccm

H2=2 sccm

40

0 H2=1sccm

 
  

v H2=0 sccm

35-

A

be

: 30-

8

E 25-

E

g 20*

5

g 15-

6

10-

 

5—

0

O

6

g

O
x:

v

1

1

1

 

50

100

150

200

250

Pressure (Torr)

FIGURE 4.32 : CARBON CONVERSION EFFICIENCY VS. PRESSURE

Ar=100 sccm, CH4=1 sccm and deposition time=8 hours by system 11.

162

 45

I 

I

I

I

b

I

l

O

H2=0 sccm

_

1: H2=1sccm

O

4O

0

H2=2 sccm

O

 
  

H2=4 sccm

v

*

a?

35-

20—

15*

)
%
(

y
c
n
e
i
c
ﬁ
f
E

n
o
i
s
r
e
v
n
o
C

n
o
b
r
a
C

 

 

O

.17“? .

l

l

L

l

l

l

l

 *7

700

800

900

1000

1100

1200

1300

1400

1500

1600

Absorbed Microwave Power (W)

FIGURE 4.33 : CARBON CONVERSION EFFICIENCY VS. MICROWAVE

AI=100 sccm, CH4=1 sccm and deposition time=8 hours by system 11.

POWER

163

4.4

Controlled Nitrogen Impurity Study By System II

Figure 4.14 displays the total growth rate versus nitrogen impurities with pressure

at 120 Torr, gas chemistry ﬁxed at Ar:H2:CH4=100:4:1, ft=105 sccm and deposition time

of 8 hours. As shown, the addition of nitrogen impurity decreases the growth rate approxi-

mately 38% as nitrogen impurity increases from a few ppm to 2500 ppm.

4

I

I

I

I

 

l

0.25

“0.2188

-0.1875 E?

8

E

-0.1563 9?

0

~0.l25

0

Q
o

40.0938 F

3

«0.0625 V

-0.0313

1) O

3.5

0

0

3-

o

0

1:

€11

E 2.5

v

<6

0:

5
g

2~

5 1.5*

o

<>

0

"a:

E“

1~

0.5-

 
  

1

1

1

0

1

1

0.00

0

500

1000

1500

2000

2500

N2 (1313“!)

FIGURE 4.34 : TOTAL GROWTH RATE VS. N2 IMPURITIES

Pressure=120 Torr, Ar:H2:CH4=100:4:1, ft=105 sccm

and deposition time=8 hours by system H.

164

Unfortunately, with the current set up of system II, the study of nitrogen impurity

on substrate temperature and absorbed microwave power can only be done from 0 to

1500ppm in one experiment. The nitrogen impurity pre-mixed gas tank will need to be

switched to a different tank to run higher nitrogen impurity experiments. To change the

gas tanks, the plasma will have to be shut down and the system will need to be pumped

down then restart the experiment. Therefore, the data presented here is only from 5 to

1500ppm of nitrogen impurity. The study of the inﬂuence of the nitrogen impurities on

substrate temperature and absorbed microwave power is shown in Figure 4.35. Figure

4.35(a) shows that nitrogen impurities (up to 1500ppm) have only a minor effect on sub-

strate temperature. Figure 4.35(b) shows that there is no changes in absorbed microwave

power as nitrogen impurities increase (up to 1500ppm). Nitrogen impurities have no effect

on absorbed microwave power and very little inﬂuence on substrate temperature.

165

 

660?

6 IN

V

 

W

a; 650-

I-

640 -

 
  

l

l

1

l

l

I

630

l

0

200

400

600

800

1000

1200

1400

1600

(a)

N2 (ppm)

 

950

I

I

I

I

I

I

I

 

2900

850

  
 

1

l

1

l

I

l

l

0

200

400

600

800

1000

1200

1400

1600

(b)

N2 (ppm)

FIGURE 4.35 : THE INFLUENCE OF N2 IMPURITIES ON Ts and Pabs

Pressure=120 Torr and Ar/HZ/CH4=100/4/l by system H.

166

Figure 4.14 displays the speciﬁc yield versus nitrogen impurities with pressure at

120 Torr, gas chemistry ﬁxed at Ar/Hz/CH42100/4/1 sccm and deposition time of 8 hours.

As shown, the addition of nitrogen impurities increase the speciﬁc yield approximately

68% as nitrogen impurity increases from a few ppm to 2500 ppm.

 500

I

I

I

I

I

450-

400 *

350 I

300

I

250<> <>

200 I

150*

)
g
/
r
h
*
W
k
(

d
l
e
i
Y

c
ﬁ
i
c
e
p
S

O

O

_

-

1

1

-

1

‘

100*

50*

 
  

*

*

0

500

1 000

1 500

2000

2500

N2 (1913"!)

FIGURE 4.36 : SPECIFIC YIELD VS. N2 HVIPURITIES

Pressure=120 Torr, Ar/Hz/CH4=100/4/1 sccm and deposition time=8 hours

by system H.

167

Figure 4.14 displays the carbon conversion efﬁciency versus nitrogen impurities

with pressure at 120 Torr, gas chemistry ﬁxed at Ar/HZ/CH4=100/4/l sccm and deposition

time of 8 hours. As shown, the addition of nitrogen impurity decreases the carbon conver-

sion efﬁciency approximately 38% as nitrogen impurity increases from a few ppm to 2500

1 24> O

I

 

I

I

I

I

0

10 *

<>

”a
E

3

8 *

0

O

0

O

0

o

_

-

-

*

_

 

5.5

35

L1]

a

8

E

8

c:

O

'8
Cd

0

6-

4 ’

2 *

 

O

l

J

l

I

l

 

0

500

1000

1 500

2000

2500

N2 (Wm)

FIGURE 4.37 : CARBON CONVERSION EFFICIENCY VS. N2 IMPURITIES

Pressure=120 Torr, Ar/H2/CH4=100/4/1 sccm and deposition time=8 hours

by system H.

168

4.5 Summary

Smooth ultra-nanocrystalline diamond ﬁlm synthesis was investigated with argon,

hydrogen, and methane gases by using MSU-MPACVD reactor systems I and H. The vari-

ables for reactor performance (Y1) of both systems are listed in Tables 4.1 and 4.2

respectively. We have learned that systems I and II behave in similar fashion as shown in

Figures 4.2 to 4.34. The results for both systems are summarized below.

(1) As hydrogen (and methane) concentration increases, the growth rate increases, the spe-

ciﬁc yield decreases, and the carbon conversion efﬁciency increases. Film growth rate

and the carbon conversion efﬁciency increases (the speciﬁc yield decreases) dramati-

cally with small increases in H2.

(2) Total gas ﬂow rate has no inﬂuence on growth rate and speciﬁc yield (It’s only true as

hydrogen to methane ratio is equal or greater than three for system I.) but a higher

total gas ﬂow rate reduces the carbon conversion efﬁciency.

(3) Higher pressure or absorbed microwave power increases growth rate and carbon con-

version efﬁciency and lower the speciﬁc yield.

(4) The experiments performed in systems I and H were both reproducible over a long

period of time. Even in system H, there were six users sharing the reactor and it did

not affect the reproducibility of system II. Several multiple data-point examples of

reproducibility of experiments are listed in Tables 4.3 (system I) and 4.4 (system 11).

169

TABLE 4.3: REPRODUCIBLE EXPERIMENT DATA FOR SYSTEM I

 

Ar

H2

CH4

f:

p

t

Ts

Pabs

Gt

Gave

8

C

(%)

(%)

(%)

(sccm)

(Torr)

(hr)

(0c)

(W)

(mg/hr)

(um/hr)

(kWh/g)

(%)

 

 

97.56

1.95

0.49

205

120

8

813

765

0.838

0.052

913

2.593

 

97.56

1.95

0.49

205

120

8

806

741

0.725

0.045

1022

2.243

 

97.56

1.95

0.49

615

120

8

830

821

0.750

0.047

1095

0.774

 

97.56

1.95

0.49

615

120

8

807

839

0.763

0.048

1100

0.787

 

97.56

1.95

0.49

615

120

4

810

821

0.700

0.044

1 173

0.722

 

97.56

1.95

0.49

615

120

4

840

858

0.675

0.042

1271

0.696

 

97.56

1.95

0.49

615

120

4

830

802

0.675

0.042

1 188

0.696

 

96.39

2.89

0.72

622.5

120

8

850

895

2.163

0.135

414

1.487

 

96.39

2.89

0.72

622.5

120

8

850

899

2.313

0.144

389

1.590

 

95.24

3.81

0.95

105

120

8

844

858

4.625

0.289

186

14.311

 

95.24

3.81

0.95

105

120

8

855

854

4.263

0.266

200

13.191

0.95

105

95.24

3.81

95.24

3.81

 

 

 
 
 
  

 

 

8

8

8

8

92.59

5.56

 

92.59

5.56

1.85

0.95

105

1.85

108

10.950

0.683

890

925

850

873

900

876

850

858

 

1 1.325

0.707

82

0.289

186

13.733

4.438

0.277

197

108

120

120

120

80

16.941

4.625

14.311

120

17.522

 

 

 

 

 

 

 

170

TABLE 4.4: REPRODUCIBLE EXPERIMENT DATA FOR SYSTEM II

 

Ar/H2/CH4

p

1

TS

Fab,

Gt

Gave

s

C

(sccm)

(Torr)

(hr)

(0C)

(W)

(mg/hr)

(um/hr)

(kWh/g)

(%)

 

 

100/4/1

120

8

626

945

3.800

0.237

249

1 1.758

 

100/4/1

120

8

650

973

3.663

0.228

266

1 1.335

 

100/4/1

120

16

629

965

3.881

0.242

249

12.009

 

100/4/1

120

16

642

965

3.825

0.239

252

1 1.836

 

100/4/1

160

8

709

1214

7.913

0.494

153

24.485

 

100/4/1

160

8

697

1223

8.238

0.514

148

25.491

 

100/2/1

160

8

696

l 147

8.063

0.503

142

24.950

 

100/2/1

160

8

706

1160

7.575

0.473

153

23.440

 

100/1/1

120

8

594

890

2.375

0.148

375

7.349

 

100/1/1

120

8

601

894

2.238

0.140

399

6.925

100/1/1

200

100/1/1

 

 

  
 
 

 

1220

6.038

1314

8.838

1227

5.300

1251

8.825

 

8

8

8

8

0.377

202

0.331

232

0.551

142

0.551

149

27.348

100/0/1

200

100/0/ 1

200

673

677

611

604

 

27.307

16.400

18.684

 

 

200

 

 

 

 

There are some distinctive differences between the two systems. These are:

(1) For System I, the plasma can not be stabilized and/or ﬂame free if hydrogen to meth-

ane ratio is under three because of the impurity in the system. But system H can run

the experiments with hydrogen to methane ratio approaches to zero.

(2) The reactor performance: growth rate, speciﬁc yield, and carbon conversion efﬁciency

are affected by deposition time about 20% to 25% between 8 and 24 hours for the

ﬁlms produced by system I but are not inﬂuenced by deposition time between 8 and

171

25 hours for the ﬁlms produced by system 11. The impurity in the system I seems to

affect the growth mechanism of ultra-nanocrystalline diamond.

(3) A quantitative comparison of the reactor performance of the two reactors with

Ar/Hz/CH4=100/4/l sccm, and deposition time of 8 hours is listed in table 4.3.

TABLE 4.5: QUANTITATIVE COMPARISON OF THE REACTOR

PERFORMANCE OF SYSTEMS I AND II

 

p

PabsI

PabsII

GTI

GTII

GaveI

GaveII

SyI

SyII

CI

CH

 

120

861

945

4.488

3.800

0.280

0.237

192

249

13.887

1 1.758

 

140

956

1053

7.163

6.413

0.447

0.400

133

164

22.165

19.844

 
  
 

 

10.788

0.494

153

33.382

24.485

160

959

1214

7.913

0.673

89

 

 

 

 

 

 

 

 

 

The ratios of microwave power of system I to system II at 120 Torr, 140 Torr, and

160 Torr are as follows:

I

(Pabs)

= 0.911,

II

(Pabs)

I

(P

abs)

 

II

(P

abs)

= 0.908, and

I

(P

abs)

 

II

(p

abs)

= 0.790.

Surprisingly, the microwave power of system I is lower than the microwave power

of system H. This tells us that system I is more energy efﬁcient than system 11. This can be

caused by a different quartz dome height or a difference in the size of the cooling base-

plate of the two microwave cavity plasma reactors or substrate temperature. To improve

the energy efﬁciency of system 11, the operator can (1) use the same size quartz dome as

system I, and (2) increase the temperature set point of the Neslab chiller of system H from

15°C to a higher degree (lowering the cooling efﬁciency).

172

The addition of nitrogen impurity decreases the growth rate and carbon conversion

efﬁciency approximately 38%, and increases the speciﬁc yield approximately 68% as

nitrogen impurity increases from a few ppm to 2500 ppm.

173

Microwave Plasma Assisted Chemical Vapor Deposition of

Ultra-nanocrystalline Diamond Films

VOLUME 11

By

Wen-Shin Huang

A DISSERTATION

Submitted to

Michigan State University

in partial fulﬁllment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Electrical and Computer Engineering

2004

5

MSU-MPACVD Reactor Experimental Output

Variables-Film Characteristics (Y2)

5.1

Introduction

Chapter 4 described the ultra-nanocrystalline diamond ﬁlm synthesis performance

(Y1) for systems I and II. The results of the experimental investigation of the ﬁlm charac-

teristics (Y2) are presented in this chapter. The parameter space investigated in this chapter

is the same as in Chapter 4 and is illustrated in Figure 4.1. The input (U), internal (X), and

output (Y) variables considered and their associated experimental range explored are dis-

played in Table 5.1 and 5.2 (Input and internal variables are the same as Table 4.1 and

4.2.). In system I, the residence time inside (1) the quartz dome is 9.31-77.74 seconds, and

(2) the system is 11.82-98.72 minutes. In system H, the residence time inside (1) the

quartz dome is 9.49-113.83 seconds, and (2) the system is 1305-15659 minutes. In this

chapter, an overview of the MSU-MPACVD ﬁlm characteristics (Y2) as a function of a

selected number of input variables is presented: Ar/H2/CH4 gas chemistry, total gas ﬂow

rate ft, deposition pressure, p, and deposition time, t. Output variables-ﬁlm characteristics

(Y2) consist of ﬁlm roughness, morphology, structure quality, ﬁlm texture and crystal size.

Section 5.2 and Section 5.3 describe the inﬂuence of gas chemistry, deposition time, total

gas ﬂow rate, and deposition pressure on the ﬁlm characteristics of system I and H respec-

tively. The controlled nitrogen impurity investigation using system H is presented in

Section 5.4. In Section 5.5, the summary of the ﬁlm characterization results of systems I

and H discussed in this chapter is presented in Section 5.4.

174

TABLE 5.1: PARAMETER SPACE FOR EXPERIMENTS PERFORMED ON SYSTEM I

 

Controllable

a) Gas Chemistry, Ar/H2/CH4=(99-91)/(1-7)/(0-2) (%)

Input

Variables

U1

b) Total Gas Flow Rate, ft=102.5-642 sccm

c) Deposition Pressure, p=120-160 Torr

d) Absorbed Microwave Power, Pabs=691-1221W

 

a) 2.45 GHz Microwave Power, ﬁxed

b) Reactor Conﬁguration, MPCR, ﬁxed

c) End Feed Excitation, ﬁxed

d) Electromagnetic Mode and Cavity Tuning, Fixed

at TM013

Ian

Reactor Design

e) Reactor Wall Material, quartz dome, ﬁxed, dia.=5”

Variables

Variables

f) Substrate Holder Set-up, molybdenum holder,

U

U2

thermally ﬂoating stage

g) Vacuum System, mechanical pump, 1~2x10'2 Torr,

leak rate: 12.6 mTorr/hr

h) Purity of gases, ﬁxed, See Section 3.2.2.1

 

Deposition

a) Deposition Time, 2-48 hours

Process

b) Substrate Material and Size, 3” Si (100) wafers

Variables

c) Seeding Procedure, mechanically scratched by

U3

0.1 um natural diamond powder

(1) Start-up and Shut-down, ﬁxed

 

Internal

Variables

X

top side)

Substrate Temperature, TS=750-950 °C (by pyrometer,

 
  
 

  (1) Crystal size: 30-2700 nm

Characteristics

b) Film roughness: 14-149 nm

Film

a) Film texture: I(220)/I(111)=0.06-4.04

Y2

c) Morphology: mostly (111)

 

175

TABLE 5.2: PARAMETER SPACE FOR EXPERIMENTS PERFORMED ON SYSTEM 11

 

a) Gas Chemistry, Ar/Hz/CH4=(99-90)/(0-9)/1 (%)

Controllable

b) Total Gas Flow Rate, ft=101-630 sccm

Input

c) Deposition Pressure, p=60-240 Torr

vanables

d) Absorbed Microwave Power, Pabs=732-1518W

 

U1

e) Controlled N2 impurity, 5-2500ppm

a) 2.45 GHz Microwave Power, ﬁxed

b) Reactor Conﬁguration, MPCR, ﬁxed

c) End Feed Excitation, ﬁxed

(1) EM Mode and Cavity Tuning, Fixed at TM013

Input

Reactor Design

e) Reactor Wall Material, quartz dome, ﬁxed, dia.=5”

Variables

Variables

f) Substrate Holder Set-up, molybdenum holder,

U

U2

thermally ﬂoating stage

g) Vacuum System, turbo pump, 10'7~10'6 Torr, leak

rate*: 4mTorr/hr

h) Purity of gases, ﬁxed, See Section 3.2.2.2

 

Deposition

a) Deposition Time, 8-25 hours

Process

b) Substrate Material and Size, 3” Si (100) wafers

Variables

c) Seeding Procedure, mechanically scratched by

U3

0.1 urn natural diamond powder

d) Start-up and Shut-down, ﬁxed

 

Internal

X

Variables

back side)

Substrate Temperature, TS=UFL-770 °C (by pyrometer,

  
 
  

a) Film texture: most (220); I(220)/I(111)=0.13-9.31

Characteristics

b) Film roughness: “-50 nm

Y;

c) Morphology: featureless

 

(1) Crystal size: 3-30 nm

Film

 

UFL: Temperature is too low to be detected by the pyrometer.

176

5.2

System I

5.2.1 Film Texture = g(Ar/Hz/CH4, t, ft, p)

Figure 5.1 displays the XRD (X-ray Diffraction) ratio of integrated intensities of

diamond <220> and <111> peaks plotted as a function of hydrogen concentration (the

percentage of hydrogen ﬂow rate to the total gas ﬂow rate). These experiments were per-

formed at a constant pressure of 120 Torr and deposition time of 8 hours. The argon

concentration was varied from 90.91-98.68%, the H2 concentration varied from

0.99-7.27%, and the CH4 concentration varied from 0.33—1.85% with the total gas ﬂow

rate varied from 102.5-642 sccm. According to ASTM (American Society for Testing

Materials) and JCPDS (Joint Committee on Powder Diffraction Standard) - International

Center for Diffraction Data, using diamond powder sample with no preferred orientation,

the XRD ratio of intensities of diamond <220> and <111> peaks is 0.25. (For statistics

reasons, <111> and <220> peaks are chosen.) A ratio larger than 0.25 means the ﬁlm has

(220) preferred orientation. A ratio less than 0.25 indicates the ﬁlm has (111) preferred

orientation.

As hydrogen concentration rises above 4%, the ﬁlms clearly have (220) preferred

orientation. Generally, the ﬁlms synthesized with higher hydrogen concentration (or

higher methane concentration) have the (220) preferred orientation. At lower hydrogen

concentration (below 4%), the value of I(220)/I(111) scatters between 0 and 0.47. The ﬁlm

corresponding to a zero ratio is too thin (average ﬁlm thickness=81 nm) and is masked

under the large Si(400) peak.

177

 

 

0

1*

CH4~0.5°0

CH4~0.95°/o

  
 

CH4~1.85°/o

O

4.5 *

4 *

3.5 *

3 *

2.5 —

2 *

1.5 *

)
l

l

1
(
I
/
)
0
2
2
(
I

8

O

,

5

-

_

-

-

-

O

1 _

0.5 5

  
 

F 1

3r

13:

('31

fr

8

ﬁ

ﬁ

_

-

fr

1

1

1

\

O

1

1

0

1

2

3

4

5

6

7

8

H2 (%)

FIGURE 5.1 : I(220)/I(111) VS. HYDROGEN CONCENTRATION

Pressure=120 Torr, deposition time=8 hours,

Ar:H2:CH4=(90.91-98.68):(0.99-7.27):(0.33-l .85)%

and total gas ﬂow rate varied from 102.5-642 sccm

by system I.

178

Figure 5.2 displays the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of the deposition time, t, with pressure at 120 Torr

Ar:H2:CH4 =100:2:0.5, and ft=615 sccm. Using this deposition condition, the ﬁlms have

the (111) preferred orientation. The ﬁlm corresponding to zero ratio is too thin (average

ﬁlm thickness=87 nm) and is masked under the big Si(400) peak.

0.2

I

I

I

I

 

0.18*

O

0.16

I

0.14

I

0.12

I

0.1

I

0.08

I

0.06

I

)
1

l

1
(
I
/
)
0
2
2
(
I

_

l

l

O

1

l

1

l

0.04

0.02

 O

I

I

1‘)

I

l

1

l

1

l

 

 

O

1 O

20

30

40

50

Time (hours)

FIGURE 5.2 : I(220)/I(111) VS. DEPOSITION TIME

Pressure=120 Torr, Ar:H2:CH4=100:2:0.5, and ft=615 sccm by system I

179

Figure 5.3 displays the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of total gas ﬂow rate, ft, with pressure at 140 Torr and

deposition time of 8 hours. Only the ﬁlms synthesized with Ar:H2:CH4=100:2:0.5 gas

chemistry ratio have the (111) preferred orientation. The ﬁlms synthesized with the gas

chemistry ratios of Ar:H2:CH4=100:3:1 and Ar:H2:CH4=100:4:1 have the (220) preferred

orientation. At higher total gas ﬂow rates, the ﬁlms’ (220) preferred orientation tends to be

weaker.

 

1.8

I

o

 

O

9

Ar:H2:CH4=100:4:1

Ar:H2:CH4=100:3:1

_

 
  

Ar:H2:CH4=100:2:0.5

:»

1.6

I

1.4”

1.2*

)
1

l

1
(
I
/
)
0
2
2
(
I

0.6

I

w

0

l

4

5

~

0.4

I

0.2 I

1 
  

D
"

O

1

0

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 5.3 : I(220)/I(111) VS. TOTAL GAS FLOW RATE

Pressure=140 Torr and deposition time=8 hours by system I

180

Figure 5.4 display the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of deposition pressure, p, with deposition time of 8

hours and ft: 102.5-630 sccm. Between the deposition pressure of 120 to 160 Torr, only

the ﬁlms synthesized with Ar:H2:CH4=100:2:0.5 gas chemistry ratio have the (111) pre-

ferred orientation. As the deposition pressure equal to or greater than 140 Torr, the ﬁlms

synthesized with the gas chemistry ratios of Ar:H2:CH4=100:3:1 and Ar:H2:CH4=100:4:1

mostly have the (220) preferred orientation; I(220)/I(111)=0.34-3.62. As the deposition

pressure equal to 120 Torr and the ﬁlms were synthesized with the gas chemistry ratios of

Ar:H2:CH4=100:3:1 and Ar:H2:CH4=100:4:1, the ratio of I(220)/I(111) scatters between

0.08-0.47. Figure 5.4 shows that the increase of the deposition pressure has more effect on

the

ﬁlms grown

with

the

gas

chemistry

ratios

of Ar:H2:CH4=100:3:1

and

Ar:H2:CH4=100:4:1 than those grown with Ar:H2:CH4=100:2:0.5 gas chemistry ratio.

The effect caused an increase in the (220) preferred orientation for the ﬁlms synthesized

with the gas chemistry ratios of Ar:H2:CH4=100:3:1 and Ar:H2:CH4=100:4:1.

181

 

4'5 _

4 -

3.5 *

3

2.5 *

2 -

1.5 *

)
1

l

1
(
I
/
)
0
2
2
(
I

 

O

Ar:H2:CH4=100:4:1

8

AIZHZICH4=1003321

n

 
  

Ar:H2:CH4=100:2:0.5

"

r?

O

-

~

-

-

—

*

o

O

it

  
 

8
g

1

g

1

-

_

.

1 ~

0.5 7

O

.

.

110

120

130

140

150

160

170

Pressure (Torr)

FIGURE 5.4 : I(220)/I(111) VS. DEPOSITION PRESSURE

Deposition time=8 hours by system I

182

5.2.2 Film Roughness = g(Ar/Hz/CH4, t, ft, p)

Figure 5.5 displays the ﬁlm roughness plotted as a function of hydrogen concentra-

tion, (the percentage of argon, hydrogen, and methane gases), with a constant pressure of

120 Torr, Ar:H2:CH4 =100:2:0.5, Ar:H2:CH4 =100:3:1, Ar:H2:CH4 =100:4:l, and with (a)

ft=102.5-105 sccm, (b) ft=615-630 sccm, and deposition time of 8 hours. The ﬁlm rough-

ness increases with the increases in hydrogen concentration.

150

r

,

,

1

 

 

0

1"

 

ft=102.5—105 sccm

ft=615—630 sccm

 

 

A 100 *

<>

0

~

0

E

5
(I:

8

t:

..C‘.

6.0

:3

O
a:

50 *

 

0

1

1

1

L

-

 

 

1 5

2

2.5

3

3 5

4

H2 (%)

FIGURE 5.5 : RMS ROUGHNESS VS. HYDROGEN CONCENTRATION

Pressure=120 Torr and deposition time=8 hours by system I

183

Figure 5.6 displays the ﬁlm roughness plotted as a function of different deposition

times, with the pressure of 120 Torr, Ar:H2:CH4 =100:2:0.5, and ft=615 sccm. The ﬁlm

roughness increases as the deposition time increases.

1 50

I

I

I

I

Q

 

100 *

E

5

E

DO

3
o

M

50 *

 

0

O

o 0

0

g

L

l

l

7

7

 

 

O

1 O

20

30

40

50

0.087 0. 1 68 0.381

1.422

3.007

ave. thickness (um)

Time (hour)

FIGURE 5.6 : RMS ROUGHNESS VS. DEPOSITION TIME

Pressure=120 Torr, Ar:H2:CH4=100:2:0.5, ft=615 sccm by system I.

184

Figure 5.7 displays the ﬁlm roughness plotted as a function of different deposition

pressure, with Ar:H2:CH4 =100:2:0.5, ft=615 sccm and a deposition time of 8 hours. The

ﬁlm roughness increases as the deposition pressure increases.

60

1

I

.

.

.

 

O

O

50—

40—

’8
5

8 30-

9

.2

DD

3

0

°‘

120

10~

 

0

1

l

l

I

1

.

~

..

-+

«

 

 

110

120

130

140

150

160

170

Pressure (Torr)

FIGURE 5.7 : RMS ROUGHNESS VS. DEPOSITION PRESSURE

Ar:H2:CH4=100:2:0.5, ft=615 sccm and deposition time=8 hours

by system I.

185

5.2.3 Film Quality = g(t, p)

Figure 5.8 displays the Raman Spectra for different deposition times, with the

pressure of 120 Torr, Ar:H2:CH4 =100:2:0.5, and ft=615 sccm. The diamond peak at 1332

cm'1 becomes sharper as deposition time becomes longer. This may be caused by the fact

that crystal size becomes larger as time increases.

 

I

I

I

I

I

I

I

:5

<55

2.5W

'

8 hours

.

2 b

1 .5 *

24 hours I

*

1 '

7
   

48 hours

0

            

41

l

l

l

I

   

l

l

   

 

1200

1250

1 300

1 350

1400

1450

1 500

1 550

1 600

Shift Wavenumber

FIGURE 5.8 : RAMAN SPECTRA

Pressure=120 Torr, Ar:H2:CH4=100:2:0.5, and ft=615 sccm by system I.

Spectra Taken by Raman System (Ar 514.5nm Laser) at MSU, ECE

186

 

5

4.5

4 ”

3.5

3 *

 

0.5 *

2 hours

4 hours

*

‘

1

_

'I

 

Figure 5.9 displays the Raman Spectra for different deposition pressures, with

Ar/Hz/CH4 =600/12/3 sccm and deposition time of 8 hours. The diamond peak at 1332

cm'1 becomes sharper as deposition pressure becomes larger. This may be caused by the

fact that crystal size increases as pressure increases.

3

 

I

I

I

I

I

I

I

2.5 *

2 _

=5
<6

1.5 *

120 Torr

140 Torr 7

‘

1 h

160 Torr ‘

0.5 * 

O

 

 

I

I

I

I

L

1

I

1200

1 250

1 300

1 350

1 400

1 450

1 500

1 550

1 600

Shift Wavenumber

FIGURE 5.9 : RAMAN SPECTRA

Ar:H2:CH4=100:2:0.5, ft=615 sccm and deposition time=8 hours

by system I.

Spectra Taken by Raman System (Ar 514.5nm Laser) at MSU, ECE.

187

5.2.4 Morphology and Crystal Size: g(Ar/H2/CH4, t, ft, p)

The AFM probes used in this research were TESP (TappingMode Etched Silicon

Probe) which had nominal tip radius of curvature of 5 -10 nm. For this thesis research, 5

pm x 5 pm images were acquired with the number of samples parameter set to 512. Then

the pixel size is 9.8 nm (5 urn/512:0.0098 um=9.8 11m). Thus, features smaller than 9.8

nm can not be resolved. As AFM scan is performing, the tip becomes dull and the radius

of curvature increases. A dull tip with a larger radius of curvature will not be able to later-

ally resolve as small a feature as a sharp tip does. Therefore, the resolution decreases as

the radius of curvature increases because of the scanning. Thus, an AFM is not a good tool

to determine the crystal size when the crystal size is small because it can not distinguish

the individual crystal’s boundary. But it’s an excellent tool to examine the ﬁlm morphol-

ogy without ﬁlm destruction and estimate the crystal size when the crystal is large enough.

Although the crystal sizes of the ﬁlms deposited by system I are larger than desired, it

offers an opportunity to study the relationship between the crystal size and the input vari-

ables for system I.

Figure 5.10 displays

the ﬁlm morphology for different gas chemistries,

Ar/HZ/CH4, (the percentage of argon, hydrogen, and methane gases), with the pressure of

120 Torr and deposition time of 8 hours. Figure 5.10(b) and Figure 5.10(c) have (111) tri-

angular morphology.

Figure 5.10(a) (Ar:H2:CH4=100:2:0.5 and f =102.5 sccm) shows that the crystal

sizes range from roughly 30 nm to 1 pm. In Figure 5.10(b) (Ar:H2:CH4=100:3:1 and

ft=104 sccm), the crystal sizes range from roughly 100 nm to 1.5 pm. In Figure 5.10(c),

188

(Ar:H2:CH4=100:4:1 and ft=105 sccm) the crystal sizes range from roughly 100 nm to 1.7

pm. With the pressure of 120 Torr, ft=102.5-105 sccm and deposition time of 8 hours,

increasing hydrogen concentration from 1.95% to 3.81% increases the low end of crystal

size around 3.3 times and 1.7 times for the high end.

189

Figure 510(a)

  

Digital instruments NanoScope

Scan 51 ze

‘

Scan rate

.000 pm

5

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

750.0 nm

“m

x

2

1.000 um/div

750.000 nm/div

ar9035.000

Figure 5.10(b)

gantghinstrumts "lawn”.

  

luage Data

Height

Data scale

1.

un

Scan rate

1.001 H2

Nuuber of sauplcs

256

  
    

X

1.000 uu/div

2

1500.000 nu/diu

vu

argo36.002

FIGURE 5.10 : FILM MORPHOLOGY VS. GAS CHEMISTRY

Pressure=120 Torr and deposition time=8 hours by system I.

190

Figure 5.10(c)

DigitalzInstruments Nanoscope

i3"?.3.

M000
1. 001 Hz

Nunber of samples

6

luage Data

Height

Data scale

1.500 uu

   

X

1.000 uu/div

uu

ar9057.000

2

1500.001 nu/div

Figure 5.10(a): Ar:H2:CH4=100:2:0.5 and ft=102.5 sccm

Figure 5.10(b): Ar:H2:CH4=100:3:1 and ft=104 sccm

Figure 5.10(c): Ar:H2:CH4=100:4:1 and f,=105 sccm

FIGURE 5.10 (cont’d)

191

Figure 5.11 displays the ﬁlm morphology for different gas chemistries, i.e. for dif-

ferent percentages of argon, hydrogen, and methane gases, with the pressure of 120 Torr

and deposition time of 8 hours. Figure 5.11(a), Figure 5.11(b) and Figure 5.10(c) have

(111) triangular morphology. In Figure 5.11(a) (Ar:H2:CH4=100:2:0.5 and ft=615 sccm),

the

crystal

sizes range from roughly 80 nm to 600 nm. In Figure 5.11(b)

(Ar:H2:CH4=100:3:1 and ft=624 sccm), the crystal sizes range from roughly 120 nm to 1

pm. In Figure 5.11(c) (Ar:H2:CH4=100:4:1 and ft=630 sccm), the crystal sizes range from

roughly 120 nm to 1.3 um. With the pressure of 120 Torr, ft=615-630 sccm and deposition

time of 8 hours, by increasing the hydrogen concentration from 1.95% to 3.81% increases

the low end of crystal size 1.5 times and almost 2.2 times for the high end.

192

ﬁgure 511(3)

Digital lnstruuents Nanoscope

32:2 2::

'

5

233‘: a

Number of sauples

    

lnage Data

Height

Data scale

350.0 nu

   

X

1.000 un/div

2

350.000 nn/div

nu

ar5012.002

'

Flgure 5'11(b)

Di ital Instruments NanoSco e

chn si 29

Scan rate

. OOODum

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

800.0 nm

1"“

x

2

1.000 urn/div

800.000 nm/div

ar9101.000

FIGURE 5.11 : FILM MORPHOLOGY VS. GAS CHEMISTRY

Pressure=120 Torr and deposition time=8 hours by system I.

 

193

Figure 5.1 I(C)

  

Digital Instruments NanoSCope

Scan size

Scan rate

5

.000 pm

1.001 H2

Number of samples

256

Image Data

Height

Data scale

1.800 um

  

W"

x

2

1.000 urn/div

1800.000 nm/div

ar9082.001

Figure 5.10(a): Ar:H2:CH4=100:2:0.5 and ft=615 sccm

Figure 5.10(b): Ar:H2:CH4=100:3:1 and ft=624 sccm

Figure 5.10(c): Ar:H2:CH4=100:4:1 and ft=630 sccm

FIGURE 5.11 (cont’d)

194

Figure 5.12 displays the ﬁlm morphology of different deposition times, with the

pressure of 120 Torr, Ar:H2:CH4 =100:2:0.5, and ft=615 sccm. Figures 5.12(a), 5.12(b),

and 5.12(c) have (111) triangular morphology. Figure 5.12(d) and Figure 5.12(e) have

mixed morphology. In Figure 5.12(a) (deposition time = 2 hours), the crystal sizes range

from 30 nm to 120 nm roughly. In Figure 5.12(b) (deposition time = 4 hours), the crystal

sizes range from 50 nm to 200 nm roughly. In Figure 5.12(c) (deposition time = 8 hours),

the crystal sizes range from 80 nm to 600 nm roughly. Figure 5.12(d) (deposition time =

24 hours), the crystal sizes range from 140 nm to 1.4 pm roughly. In Figure 5.12(e) (depo-

sition time = 48 hours), the crystal sizes range from 200 nm to 2.7 mm roughly. With the

pressure ﬁxed at 120 Torr, Ar:H2:CH4 =100:2:0.5, and ft=615 sccm, and by varying the

deposition time from 2 hours to 48 hours increases the low end of the crystal size almost 7

times and 22.5 times for the high end.

195

    

Figure 5.12(a)

Digital Instruments Nanoscope

gggn :3:

Number of samples

Image Dat

1. 000 um

1.001 H2

256

lght

Data scale

150.0 nm

0.8 \ \L

K_\

‘

"‘8

/’

V"

x

2

0.200 um/diu

150.000 nm/diu

arg022.001

  

Scan rat

Image Data

Data scale

Number of samples

1.

Height

400.0 nm

Figure 5.12(b)

ggitalzeInstruments "02°38‘01"”.

”"

X

2

1.000 um/div

400.000 nm/div

ar9026.001

FIGURE 5.12 : FILM MORPHOLOGY VS. DEPOSITION TIME

Pressure=120 Torr, Ar:H2:CH4=100:2:0.5, and ft=615 sccm by system I.

196

I?

5 12

Digital Instruments Nagogggpe

lgure

'

(C)

Scan size

Scan rate

Number of samples

Image Data

     

Data scale

1.001 112

258

Ight

350.0 nm

  

350.000 nm/div

1.000 um/div

um

ar9012.002

Flgure 5,12(d)

glgi tglz.Instruments Nari-353813;”

   

Image Dat

Height

Data scale

1.000 um

Scan rat:1.001

Number of asamples

256

  

\I

3\

\

um

arg014.001

x

2

2.000 Inn/div

1000.000 nm/div

FIGURE 5.12 (cont’d)

197

 

.

Flgure 5.12(e)

Digital Instruments NanoScope

Scan size

Scan rate

10.00 pm

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

900.0 nm

  

1"“

X

2

2.000 urn/div

900.000 nm/div

ar9017.000

Deposition Time & AFM Scan Size:

Figure 5.12(a): 2 hours; 1x1 umz

Figure 5.12(b): 4 hours; 5x5 umz

Figure 5.12(c): 8 hours; 5x5 umz

Figure 5.12(d): 24 hours; 10x10 umz

Figure 5.12(e): 48 hours; 10x10 umz

FIGURE 5.12 (cont’d)

198

  

Figure 5.13 displays the ﬁlm morphology versus different deposition pressures,

with Ar:H2:CH4 =100:2:0.5, and ft=615 sccm and deposition time of 8 hours. Figure

5.13(a), Figure 5.13(b) and Figure 5.13(c) have mostly (111) triangular morphology. In

Figure 5.13(a) (deposition pressure = 120 Torr), the crystal sizes range from roughly 80

nm to 600 nm. In Figure 5.13(b) (deposition pressure = 140 Torr), the crystal sizes range

from roughly 100 nm to 700 nm. In Figure 5.13(c) (deposition pressure = 160 Torr), the

crystal sizes range from roughly 100 nm to 1.2 pm. With Ar:H2:CH4 =100:2:0.5, ft=615

sccm and deposition time of 8 hours, by varying the deposition pressure from 120 Torr to

160 Torr increases the low end of the crystal size 1.25 times and almost 2 times for the

high end.

199

'

Figure 543(8)

Digital Instruments Nanoscope

33:2 2::

m a

Number of samples

255

Image Data

Height

   

Data scale

350.0 nm

X

1.000 um/diu

”I

2

350.000 nm/div

   

arg012.002

Scan size

Scan rate

Image Data

Data scale

.

Frgure 5.13(b)

Digital Instruments NanoScope

Number of samples

5.000 urn

1 001 Hz

256

Height

500.0 nm

X

2

1.000 pm/div

500.000 nm/div

ar9027.001

FIGURE 5.13 : FH3M MORPHOLOGY VS. DEPOSITION PRESSURE

Deposition Pressure: (a): 120 Torr, (b): 140 Torr, (c): 160 Torr

Ar:H2:CH4=100:2:0.5, ft=615 sccm and deposition time=8 hours

by system I.

200

 

500.0 nm  

Number of samples

Data scale

Image Data

Scan size

Scan rate

Digital Instruments NanoScope

Figure 5.13(C)

5.000 urn

1 001 Hz

256

Height

”m

x

2

1.000 um/div

500.000 rim/div

arg032.000

FIGURE 5.13 (cont’d)

201

5.3

System 11

5.3.1 Film Texture = g(A r/Hz/CH4, ft, p)

Figure 5.14 displays the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of the hydrogen ﬂow rate, with argon ﬂow ﬁxed at 100

sccm, methane ﬂow ﬁxed at 1 sccm, pressure of 120 Torr and deposition time of 8 hours.

The XRD ratios of intensities of diamond <220> and <111> peaks are all above 0.25.

Therefore, all the ﬁlms have the (220) preferred orientation.

3

I

I

I

 

I

I

I

2.5 *

2 *

5:

O

0

<>

o

E“ 1.5 —
5

o

0

N
(3

1 _

o

T

_

~

_

_

 

0.5 *

 

O

l

I

I

I

I

1

 

O

2

4

6

8

10

H2 (sccm)

FIGURE 5.14 : I(220)/I(111) VS. H2 FLOW RATE

Pressure=120 Torr, Ar=100 sccm, CH4=1 sccm and deposition time=8 hours

by system H

202

Figure 5.15 displays the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of total gas ﬂow rate, ft, with pressure of 120 Torr and

deposition time of 8 hours. Only the ﬁlms synthesized with Ar:H2:CH4=100:4:1 gas

chemistry ratio and total gas ﬂow rates equal to 400 sccm or 600 sccm have the (111) pre-

ferred

orientation. The ﬁlms synthesized with the gas chemistry ratios of (a)

Ar:H2:CH4=100:2:1; and (b) Ar:H2:CH4=100:4:1 and total gas ﬂow rate equal to 100 or

200 all have the (220) preferred orientation. At higher total gas ﬂow rates, the ﬁlms’ (220)

preferred orientation tends to be weaker.

 

5

I

I

I

I

L

I

 

o

Ar:H2:CH4=100:4:1 sccm

Ar:H2:CH4=100:2:1 sccm _
 
  

o

4.5-

3.5 l

2.5

l

1.5-

)
l

1

1
(
I
/
)
0
2
2
(
I

 

J

l

-

I

 

0.5

I

O

O

O

 

O

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 5.15 : I(220)/I(111) VS. TOTAL GAS FLOW RATE

Pressure=120 Torr and deposition time=8 hours by system 11

203

Figure 5.16 display the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of deposition pressure, with argon ﬂow equal to 100

sccm, methane ﬂow equal to 1 seem and deposition time of 8 hours. The ﬁlm corresponds

to zero ratio is too thin and is masked under the big Si(400) peak. The XRD ratios of inte-

grated intensities of diamond <220> and <111> peaks usually have a maximum around

140 Torr or 160 Torr except when the hydrogen ﬂow equal to 1 sccm.

204

)
l

1
1
(
I
/
)
0
2
2
(
I

)
1

1

1
(
I
/
)
0
2
2
(
I

)
1

l

l
(
I
/
)
0
2
2
(
I

)
l

l

1
(
I
/
)
0
2
2
(
I

 1O

E O

H2=4 sccm

 

5 —

o

O

O

a

O

O

 
  

J)

C

O

O

9

I

50

1 00

1 50

200

250

1

I

I

1 0

I

 

1

 

5 )—

*

a:

I

w

H2=2 sccm I

 

*

*

‘1

*

 
  

I

1*

a

ﬁ'

I

I

50

1 00

1 50

200

250

1 O

r

 

1

 

5 *

I

«a

H2=1 sccm I

 

_

 
  

50

1 00

1 50

200

250

 

1 O

l

l

V

v

I

 

I

v

H2=O sccnﬂ

 

5 ”

V

V

_

 
  

V

50 V

V

160

150

200

250

Pressure (Torr)

FIGURE 5.16 : I(220)/I(l l 1) VS. DEPOSITION PRESSURE

Ar/CH4=100/1 sccm and deposition time=8 hours by system II

205

5.3.2 Film Roughness = g(Ar/Hz/Clh, t, ft, p)

Figure 5.17 displays the ﬁlm roughness plotted as a function of hydrogen ﬂow rate,

with argon ﬂow ﬁxed at 100 sccm, methane ﬂow ﬁxed at 1 seem, pressure at 120 Torr and

deposition time of 8 hours. The ﬁlm roughness increases dramatically with the increases

in hydrogen concentration.

 

50

l

T

l

I

45~

40—

35-

’a‘ 30~

5

g 25”

.C.‘

Di)

8 20-
a:

15-

O

O

0

O

O

I

<>

—

J

-

.

‘

_

1

1042

 

5-

O

l

l

l

l

l

a

a

 

 

O

2

4

6

8

10

H2(sccm)

FIGURE 5.17 : RMS ROUGHNESS VS. H2 FLOW RATE

Pressure=120 Torr, Ar=lOO sccm, CH4=1 sccm and deposition time=8 hours

by system 11

206

Figure 5.18 displays the ﬁlm roughness plotted as a function of the deposition

time, t, with pressure of 160 Torr and Ar/Hz/CH4 =lOO/4/l sccm. As shown, the ﬁlm

roughness increases as deposition time increases.

 

T

T

l

l

l

:;;

50

l

l

r

45.

40

l

15

1

)
m
n
(

s
s
e
n
h
g
u
o
R

—

l

L

L

 

10

 0

l

l

I

J

I

I

I

I

I

l

 

8

1 0

12

14

1 6

1 8

20

22

24

26

Time (hours)

FIGURE 5.18 : RMS ROUGHNESS VS. DEPOSITION TIME

Pressure=160 Torr, Ar/HZ/CH4=100/4/l sccm by System H

207

Figure 5.19 displays the ﬁlm roughness plotted as a function of total gas ﬂow rate,

ft, with pressure of 120 Torr and deposition time of 8 hours. the data indicate that the ﬁlm

roughness increases as total gas ﬂow rate increases.

100

l

l

I

 

 

o

Ar:H2:CH4=100:4:1 Roughness

Ar:H2:CH4=100:2:1 Roughness _ 
90  

 

l

80

I

70

l

60

l

50

l

40-

30

I

)
m
n
(

s
s
e
n
h
g
u
o
R

o

O

I

I

I

I

~

I

:20 __

10

l

4

 

A

I

‘.

,'»,

(

,1

..I

I

 

 

O

1 00

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 5.19 : RMS ROUGHNESS VS. TOTAL GAS FLOW RATE

Pressure=120 Torr and deposition time=8 hours by System II

208

Figure 5.20 display the ﬁlm roughness plotted as a function of deposition pressure,

p, with argon ﬂow constant at 100 sccm, methane ﬂow equal to 1 sccm and deposition

time of 8 hours. Figure 5 .20 shows that between 100-120 Torr, the ﬁlm roughness achieves

minimum. As pressure increases beyond 160 Torr, the ﬁlm roughness increases. Between

60 to 100 torr, usually, the ﬁlm roughness decreases as the deposition pressure increases.

The smoothest ﬁlms have a surface roughness of 10-15 nm between 100-160 Torr.

209

 50

 

o

t

0

H2=4 sccm

H2=2 sccm

H2=1 sccm

 
  

H2=O sccm

v

45 -

4O _

35 —

30 ~

O

g

25 - Z

20 '-

15 —

O

O

V

{‘2

o

t

V

)
m
n
(

s
s
e
n
h
g
u
o
R

O

_

-

a

o

‘

-

~

_

.

O

*

V

n

 
  

l

-

_

l

l

10 —

5 ~

0

50

100

150

200

250

Pressure (Torr)

FIGURE 5.20 : RMS ROUGHNESS VS. DEPOSITION PRESSURE

Ar/CH4=lOO/1 sccm and deposition time=8 hours by System H

210

5.3.3 Film Quality = g(Ar/HZ/CH4, t, p)

Figure 5.21 displays the Raman Spectra for different hydrogen ﬂow rates, with

argon ﬂow ﬁxed at 100 sccm, methane ﬂow ﬁxed at l sccm, pressure held constant at 120

Torr and deposition time of 8 hours by system I] [Huan 2000]. The intensity of the gra-

phitic/amorphous carbon peak around 1560 cm'1 increases as the input H2 concentration

decreases. By varying the H2 concentration, the variation of the graphitic peak at 1580

cm"1 has been observed. Lower H2 concentrations tend to have a higher graphitic peak.

This usually indicates more graphite content in the ultra-nanocrystalline diamond ﬁlms.

The FWHM of the 1332 cm"1 diamond peaks are the same. This may be an indication that

the crystal size remains the same.

211

 4

I

I

I

II

I

1332 cm’1 —’I'I

LIN," .

\

I

2.4:-re

4..."...‘v

_

b4.“

'1 VVVVV sewn! .'

"v.

3.5 I.

’

,'

I.

'

/

g,

H2=4 sccm

\*“'¢vt.\_

I

‘ \ I 'M

l W» - -. 9 . v-

1560 cm‘1 “I

H2=2 sccm

.
0
.
3

2 I"

I

X

/

Iii-“‘1'"

1 5 h

I

"' "Pr

‘

I/' ‘

\

A

J.

4‘;

1’

;——1580cnr1

H

4" W‘

2""

‘1‘

“'6‘ 4“ . w» 1’

‘1.

H2=1 SCCm

‘

.L

ﬁ-V\‘M~h‘ﬁ~d‘ ’

I

“EMF“...

—.

_

-t

w

1- W

'

H

'

r

k

In"!

I

I’llpﬂlﬂ

‘

 
  

112:0 sccm

' MWHHM‘IM

“m

‘u t "H H g

~ ‘5 u A .- J a

I
I]

I‘ a.

1.

‘

IL

II

"‘H.

III")

1

‘I

/

_

1'}

/

/

it

|

0 5

'

O

I

1000

1200

1400

1600

1800

2000

Shift Wavenumber

Pressure=120 Torr and AFIOO sccm, CH4=l sccm and deposition time =8 hours

FIGURE 5.21 : RAMAN SPECTRA

Spectra Taken by UV-Raman System at Argonne National Lab.

by system H.

212

Figure 5.22 displays the Raman Spectra for different deposition times, with pres-

sure held constant at 120 Torr and Ar/Hz/CH4 =100/4/l sccm. The FWHM of the 1332

cm’1 diamond peaks are the same. This may be an indication that the crystal size remains

the same.

 

2

I

I

l

l

1332 cm’1 —>

1560 cm'1

1.6L

r— 1580 crn'1

.

1.2 _

0.8

l

.
u
.
a

8 hours _

0.6

0.4

 

0.2 r

O

  

16 hours

  

   

l

I

     

l

 

 

1 000

1200

1400

1 600

1 800

2000

Shift Wavenumber

FIGURE 5.22 : RAMAN SPECTRA

Pressure=120 Torr and Ar/HZICH4=100/4/l sccm by system II.

Spectra Taken by UV-Raman System at Argonne National Lab.

213

Figure 5.23 displays the Raman Spectra for different deposition pressures, with

Ar/HZICH4 =100/2/l sccm and deposition time of 8 hours. The FWHM of the diamond

peaks are the same. This may be an indication that the crystal size remains the same.

 3

I

I

I

I

1332 cm”1 ~—*

      

1560 cm‘1

2.5 ‘

.
u
.
a

1.5

—

120 Torr

—1

140 Torr

 
  

I

I

l

l

0.5

0

1 000

1200

1400

1 600

1 800

2000

Shift Wavenumber

FIGURE 5.23 : RAMAN SPECTRA

Ar/H2/CH4=100/2/l sccm and deposition time=8 hours

by system 11.

Spectra Taken by UV-Raman System at Argonne National Lab.

 

  

214

5.3.4 Morphology = g(Ar/Hz/CH4, t, ft, p)

The ﬁlms deposited by system II are ﬁne grained with small smooth projections.

The small crystal sizes of the ﬁlms make it inaccurate to use AFM to determine the crystal

size.

Figure 5.24 displays the ﬁlm morphology for different hydrogen ﬂow rate with

argon ﬁow ﬁxed at 100 sccm, methane ﬂow ﬁxed at 1 sccm, pressure held constant at 120

Torr and deposition time of 8 hours. From Figure 5.24(a) to Figure 5.24(g), the ﬁlms all

have a featureless (not showing any crystalline facets) morphology. Therefore, the mor-

phology does not vary with changes in gas chemistry. However, the contour of the ﬁlm is

affected by the gas chemistry. They change from numerous small round protrusions (Fig-

ure 5.24(a), (b), (c), and (d)) to many small and medium-sized mounds (Figure 5.24(e), (f),

and (g)). The ﬁlm roughness (See Figure 5.17.) is higher for those ﬁlms which have more

irregular surface contours.

215

     

_

Figure 5.24(a)

H2=0 SCCIl'l

Digital Instruments Nanoscope

Sean Size

Sean rate

Number of sauples

Image Data

5.000 on

1.00! Hz

258

Height

Data scale

700.0 nu

DI!

X
N

1.000 vu/diu

700.000 nu/diu

arn039.000

Figure 5240))

H2=1 sccm

gigita! Instruments Nagoggspe

33:: :3:

13001 g

es

Image Data

Height

Data scale

700.0 nu

”"

X

2

1.000 vn/div

700.000 nun/div

arn028.002

FIGURE 5.24 : FILM MORPHOLOGY VS. H2 FLOW RATE

Pressure=120 Torr, Ar=100 sccm, CH4=1 sccm and deposition time=8 hours

by system H.

 

216

Figure 5.24(c)

H2=2 SCCIII

    

Digital zInstruments Nanoscope

Scans

Scan rate

Number Dot.samples

Image

5. 0m

1. 001 Hz

256

Height

Data scale

900.0 nu

  

”'

x

2

1

.000 um/diu

900.000 nm/diu

arn023.001

Figure 524(d)

H2=4 sccm

   

Digital zInstruments Nanoscope

Scan

Scan rate

.

um

1.001 Hz

Number of samples

2

Image Data

Height

Data scale

700.0 nm

  

”"

1. 000 um/diu

z

700. 000 nm/diu

arn008.000

FIGURE 5.24 (cont’d)

217

 

 

 

Digital Instruments NanoScope

Scan size

Scan rate

Number of samples

5.000 um

1.001 H2

.256

Height

.

Flgurc 524(6)

H2=6 SCCIII

  

Image Data

Data scale

400.0 an

  

1""

x

2

1.000 um/div

400.000 nm/div

arn066.000

'

Flgure 524(f)

H2=8 sccm

Digital Instruments NanoScope

Scan size

Scan rate

5.000 um

1.001 Hz

Number of samples

256

    

Image Data

Height

Data scale

300.0 nm

  

1‘“

x

2

1.000 urn/div

300.000 nm/div

arn067.001

FIGURE 5.24 (cont’d)

218

 

   

Scan si e

Scan rate

Image Data

Data scale

_

Flgufe 524(g)

H2=10 sccm

Number of samples

Digital Instruments NanoScope

5.000 um

1.001 H2

256

Height

600.0 nm

X

2

1.000 urn/div

600.000 nm/div

arn106.000

FIGURE 5.24 (cont’d)

219

Figure 5.25 displays the ﬁlm morphology for different deposition times with pres-

sure held constant at 160 Torr and Ar/Hz/CH4 =100/4/1 sccm. From Figure 5.25(a) to

Figure 5.25(c), the ﬁlms all have a featureless morphology. Therefore, the morphology

does not vary with changes in deposition time. However, the contour of the ﬁlm is differ-

ent for each ﬁlm. The contour of the ﬁlm changes from many round protrusions (Figure

5.25(a)) to numerous small smooth projections (Figure 5.25(b)) and to small and

medium-sized mounds (Figure 5.25(c)). The ﬁlm roughness (See Figure 5.18.) is higher

for those ﬁlms which have more irregular surface contours.

220

Figure 5.25(a)

t=8 hours

   

Digital Instruments Hanoscope

gem

size

t

can ra O

Number of samples

1'00: u";
5 000

-

256

Image Data

Height

Data scale

700.0 nu

  

”I

X

2

1.000 vm/dio

700.000 nm/diu

arn03?.001

Figure 5250’)

i=1] hours

   

2;:;*;:.L"S*M“ “sass"...

Scan rate

1.001 H2

Number of samples

Image Data

Height

Data scale

800.0 nm

4

X

2

1.000 Dm/div

800.000 nm/diu

DI arn051.000

FIGURE 5.25 : FH.M MORPHOLOGY VS. DEPOSITION TIME

Pressure=160 Torr, Ar/Hz/CH4=100/4/1 sccm by system II

 

  

221

Height  Data scale

Number of samples

Image Data

400.0 nm

Scan size

Scan rate

Digital Insrruments NanoScope

5.000 pm

1.001 H2

256

.

Flgure 5.25(C)

t=25 hours

1"“

x

2

1.000 urn/div

400.000 rim/div

arn131.002

FIGURE 5.25 (cont’d)

222

Figure 5.26 displays the ﬁlm morphology for different total gas ﬂow rates with the

pressure held constant at 120 Torr, Ar:H2:CH4=100:2:1 sccm and deposition time of 8

hours. From Figure 5.26(a) to Figure 5.26(e), the ﬁlms all have featureless morphology.

Therefore, the morphology is not affected by total gas ﬂow rate. However, the total gas

ﬂow rate does affect the contour of the ﬁlm. The contour of the ﬁlm changes from small

smooth projections (Figure 5.26(a)) to small round protrusions (Figure 5.26(b)) and to

small and medium-sized mounds (Figure 5.26(e), (d), and (e)). The ﬁlm roughness (See

Figure 5.19.) is higher for those ﬁlms which have more irregular surface contours.

223

Figure 5.26(a)

ft=103 SCCIIl

Digital Instruments Hanoscope

Scan size

Scan rate

5 000

.

pm

1.001 Hz

Number of samples

Image Data

Height

   

Data scale

900.0 nm

D"

X

2

1.000 um/div

900.000 nm/div

arn023.001

Figure 5.260))

ft=206 sccm

Digital Instruments NanoScope

Scan size

Scan rate

5.000 um

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

300.0 nm

1"”

x

2

1.000 urn/div

300.000 nm/div

arn089. 000

FIGURE 5.26 : FILM MORPHOLOGY VS. TOTAL GAS FLOW RATE

Pressure=120 Torr, Ar:H2:CH4=100:2:1 sccm and deposition time=8 hours

by system H

 

224

.

Flgure 5.26(c)

ft=309 sccm

Digital Instruments NanoScope

Scan size

Scan rate

5-000

1.001 H2

Number of samples

256

Image Data

Height

Data scale

500.0 nm

   

”m

X

2

1.000 um/div

500.000 nm/div

arn088.002

'

Flgure 526(d)

ft:412 sccm

  

  

Digi tal Instruments NanoStope

Scan size

Scan rate

5.000 um

1.001 NZ

Number of samples

256

Image Data

Height

Data scale

400.0 nm

  

W"

X 1.000 um/div

2

400.000 nm/div

arn091.000

FIGURE 5.26 (cont’d)

225

400.0 nm  

Number of samples

Data scale

Image Data

Height

Scan size

5.000

Digital Instruments NanoScope

Scan rate

1.001 Hz

256

Figure 5_26(e)

{1:618 sccm

“m

x

2

1.000 um/div

400.000 nm/div

arn086.001

FIGURE 5.26 (cont’d)

226

Figure 5.27 display the ﬁlm morphology for different deposition pressures with

Ar/HZICH4=100/4/l sccm and deposition time of 8 hours. As shown in Figures 5.27(a) to

5.270), the ﬁlms all have featureless morphology. Therefore, the morphology does not

vary with changes in deposition pressure. However, the contour of the ﬁlms are somewhat

different as the deposition pressure changes. The contour of the ﬁlm changes from numer-

ous smooth projections (Figure 5.27(a)) to small-sized mounds (Figure 5.27(b)) to tiny

and small-sized protrusions (Figure 5.27(c)) to numerous small smooth projections (Fig-

ure 5.27 (d)) to small smooth protrusions (Figure 5.27(e), (f), (g), and (h)) to small and

medium-sized hills (Figure 5.27(i)) and to small and medium-sized mounds (Figure

5.270)). The ﬁlm roughness (See Figure 5.20.) is higher for those ﬁlms which have more

irregular surface contours.

227

Fl UI'C

.27 a 2

Torr

g

5

( ) 60

Disi tal Instruments Nanosoope

Scan size

Sean rate

5.000 nu

LN: Hz

Number or sanples

luage Data

Height

Data scale

500.0 nu

arn159.000

‘

Figure 5.27(b). 80 Torr

.

Digital Instruments NanoScope

Sm Sm

5.000 0...

Scan rate

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

400.0 nm

”m

arn094.003

x

2

1.000 tam/div

400.000 nm/div

FIGURE 5.27 : FILM MORPHOLOGY VS. DEPOSITION PRESSURE

Ar/H2/CH4=100/4/1 sccm and deposition time=8 hours

by system II.

 

 

228

Figure 5.27(c): 100 Torr

Digital Instruments NanoScope

Scan we
Scan rate

5.000 um
1.001 H2

Number of samples

256

Image Data

Height

Data scale

300.0 nm

”m

arn093.001

x

2

1.000 urn/div

300.000 nm/div

Figure 5.27(d): 120 Torr

   

Digital Instruments Nanoscope

Scan st

'2e

Scan rate

5.

000 ”I

1.001 Hz

Nunber of samples

256

  

Data scale

luage Data

Height

700.0 nu

  

”"

X

2

1.000 uu/diu

700.000 nu/aiu

arn008.000

 

  

FIGURE 5.27 (cont’d)

229

Figure 5.27(e). 140 Torr

'

-

Digitalz Instruments Nanoscope

Sea:If:
Scan

5. 000 um
1. out "2

mber of samples

256

Image Data

Height

Data scale

700.0 nm

  

”I

X

2

1.000 um/diu

700.000 nm/div

arn030.000

Figure 5.27(f): 160 Torr

  

  

Digital Instruments Hanoscope

Scan

Scan rate

5. 000m

1.00! Hz

Number of samples

2

Image Data

Data scale

Height

700.0 nm

”"

X

z

.000 um/diu

700. 000 nm/dio

arn037.001

FIGURE 5.27 (cont’d)

230

Figure 5.27(g): 180 Torr

  

Digital Instruments Nanoscope

$2: 3;:

'

1:001 "2
5 000 um

Number of samples

258

Image Data

Height

Data scale

700.0 nm

”"

x

2

1.000 um/div

700.000 nm/div

arn055.000

Figure 5.27(h): 200 Torr

gggj‘tggzénstmnts Nag-1:33P;

Scan rate

1.001 Hz

Number of samples

Image Data

Height

Data scale

700.0 nm

  

”“

x

2

1.000 um/div

700.000 nm/diu

arn047.001

FIGURE 5.27 (cont’d)

231

  

Figure 5.27(i): 220 Torr

Digital Instruments NanoScope

scan size

Scan rate

5.000 um

1.001 H2

Number of samples

256

Image Data

Height

Data scale

500.0 nm

  

“m

X

2

1.000 um/div

500.000 nm/di v

arn074.000

Figure 5,270): 240 Torr

Digital Instruments NanoScope

  

Scan rate

ize

5.000 um

1.001 Hz

  

Number of samples

256

Image Data

Height

   

Data scale

500.0 nm

u"

x

2

1.000 um/div

500.000 nm/div

arn072.001

FIGURE 5.27 (cont’d)

232

Figure 5.28 display the ﬁlm morphology for different deposition pressures with

Ar/H2/CH4=100/O/1 sccm and deposition time of 8 hours. As shown in Figure 5.28(a) to

5.28(h), the ﬁlms all have a featureless morphology. Therefore, the morphology does not

vary with changes in deposition pressure. However, the contour of the ﬁlms varies as dep-

osition pressure changes. The contour of the ﬁlm changes from small round protrusions

(Figure 5.28(a) and (h)) to small-sized hills (Figure 5.28(c)) and to small smooth projec-

tions (Figure 5.28(d), (e), (f), (g), and (h)). The ﬁlm roughness (See Figure 5.21.) is higher

for those ﬁlms which have more irregular surface contours.

233

'

Figure 5.28(a): 60 Torr

Digital Instruments Nanoscope

Scan size

Scan rate

5 0

0 um

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

500.0 nm

“m

x

2

1.000 um/div

500.000 nm/div

arn102.003

'

.

Digital Instruments NanoScope

Flgure 5'28(b)' 80 Torr

Scan size

5.000 um

Scan rate

1.001 H2

Number of samples

256

Image Data

Height

Data scale

500.0 nm

””

x

2

1.000 um/div

500.000 nm/div

arn101.000

FIGURE 5.28 : FILM MORPHOLOGY VS. DEPOSITION PRESSURE

Ar/HZICH4=100/O/l sccm and deposition time=8 hours

 

 

by system II

234

Figure 5.28(c): 100 Torr

Digitalz Instruments Nanoscope

Scan

Scan rate

5. 000 I...

1.001 H2

Number of samples

258

Image Data

Height

Data scale

200.0 nm

arn100.002

X

2

1.000 Dm/div

200.000 nm/div

Figure 5.28(d): 120 Torr

gégital zInstruments Nagoscogpem

   

Data scale

700.0 nm

Scan raate

1.001 H2

Number of samples

258

Image Data

Height

”'

x

2

1.000 um/div

700.000 nm/diu

arn039.000

 

 

FIGURE 5.28 (cont’d)

235

Figure 5.28(c): 140 Torr

   

Digital Instruments NanoScope

e

1.001 H2

5.000

Number of samples

256

Image Data

Height

Data scale

0.0

  

W"

x 1.000 um/div

2

500.000 nm/div

arn098.000

Figure 528(1): 160 Torr

Digital Instruments NanoScope

Scan size

Scan rate

5.000 u

1.001 Hz

Number of samples

256

Image Data

Height

Data scale

400.0 nm

“m

x 1.000 urn/div

2

400.000 rim/div

arn099.001

FIGURE 5.28 (cont’d)

 

 

236

Figure 5.28(g): 180 Torr

  

Digital Instruments NanoScope

e

1.001 Hz

5

00

Number of samples

256

Image Data

Height

   

Data scale

500.0 nm

  

“m

x

2

1.000 um/div

500.000 nm/div

arn103.001

Figure 5.28(h): 200 Torr

  

Digital Instruments NanoScope

Scan size

Scan rate

5 000

-

Um

1.001 H2

Number of samples

256

Image Data

Height

   

Data scale

500.0

  

“m

X

2

1.000 um/div

500.000 nm/div

arn105.000

FIGURE 5.28 (cont’d)

237

5.3.5 Growth Mechanism & Crystal Size: g(Ar/Hz/CH4)

To investigate the growth mechanism and to observe the microstructure of the

ultra-nanocrystalline diamond ﬁlm on the silicon substrate, cross-section TEM images

were taken. The ultra-nanocrystalline diamond ﬁlm was deposited in system II with the

following experimental conditions: pressure=120 Torr, argon ﬂow rate=lOO sccm, hydro-

gen ﬁow rate=4 sccm, methane ﬂow rate=l sccm, time=8 hours, and no additional

nitrogen.

A cross-section TEM bright ﬁeld image of the diamond ﬁlm on a silicon substrate

is displayed in Figure 5.29. It is clearly shown in the image that there is a transitional layer

(~50 nm) between the silicon substrate and the diamond ﬁlm. The nature of this interlayer

is unknown. Further probing using EELS (Electron Energy Loss Spectroscopy) is recom-

mended to identify the nature of this interlayer. Further away from the silicon boundary

and the transitional interlayer (around 280 nm to the right, see the white arrow), there

seems to be some cone-shaped features. According to the results of the researchers from

ANL (Argonne National Laboratory), the cone-shaped features (see the black arrow) that

appear in Figure 5.29 are not columnar-type single crystals. It actually consists of many

nanosized grains [J iao 1999]. From Figure 5.29, diamond appears to deposit on the silicon

substrate directly (black arrow), in the transitional layer, and also in the ﬁlm beyond the

transitional layer (re-nucleation).

238

 

Ila-03ml

 

FIGURE 5.29 : BRIGHT FIELD TEM IMAGE

Black arrow: interlayer, white arrow: cone-shaped feature.

Sample: Deposition pressure=120 Torr, Ar/H2/CH4=100/4/l sccm and

deposition time=8 hours by system H

239

A pair of bright ﬁeld and dark ﬁeld cross-section TEM images of the diamond/sili-

con interface taken in the different area of the same specimen are displayed in Figure 5.30.

From the images, the transitional interlayer is invisible at this area. Diamond appears to

deposit on the silicon substrate directly. In the images, the MSU-ultra—nanocrystalline dia-

mond grains are equiaxed and aspect ratio. From Figure 5.30 (B), it is clearly shown that

there are diamond grains which re-nucleate after the boundary of silicon and diamond.

240

Ila-00ml

(A)

(B)

 

FIGURE 5.30 : BRIGHT FIELD(A) & DARK FIELD(B) TEM IMAGES

Sample: Deposition pressure=120 Torr, Ar/H2/CH4=100/4/1 sccm and

deposition time=8 hours by system H

241

Figure 5.31 displays a cross-section TEM dark ﬁeld image of Figure 5.30 at higher

magniﬁcation. It shows many small diamond grains agglomerated together. The

comet-tailed features (see the white arrow) are not single crystals because there are many

spots inside one tail. The spot inside the tails can be an individual crystal or groups of

crystals with the same orientation with respect to the diffraction condition.

Ila-@0101

 

FIGURE 5.3] : DARK FIELD TEM IMAGE

Sample: Deposition pressure=120 Torr, Ar/Hz/CH4=100/4/l sccm and

deposition time=8 hours by system II

242

The crystal sizes of the ﬁlms deposited by system 11 are very small. Plan-view

TEM (Transmission Electron Microscopy) images were taken to observe the microstruc-

ture and to measure the crystal sizes. Based on the results by cross-section TEM, the

cone-shaped and comet-tailed features (conglomerate structures) are not big single crys-

tals. Then the crystal size determined by plan-view TEM is meaningful.

A dark ﬁeld TEM image, Figure 5.32, was obtained by selecting a fraction of the

(11]) ring of the SAD (Selected Area Diffraction) with an objective aperture. It shows dif-

ferent orientations of the ultra-nanocrystalline diamond [Huan 2000]. Crystallinities

contributing to this fraction of diffraction illuminate brightly over the dark background.

The differences in the orientation with respect to the diffraction condition has caused the

differences in the contrast between the individual grains in the image. The upper bound

limit of the crystal sizes is about 20 nm.

243

 

Scale Bar=10 nm

FIGURE 5.32 : DARK FIELD TEM IMAGE BY ANL [HUAN 2000]

Sample: Deposition pressure=120 Torr, Ar/Hz/CH4=100/4/l sccm and

deposition time=8 hours by system 11

244

Figure 5.33 is a HRTEM (High Resolution TEM) image which shows planar

defects such as stacking faults (on the order of angstroms) and twinning [Huan 2000]. It

appears that there is not much amorphous carbon in the intergranular boundaries and there

is no evidence of segregation. The microtwins in the ﬁlm have been interpreted as a sec-

ondary nucleation mechanism. Such a secondary nucleation mechanism could be the

cause of the high nucleation rate which in turn produces nano-sized diamond grains [Luo].

In Figure 5.33, the crystal sizes range from 1—15 nm.

 

FIGURE 5.33 : HIGH RESOLUTION TEM IMAGE BY ANL [HUAN 2000]

Sample: Deposition pressure=120 Torr, Ar/H2/CH4=100/4/1 sccm and

deposition time=8 hours by system II

245

Due to the better contrast (than bright ﬁeld TEM) and higher statistics (capable of

covering hundreds or thousands of crystals rather than several in a typical HRTEM), a

dark ﬁeld image is used to determine grain size distribution for ultra-nanocrystalline mate-

rials. Figure 5.34 shows an automapping of histogram of grain size distribution of the ﬁlm

in Figure 5.32. The data was obtained with NIH imaging analysis software, showing that

the crystals are predominantly of the size S 7 nm [Huan 2000], [Huan 2001].

 

FIGURE 5.34 : A HISTOGRAM OF FIGURE 5.32 BY ANL [HUAN 2000]

Grain size distribution of MSU ultra-nanocrystalline diamond ﬁlm

Sample: Deposition pressure=120 Torr, Ar/Hz/CH4=100/4/l sccm and

deposition time=8 hours by system II

246

To investigate the inﬂuence of additional hydrogen on the growth mechanism,

another ultra-nanocrystalline diamond ﬁlm was deposited by system 11 with the following

experimental conditions: pressure=120 Torr, argon ﬂow rate=lOO sccm, methane ﬂow

rate=1 sccm, time=8 hours, and no additional nitrogen. To observe the microstructure of

the ultra-nanocrystalline diamond ﬁlm on the silicon substrate, cross-sectional TEM

images were taken.

A dark ﬁeld cross-section TEM image of the diamond ﬁlm on a silicon substrate is

displayed in Figure 5.35. The transitional interlayer (between two white arrows, ~70 nm)

between the silicon substrate and the diamond ﬁlm is larger than the ﬁlm deposited with 4

sccm of hydrogen ﬂow rate (see Figure 5.29). Diamond appears to deposit on the silicon

substrate directly, in the transitional layer, and in the ﬁlm beyond the transitional layer

(re-nucleation). As shown, there are many small crystals within a large cone-shaped fea-

tures. The inserted SAD (selected area diffraction) pattern shows the nature of diamond

but does not show any texture.

247

Din-(mail 

FIGURE 5.35 : DARK FIELD TEM IMAGE

Sample: Deposition pressure=120 Torr, Ar/CH4=100/1 sccm and

deposition time=8 hours by system 11

248

A bright ﬁeld plan-view TEM image with a corresponding SAD patter is displayed

in Figure 5.36. The diffraction rings, indexed as diamond, are visible up to (531). Both the

image and the SAD pattern show the polycrystalline nature of the ultra-nanocrystalline

diamond ﬁlm and indicate the randomness of the crystal orientations. As shown, the

ultra-nanocrystalline diamond grains grown by system II (pressure=120 Torr, argon ﬂow

rate=100 sccm, methane ﬂow rate=1 sccm, time=8 hours, and no additional nitrogen) has

both aspect ratio and equiaxed grains. Similar to the ultra-nanocrystalline diamond grains

grown by system II (pressure=120 Torr, argon ﬂow rate=100 sccm, hydrogen ﬂow rate=4

sccm, methane ﬂow rate=1 sccm, time=8 hours, and no additional nitrogen). The compar-

ison of this study with the results of similar research at ANL shows the ﬁlm grown with

Ar/CH4=100/1 sccm (other deposition conditions are the same) has fewer aspect ratio

grains [Jiao 1999].

249

Din-com

 

FIGURE 5.36 : PLAN VIEW BRIGHT FIELD TEM IMAGE

Sample: Deposition pressure=120 Torr, Ar/CH4=100/1 sccm and

deposition time=8 hours by system H

250

5.4

Controlled Nitrogen Impurity Study By System II

Figure 5.37 display the XRD ratio of integrated intensities of diamond <220> and

<111> peaks plotted as a function of nitrogen impurities with pressure at 120 Torr, gas

chemistry ﬁxed at Ar/H2/CH4=100/4/1 sccm and deposition time of 8 hours. The XRD

ratios of intensities of diamond <220> and <111> peaks are all above 0.25. Therefore,

most of the ﬁlms have the (220) preferred orientation. Between 0 and 2500 ppm of nitro-

gen impurity, the XRD ratios of intensities of diamond <220> and <111> peaks reach the

minimum as nitrogen impurity equal to 1200 ppm.

1 .5

I

 

I

I

I

1

a

:1

S

O

S

O

<>

0

1 .—

o-

0

°

0. 5 _

<>

0

 

O

O

l

L

l

l

—1

O

_

 

 

O

500

1 000

1 500

2000

2500

N2 (PPm)

FIGURE 5.37 : I(220)/I(111) VS. NITROGEN IMPURITIES

Pressure=120 Torr, Ar/Hz/CH4=100/4/l sccm and deposition time=8 hours

251

Figure 5.38 displays the ﬁlm roughness plotted as a function of nitrogen impurities

with pressure held constant at 120 Torr, gas chemistry ﬁxed at Ar/HZ/CH4=100/4/l sccm

and deposition time of 8 hours. As shown, between 0 and 2500 ppm of nitrogen impurity,

the ﬁlm roughness scatters between 18 to 25 nm.

25

.

.

 

,

.

.

<>

O

20—

o O

O

o

O

O

O

g 15-

5

a

.2

D1)

5‘ 1o~

5~

 

O

l

l

l

l

0

—

J

—

_

 

 

o

500

1000

1500

2000

2500

N2 (ppm)

FIGURE 5.38 : RMS ROUGHNESS VS. NITROGEN IMPURITIES

Pressure=120 Torr, Ar/H2/CH4=100/4/1 sccm and deposition time=8 hours

by System 11.

252

_
t
l
i
M

Figure 5.39 displays the ﬁlm morphology versus different nitrogen impurities, with

pressure held constant at 120 Torr, gas chemistry ﬁxed at Ar/H2/CH4=IOO/4/ 1 sccm and

deposition time of 8 hours. From Figures 5.39(a) to 5.390), the ﬁlms all have featureless

morphology. Therefore, the morphology does not change by adding nitrogen impurity up

to 2500 ppm. The contour of the ﬁlms varies slightly as different nitrogen impurity con-

centrations are added. The contours of the ﬁlms are mostly small smooth projections. The

ﬁlm roughness (See Figure 5.38.) does not change much within these ﬁlms because the

contours of the ﬁlm are quite similar.

The inﬂuence of nitrogen on the growth, morphology, and crystalline quality of

MPACVD diamond ﬁlms has been studied by Asmussen et a1. [Asmu 1998]. It was found

nitrogen impurities as low as 20 ppm could affect the morphology and quality of the ﬁlms.

Based on the results of controlled nitrogen impurity study, there is no observable morphol-

ogy changes of ultra—nanocrystalline diamond up to 2500 ppm. This may be because the

very small crystal sizes of UNCD. Therefore, one can not tell from the AFM inspections.

However, the ﬁlms do begin to look opaque. Therefore, the optical properties of UNCD

have been inﬂuenced by the additional nitrogen impurities.

253

Figure 5.39(a)

'

N25 ppm

Digital lnstruncnts Nanoscope

s
33%;: in:

I

5 000
1.001 n";

.

..Zgiro‘it."" “

Height

   

Data scale

700.0 nu

  

W

X

2

1.000 vn/div

700.000 nun/div

amOOS.000

'

Flgure 539(1))

Nz'loo ppm

Digital Instruments NanoScope

Scan size

Scan rate

5.000 pm

1.001 H2

Number of samples

256

Image Data

Height

Data scale

400.0 nm

W"

x

2

1.000 um/div

400.000 rum/div

arn012.001

FIGURE 5.39 : FILM MORPHOLOGY VS. NITROGEN IMPURITIES

Ar/HZ/CH4leO/4/l sccm and deposition time=8 hours

by system II

 

254

.

Flgure 5.39(C)

N2.200 ppm

Digital Instruments NanoScope

Scan size

Scan rate

S. 000 um

1. 001 Hz

Number of samples

256

Image Data

Height

Data scale

300.0 nm

x

1.000 urn/div

300.000 nm/div

2

”m

    ra

arn013.000

Flgure 5.39(d)

'

N22300 ppm

83“ aif'ns runents
D it I

Q

l

Scan
Number of samples

Image Data

Ma

M333"...

1. 001 Hz

Height

Data scale

700.0 nm

DH

2

700.000 nm/div

1.000 um/diu

arn014.001

FIGURE 5.39 (cont’d)

 

255

Figure 5.39(e)

N2:600 ppm

Digital Instruments NanoScope

Scan size

Scan rate

5.000

1.001 H2

Number of samples

256

   

Image Data

Height

Data scale

400.0 nm

   

1.000 pm/d1v

X

2

400.000 nm/div

arn016.000

Figure 5.39(f)

N2:800 ppm

Digital Instruments Nanoscort

Scan 51

ze

Scan rate

5.000 um

1.001 N:

Number of samples

Image Data

Height

Data scale

700.0 nm

DH

1.000 pm/div

2

700.000 nm/div

arn017.000

FIGURE 5.39 (cont’d)

 

 

256

_

Flgufe 539(g)

N2.1200 ppm

Digital Instruments Nanoscope

Scan size

Scan rate

5.000

1.001 H2

Number of samples

256

Image Data

Height

Data scale

400.0 nm

  
   1"“

arn018.001

1.000 pm/div

X

2

400.000 nm/div

'

Figure 5.39(h)

N221500 ppm

Digital Instruments Nanosco

5.000 w

sin

Scan rate

1.001 "2

Number of sample

Image Data

Height

Data scale

700.0 nm

”"

X

2

1.000 vm/diu

700.000 nut/div

arn019.000

FIGURE 5.39 (cont’d)

257

Figure 5.39(i)

N222000 ppm

Digital zInstruments Nanoscope

3:2,;

5.0001101
1.001 n:

Number aof asamples

Image Dat

Height

Data scale

700.0 nm

1’"

X

2

1.000 um/diu

700.000 nm/div

arn081.000

700.0 nm  

Number.of samples

Image Data

Height

Data scale

1. 001 Hz

258

Flgurc 5.39(i)

g‘gitzlieInstruments Nagoggpew

N222500 ppm

Scanr

”"

X

2

1.000 um/div

700.000 nm/div

FIGURE 5.39 (cont’d)

 

258

-
'

m
—
s
“
t
i

5.5 Summary

Ultra-nanocrystalline diamond ﬁlms synthesized by systems I and H have a variety

of ﬁlm textures, surface roughness, ﬁlm structure qualities, morphologies, and crystal

sizes. The comparisons of the ﬁlm characteristics of these two systems are presented as

follows.

5.5.1 System I

(a) Effect of gas chemistry

At deposition pressure of 120 Torr and deposition time of 8 hours, the ﬁlms have

preferred orientation in (220) direction when hydrogen concentrations are between 5-8%.

At deposition pressure of 120 Torr and deposition time of 8 hours, increasing hydrogen

concentration from 1.95% to 3.81% has increased the ﬁlm roughness 1.2 times for

ft=102.5-105 sccm experiments and 3.9 times for ft=615-630 sccm experiments. As

hydrogen concentration increases, the ﬁlm’s crystal size increases. The ﬁlms have (111)

triangular morphology with a deposition pressure of 120 Torr, deposition time of 8 hours,

and hydrogen concentration between 1.95% and 3.81%.

(b) Effect of deposition time

With deposition pressure at 120 Torr, Ar:H2:CH4=100:2:0.5, ft=615 sccm, the

ﬁlm’s preferred orientation did not change as the deposition time varied between 2 and 48

hours, but the ﬁlm roughness increased 10.6 times as the deposition time increased from 2

to 48 hours. As the deposition time increased, from 2 to 8 hours, the crystal size increased

259

-
1

.
.
.
.
.
_
.
_
,
_
_
_

and the morphology was kept constant at a (l 1 l) triangular morphology. After 8 hours of

deposition time, the ﬁlm morphology became mixed morphology.

(c) Effect of total gas ﬂow rate

The ﬁlm’s (220) preferred orientation became weaker with higher total gas ﬂow

rates. The crystal size variation became larger as total gas ﬂow rate was reduced.

((1) Effect of deposition pressure

Film roughness increases 1.9 times as deposition pressure increases from 120 to

160 Torr with Ar:H2:CH4=100:2:0.5, ft=615 sccm and deposition time of 8 hours. The

crystal size increased when deposition pressure increased. This was also supported by the

ﬁlms’ Raman Spectra. With Ar:H2:CH4=100:2:0.5, ft=615 sccm and deposition time of 8

hours, the ﬁlm has (111) triangular morphology between the deposition pressure of 120 to

160 Torr.

5.5.2 System II

(a) Effect of gas chemistry

At deposition pressure of 120 Torr, deposition time of 8 hours, AI=100 sccm,

CH4=1 sccm, and ft=101-111 seem, the ﬁlms have (220) preferred orientation. When

hydrogen ﬂow rate was increased from O to 10 sccm, the ﬁlm roughness increased 4 times,

and the ﬁlms had a featureless morphology. The morphology did not vary with changes in

gas chemistry. The surface contour of the ﬁlms became more irregular as hydrogen con-

260

centration increased. As hydrogen ﬂow rate increased from O to 4 sccm, the ﬁlms had less

graphitic/amorphous material.

(b) Effect of deposition time

With deposition pressure at 160 Torr and Ar/Hz/CH4=100/4/l sccm, the ﬁlm

roughness increases 1.5 times as the deposition time increases from 8 to 25 hours, but the

morphology remained featureless. Therefore, the morphology did not vary with changes in

deposition time. However, the surface contour of the ﬁlms was more irregular as deposi-

tion time increased.

(c) Effect of total gas ﬂow rate

The ﬁlm’s preferred orientation in (220) direction became weaker with higher total

gas ﬂow rates. At a deposition pressure of 120 Torr and deposition time of 8 hours, as total

gas ﬂow rate increases, the ﬁlm roughness increased 3.4 times for Ar:H2:CH4=100:4:1 gas

chemistry ratio. At a deposition pressure of 120 Torr and deposition time of 8 hours, as

total

gas ﬂow rate

increases,

the ﬁlm roughness

increased

2.1

times

for

Ar:H2:CH4=100:2:1 gas chemistry ratio. The ﬁlm morphology was featureless and the

surface contour of the ﬁlms was more irregular with higher total gas ﬂow rates. The mor-

phology was not affected by total gas ﬂow rate.

((1) Effect of deposition pressure

261

When the argon ﬂow rate=100 sccm, methane ﬂow rate=1 sccm, and deposition

time of 8 hours, the ﬁlms have (220) preferred orientation. The (220) preferred orientation

was maintained as the deposition pressure was varied between 100 and 240 Torr. A maxi-

mum occurs around 140 or 160 Torr except for hydrogen ﬂow rate of 1 seem. As the

deposition pressure increased, the ﬁlm roughness dropped at the beginning and then con-

tinuously increased. The ﬁlm morphology is featureless despite the variation of the

deposition pressure. The surface contour of the ﬁlms becomes more irregular as the depo-

sition pressure increases.

(e) Effect of nitrogen impurity

The minimum of I(220)/I(1 1 1) occurs at 1200 ppm of added nitrogen impurity and

the ﬁlms all have (220) preferred orientation between nitrogen impurities of 0-2500 ppm.

There is no signiﬁcant effect on ﬁlm roughness with nitrogen impurity concentrations up

to 2500 ppm. The ﬁlms all have featureless morphology. Therefore, the morphology does

not change by adding nitrogen impurity up to 2500 ppm. The contour of the ﬁlms varies

slightly as different nitrogen impurity is added.

(e) Growth mechanism and crystal size study by TEM

Based on TEM observation, the ultra-nanocrystalline diamond ﬁlms synthesized

by system H consist of equiaxed and aspect ratio grains. The existence of a transitional

interlayer between the silicon substrate and diamond ﬁlm is clearly shown by the TEM

cross-section images. Diamond appears to be able to nucleate on the silicon, in the transi-

tional layer, on the diamond powder residue from the mechanical scratch process, and

262

'
l

continues to re-nucleate as the ﬁlm grows. The SAD patterns indicate that ultra-nanocrys-

talline diamond has random orientation. However, the XRD results show mostly preferred

(220) orientation. Neither SAD pattern nor the X-ray two-theta scan is able to quantita-

tively determine the

texture of the ﬁlms. Therefore,

to

verify

the

texture of

ultra—nanocrystalline diamond ﬁlms quantitatively, a more detailed study by X-ray pole

ﬁgures and inverse pole ﬁgures is recommended.

Planar defects of ultra-nanocrystalline diamond are shown by HRTEM technique.

Not much amorphous carbon was shown in the intergranular boundaries. The small activa-

tion energy involved in defect formation attributed to the forming of these structure

imperfections [Zhu 1983]. The interaction of twins in diamond crystals has been studied

by Shechtman et a1. and it’s been found that the meeting points of twins could play a very

important role for the nucleation and growth of diamond crystals [Shec 1993], [Luo],

[Grue 1994]. It’s been demonstrated and claimed that the vertices, or edges of such

twinned regions provide the nucleation sites for the new planes [Rang 1966] [Hami 1960].

The planar faults are likely formed during growth of secondary renucleation crystallinities

[Qin 1998].

Because TEM is a two dimensional imaging technique applied to see a three

dimensional structure. The grain overlapping may give rise to uncertainties in distinguish

the grain boundary which leads to grain size determination. Besides that, the grain visibil-

ity with a diffraction condition relies on the resolution of the diffraction contrast which is

attributed to the differences in crystals’ orientations with respected to the incident electron

beams. Therefore, combining the information from Figure 5 .32, 5.33, and 5.34, the crystal

263

size of the ultra-nanocrystalline diamond is in the range of 3-20 nm with an average of 7

nm.

(e) Film uniformity study by SEM

Film uniformity was determined by breaking the wafer and measuring the thick-

ness of the ﬁlm using the cross-sectional SEM imaging technique at the following

locations: (1) center, (2) one inch away from the center, (2) two inches away from the cen-

ter, and (4) edge of the substrate. Percent uniformity is deﬁned by the minimum thickness

divided by maximum thickness. The uniformity and the corresponding deposition condi-

tions are listed in Table 5.3.

264

TABLE 5.3: FILM UNIFORMITY STUDY

 

Ar/Hz/CH4

Pressure

t

Ts

Pabs

Uniformity

(sccm)

(Torr)

(hf)

(°C)

(W)

(%)

 

100/4/ 1

60

8

UFL

768

94

 

100/0/1

80

8

UFL

772

100

 

100/1/ 1

80

8

UFL

768

100

 

100/2/ 1

80

8

UFL

772

95

 

100/0/1

100

8

UFL

803

88

 

100/1/1

120

8

594

890

76

 

100/2/1

120

8

616

906

70

 

100/1/1

140

8

621

1005

89

 

100/0/1

160

8

608

1037

80

 

100/1/1

160

8

655

1119

74

 

100/2/1

160

8

696

1147

71

 

100/4/1

160

8

709

1214

77

 

 

180

180

 

100/4/1

100/2/1

  
 
  

100/4/1

100/1/1

100/4/1

8

8

8

8

8

 

77

88

75

87

770

730

200

240

681

709

673

 

1211

1361

1251

 

 

200

 

1429

1518

70*

*: There is a spot at the center of the deposited ﬁlm caused by the small ﬂame at the center

of the discharge with this deposition condition.

265

6

Plasma Diagnostics

6.1

Introduction

Knowledge of the internal variables is important in order to develop an understand-

ing of the ultra—nanocrystalline diamond deposition process and to relate the experimental

data to the theoretical simulations of the conditions. MSU ultra—nanocrystalline diamond

ﬁlms were deposited with a mainly carbon containing argon plasma and with only a small

fraction of hydrogen feed gas. The presence of argon in the input gases has been found to

inﬂuence the plasma chemistry. Therefore, it indirectly inﬂuences the characteristics of

diamond ﬁlms, such as surface morphology, grain size, and even affects diamond growth

mechanism [Zhou 1998], [Grue 1995]. In particular, it was found by Zhu et a1. [Zhu 1990]

that argon has an important inﬂuence on the plasma chemistry, including ionization and

dissociation of reactant molecules. The carbon dimer, produced by the dissociation of

methane molecules in a low hydrogen content argon plasma, is hypothesized to be the pri-

mary growth species in ultra-nanocrystalline diamond synthesis. Goyette et a]. [Goyl

1998], [Goy2 1998] have shown that the C2 emission intensity is linearly proportional to

the absolute carbon dimer density (as measured by a high-sensitivity white-light absorp-

tion spectroscopy apparatus) in the Ar-Hz-CH4 plasma discharges.

In order to understand the deposition process, a study of the C2 emission intensity

versus various process conditions was performed in system 11 and the results are described

in this chapter. The deposition process is sensitive to the gas kinetic temperature. The

dependence of substrate temperature versus various input parameters have been presented

in Chapter 3 (Figure 3.12 and Figure 3.13) and Chapter 4 (Figure 4.18 and Figure 4.21).

266

Therefore, only the gas neutral temperature versus various operating conditions is pre-

sented in this chapter. The theory of OES used in the gas temperature measurement is

described in Section 6.3 and the experimental method and measurements are described in

Section 6.4. The experimental data, including C2 emission intensity, gas temperature, and

controlled nitrogen impurities are presented in Section 6.5. The kinetic gas temperature of

discharges was measured by observing Swan band (0,0) emission intensity using OES

(optical emission spectroscopy).

Section 6.6 summarizes the plasma diagnostics

measurements.

The inﬂuence of controlled input nitrogen impurities on Ar—Hz-CH4 plasma is also

presented at the end of this chapter. The motivation for this set of measurements was to

understand how the addition of nitrogen inﬂuences the gas chemistry. It has been shown

that the nitrogen additions produce changes in the bulk plasma properties [Clay 1996] and

[Hong 1995]. However, one should not overlook that the nitrogen has an inﬂuence on gas

phase chemistry as well as the deposition surface [Hong 1995] and [Moss 1999]. The CBS

measurements presented in Section 6.5 investigate the bulk plasma property changes.

6.2 Background Fundamental of C2 Emission Spectroscopy

The optical emission intensity of the d3II-a3II (0,0) vibrational band of the C2

Swan system as a function of C2 concentration in an argon-hydrogen-methane microwave

plasma was studied by Goyette et a1. [Goye 1998]. The argon-hydrogen-methane plasma

was created in a commercial ASTeX PDS—17 MPECVD system. The substrate tempera-

267

ture was independently controlled by a rf-heater. The ranges of plasma processing

parameters used in the measurements are listed in Table 6.1.

_

.
.

?
"
i
u
a
M

.
-

TABLE 6.1: EXPERIMENTAL PARAMETERS [Goyl 1998], [Goy2 1998]

 

Ar

H2

CH4

Pressure

Substrate

Power

sccm

sccm

sccm

Torr

0C

W

 

97

2

1

55-150

800

800

 

97

2

1

100

800

500—1500

 

79-99

0-20

1

100

800

800

 

 

 

97

97

 

2

2

 

2

2

 

 

 

85

400-900

800

100

400-900

800

 

 

In Goyette’s experiments, two viewports equipped with fused silica windows pro-

vide in situ optical access. The emission intensity is measured using the (0,0) bandhead. A

typical spectrum of the carbon dimer d3l'I-a3l'l Swan band emission intensity around

516.5 nm is shown in Figure 6.1. The absolute density of C2 is measured using high-sensi-

tivity white-light absorption spectroscopy obtained from the continuum source of a

stabilized high-pressure Xe lamp. The absolute densities of C2 in the a3l'l state was car-

ried out using the (0,0) vibrational band of the Swan system and used 1300 °C (1573 K,

rotational temperature determined from d3TI -a3 II (0,0) vibrational band transition of C2,

the conditions, such as pressure, gas chemistry, microwave power, etc. of the measure-

ments were not reported) in determining the Boltzmann factors (Among the vibrational

levels of the d3l'lu state, this statistical fraction includes the relative population distribu-

tion [Goye 1998].)

268

140

 

r

r

(0,0)

R25

.

'

-

'

'

 

 

R45

‘

l

1

1

.

r

-

"‘

-

4

 

A

33

'3

:3

b
a

H._

1%
>5

v

1::
'aB

t:

120r

100—

so ~

60 ”

40 ~

20~

 

o

 

5000

5050

5100

5150

5200

A (an gstrom)

FIGURE 6.1 : CARBON DIMER d3r1 -a3rI SWAN BAND EMISSION

As the various plasma parameters were varied, the spectroscopically determined

C2 densities were correlated with the observed Swan band emission intensity from the

microwave discharge. Goyette et a1. [Goye 1998] observed a strong linear correlation

between the experimentally measured bandhead emission intensity and the experimentally

measured C2 concentrations. They observed the C2 densities were linearly related to C2

emission intensity as chamber pressure, substrate temperature, H2 percentage in the feed

gas were varied over the experimental parameter range shown in Table 6.1.

Based on the work of Goyette et al., in this thesis, it is assumed that C2 emission

intensity measurements are linearly related to C2 concentrations in the discharge. This

may not be a valid assumption since the experimental parameter space, i.e. pressure,

269

microwave, and reactor geometry were quite different from those used by Goyette et a1.

and were performed across a larger parameter space. Thus, C2 emission intensity may not

be exactly linearly related to C2 concentrations. In the future, this should be veriﬁed for

the experimental conditions presented in this thesis.

6.3 Gas Kinetic Temperature Measurement Theory

The rotational temperature was measured by CBS and then was used to determine

the gas temperature. The energy separations between rotational levels in a given vibra-

tional state are relatively small. Typically they are about a few hundred cm‘l. In contrast,

the vibrational frequency corresponds to an energy of the order 1000 cm’1 and the total

changes in vibrational energy may be up to an order higher [Gayd 1968].

Nearly all gas kinetic collisions produce a change in the rotational quantum num-

ber of diatomic molecules, whereas collisions producing a change in the vibrational or

electronic quantum numbers occur far less frequently. The relative rotational population

distribution in a sufﬁciently long-lived vibrational state is a Boltzmann distribution func-

tion which reﬂects the gas kinetic temperature. In the Boltzmann distribution, the

temperature parameter may be measured in many ways, depending on the plasma constit-

uents, the temperature, etc. Generally, emission methods are the most applicable [Thor

1988] and thus the methodology applied in this chapter is conﬁned to emission. The rela-

tive emission intensities of rotational lines within a single vibrational band is described by

[Thor 1988]:

270

—th .J’(J’+1)

I = K043 , ”Exp

J]

  (6.1)

V
kT

rot

Here,

K is a constant for all lines originating from the same electronic and vibrational level.

1) is the frequency of the radiation.

81,1” is the appropriate Honl-London factor.

h is the Planck’s constant.

c is the speed of light.

Bv, is the molecular rotational constant for the upper vibrational level.

I is the rotational quantum number. J’ represents the upper level and J” represents the

lower level.

k is the Boltzmann constant.

Tmt is the rotational temperature.

By

ﬁtting

the

experimental

data

to

the

expression

I

~

Sm"

exp[-BV’J’(J’+1)/(k*Tmt)] using the R25 to R45 emission lines [Goye 1998], [Pras 1994],

[Grot 1999], [Huan 2000], [Huan 2001], the rotational temperature is determined. The

rotational temperature was measured for the d3 H -a3l'I (0,0) vibrational band transition of

C2. The a3l'l level of C2 is very long-lived and almost surely equilibrated. But the upper

271

electronic level, d3H , of this transition radiates and is short-lived. At the pressure range

investigated in this chapter, it is likely that the collision rate is high enough to equilibrate

the rotational modes of this electronic excited state and the rotational populations of this

electronic state equilibrate to the gas kinetic temperature [Goye 1998] and [Gicq 1996].

6.4

Experimental Method and Measurement

The experiments were performed in the MSU-MPACVD reactor system II shown

in Figure 6.2. Since in these experimental measurements, the C2 emission intensity

depends on the volume of plasma discharges. Therefore, the discharge volume was kept

constant. The details of the reactor were previously described in Section 3.2, Section

3.2.1, Section 3.2.2.2, Section 3.2.3.2, and Section 3.2.4.1. The starting procedure was

described in Section 3.3.2.2. The operating pressure is 60-240 Torr, the gas chemistry is

conﬁned: Ar/Hz/CH4=99-90/0-9/1 (%), the absorbed microwave power is between

681-1520 W, and the total gas ﬂow rate is 100-600 sccm.

The plasma is very bright, therefore, there is no need to perform any optical

arrangement to focus the emission into the core of the ﬁber (dia.=0.8mm). A spectrometer:

Spex I, 1 meter, f/9 plane grating scanning monochromator with 1200 grooves/mm grating

and an EMI-GENCOM FACT-50 MKIH 3010 photomultiplier tube were used to detect the

signal. The entrance slit was set to 100 pm for acquiring CN spectrum, 20 pm for acquir-

ing C2 spectrum, and 2-5 pm for

rotational

temperature measurements. The

photomultiplier tube is mounted at the exit slit. For all the experiments, a bias voltage of

-800 Volts by an EMI-GENCOM high voltage power supply model 3000R was applied to

272

-

'

“
?
f
i
”

the photomultiplier tube in order to obtain a good signal-to-noise ratio. The output current

from the photomultiplier tube was read by a Keithley 480 Autoranging picoammeter. The

output current was usually in the range of a few nanoamperes and should not exceed a

microampere. The picoammeter was interfaced to a computer. The data acquisition and

processing were carried out by the computer. The software consisted of a Quick-Basic

program [Moss 1998] and was used to collect the data at the desired interval. The data

analysis applying the intensity function of Boltzmann distribution with R25 to R45 emis-

sion lines was used to ﬁt the experimental data to determine the rotational temperature is

written in MATLAB [Grot 1999], [Huan 2000], [Huan 2001]. The molecular rotational

constant for the upper vibrational level is 1.7456 and for the lower vibrational level is

1.62405 [Pras 1994]. The Honl-London factor is, S=(J’+1)(J’-1)/J’.

273

a“

mill-
 
  

 

-

,

\
K

N

N

:
a

 

 

 

,3

 

J;:-.__—_.:

Optical Fiber

' I”. ..------

  
 

 

 

 

—F""

 

 

Spectrometer

 
  

1P

 

Mirrox""""""" ’ ' ‘ Pyrometer

   

 

 

Picoammeter

 

 

 

 

Computer

 
  

FIGURE 6.2 : EXPERIMENTAL SET-UP FOR OES MEASUREMENTS

System 11: 2.45 GHz resonant cavity reactor,

Thermal ﬂoating substrate holder,

Ar-Hz-CH4 plasma discharges

 

274

6.5

Experimental Results

6.5.1 C2 Emission Intensity

Figure 6.3 displays the C2 emission intensity plotted as a function of hydrogen

ﬂow rate with argon ﬂow at 100 sccm, methane ﬂow at l sccm, and pressure at 120 Torr.

AS Shown, starting from the hydrogen ﬂow of 2 sccm, the C2 emission intensity decreases

with the increases in the hydrogen concentration. However, when hydrogen ﬂow is below

2 seem, the C2 emission intensity increases as the hydrogen concentration increases.

The intensity that is detected depends not only on the plasma emission but also

upon the visibility of the quartz dome. At low hydrogen ﬂow experiments, i.e. H2=0 sccm

experiment, a smoke-like vapor was produced. This smoke-like vapor deposits a ﬁlm on

the quartz dome which inhibits the transmission of plasma emission through the quartz

dome. Part of the emission was blocked by the ﬁlm and the measurements were affected.

The quartz dome was clear by the time when the hydrogen ﬂow reached 2 sccm. From 0 to

20% of H2 (vol.%), there is a strong linear correlation between emission intensity and C2

density in Ar-Hz-CH4 plasma [Goye 1998]. Therefore, the result indicates that the C2 con-

centration decreases with increases in hydrogen percentage in the input gases between

2-9% Of H2.

275

 

-
m

I

O

I

O

I

O

I

I

0

<>

_

.

-

.

<>

6

5 r

4 ~

3 -

<>

2 ~

)
.
u
.
a
(

y
t
i
s
n
e
t
n
I

n
o
i
s
s
i
m
E

2
C

1 _

 
  

J—

_.

l

O

1

l

l

0

2

4

6

8

1 0

H2 (sccm)

851

863

871

894

908

922

934

Absorbed Microwave Power (W)

FIGURE 6.3 : C2 EMISSION INTENSITY VS. H2 FLOW RATE

Pressure=120 Torr, Ar=100 sccm and CH4=1 sccm by system I].

276

Figure 6.4 displays the C2 emission intensity plotted as a function of total gas ﬂow

rate, ft, with Ar:H2:CH4=100:4:1, pressure at 120 Torr, and ft=105-630 sccm. It indicates

that the C2 emission intensity is essentially constant as the total gas ﬂow rate increases.

-
1
‘

 

4.5 r

O

3.5 r

2.5 r

1.5-

r

-

r

r

)
.
u
.
a
(

y
t
i
s
n
e
t
n
I

n
o
i
s
s
i
m
E

2
C

 

0.5 -

 

-

 

J

O

1 00

200

300

400

500

600

700

Total Gas Flow rate (sccm)

908

908

908

908

908

Absorbed Microwave Power (W)

FIGURE 6.4 : C2 EMISSION INTENSITY VS. TOTAL GAS FLOW RATE

Pressure=120 Torr, Ar:H2:CH4=100:4:1, ft=105-630 sccm by system 11.

277

Figure 6.5 and 6.6 display the C2 emission intensity plotted as a function of pres-

sure, p, and absorbed microwave power, Pabsr respectively, with argon ﬂow rate and

methane ﬂow rate held constant at 100 sccm and 1 sccm, respectively. The microwave

power input was adjusted (increased) as the pressure increases to maintain a discharge

area Ad that covered the three inch substrates because the discharge area Ad was varied

with the deposition pressure, p, and the absorbed microwave power, Pabs.

AS shown in Figure 6.5, the C2 emission intensity generally increases as pressure

increases. A small ﬂame starts to Show at the center of the plasma at a different pressure

for each different hydrogen gas ﬂow rate. For the experiments of H2=4 sccm, it’s around

240 Torr. For the experiments of H2=2 sccm, it’s around 220 Torr. For the experiments of

H2=1 and 0 sccm, it’s around 200 Torr. For the experiments with H2=0 seem, the quartz

dome got quite dirty toward the end of the experiments. Therefore, the decreasing trans-

mittance of the quartz dome decreased the emission intensity measurements at higher

pressure regime of this set of measurements. For the measurements of the experiments of

H2=4 sccm, the ﬂame appeared at 240 Torr. It could have inﬂuenced the plasma emission.

From 55 to 150 Torr, there is a strong linear correlation between emission intensity and C2

density in Ar—Hz-CH4 plasma [Goye 1998]. Therefore, the result indicates that the C2 con-

centration increases with increases between 60-160 Torr [Goye 1998], [Grot 1999]. The

data between 180—240 Torr could have the similar behavior. Nevertheless, a veriﬁcation by

an absorption spectrum from a high-sensitivity absorption Spectroscopy apparatus is rec-

ommended. AS shown in Figure 6.6, the C2 emission intensity increases with increases in

absorbed microwave power.

278

The same experimental data points that are shown in Figure 6.6 are also displayed

in Figure 6.5. Thus, it is important to note that as microwave power increases in Figure

6.6, the deposition pressure is also increasing.

279

l

I

1

1

1

I

1

1

1

I

1

1

1

1 00

 

 

[

o

H2=4 sccm

 

O

o

o

50 ‘

 

O

O

O

43

Q

Q

r

1

O

‘

 

 

50

1 00

1 50

200

250

 

200

 

1 a

H2=2 sccm ]

150-

 

100 '

‘k

a

ﬁ'

L

*

‘k

I

-

‘1

-

 

0

a

4:

t

 

50

1 00

1 50

200

250

 

 

1

O

H2=1 sccm I

 

c

9

50 “

‘

 

4

'

 

A;

;

4

l

 

50

1 00

1 50

200

250

 

 

f v

H2=O sccm I

 

20 —

10 -

V

V

v

V

O

V

V

Y

1

1

 

50

1 00

1 50

200

250

Pressure (Torr)

FIGURE 6.5 : C2 EMISSION INTENSITY VS. PRESSURE

Ar=100 sccm and CH4=1 sccm by system II.

280

)
.
u
.
a
(

y
t
i
s
n
e
t
n
I

n
o
i
s
s
i
m
E

2
C

50 r

 

1 00

 

0

30

 

l

l

100

 

1

 
[
H2=4 sccm j
 

o

0

o

50 r

O

_

  
 

n n

O O

O

I

1

r

800

800

1 000

1 200

1400

1 600

 

200

 
H2=2 sccm J
[ w

I

 

150 ‘

A 1 00 h

3:

V

I

I

‘k

ﬁ’

‘

"

3 50 —

  
 

g: i ﬁ ﬁ'

‘k

‘9:

I

r

‘k

_.

800

800

1 000

1 200

1400

1 600

 

1

.

 
H2=1 sccm j
1

«‘>

 

1

.

’3’

.

b

g

3...;

g 100

"a

E
(1.1

 

L3“

50 ~

1

 

 

800

800

1 000

1 200

1400

1600

 

30

 
H2=0 sccm ]
r v
 

I

20 '

10 ~

I

I

I

v

v

V

_

-

  
 

17 ‘7

V

.

,

.

0

600

800

1 000

1 200

1400

1 600

Absorbed Microwave Power (W)

FIGURE 6.6 : C2 EMISSION INTENSITY VS. MICROWAVE POWER

AFIOO sccm and CH4=1 sccm by system 11.

281

-
'

Figure 6.7 displays the growth rate plotted as a function of C2 emission intensity

with an argon ﬂow rate of 100 sccm, a methane ﬂow rate of 1 sccm, t= 8 hrs, hydrogen

ﬂow rate of 2 sccm (Figure 6.7 (a)) and 1 sccm (Figure 6.7(b)). The same experimental

data points that are shown in Figure 6.7 are also displayed in Figures 6.5 and 6.6. thus, it is

important to note that as C2 emission increases in Figure 6.7, the microwave power and

pressure are also increasing. Figure 6.7 shows the close relationship between the growth

rate and C2 concentrations, and suggests that the concentration of C2 in the argon-methane

plasma may be an important factor in determining the ﬁlm growth rate under the low and

constant hydrogen concentrations.

282

-
'

ﬂ
_
-
T
i
.
.
v
a
n
g
‘
-

n
u
e
m
)
g
r
o
m
r
q

3
2
1
9

)
1
1

u
I
/
q
r
(

15

 

I

I

r

0.942

 

a

H2=2 sccm

 
  

10 -

*

10.628

II

ﬁ

t

a

5 '

 

1?

1.

C

51

5:; 0L

1

1

1

40.314

 

0.000

 

0

50

100

(a)

150

200

2

G!

a:

:5

 

I

I

I

I

0.942

 

0

H2=1 sccm

 
  

E

15

g

ﬁ

10 r

5 -

g

T

 

0 ; Iv;

1

l

l

1

10.628

-0.314

 

0.000

 

0

20

4O

60

80

100

(b)

C2 Emission Intensity (a.u.)

FIGURE 6.7 : GROWTH RATE VS. C2 EMISSION INTENSITY

Ar=100 sccm, CH4=1 sccm, and t=8 hours by system 11.

283

-
1
-
1
+
‘
.
‘
W
‘

6.5.2 Gas Temperature

Figure 6.8 displays the gas temperature plotted as a function of hydrogen ﬂow rate,

with argon ﬂow ﬁxed at 100 sccm, methane ﬂow ﬁxed at 1 sccm, and pressure at 120 Torr.

As shown, the gas temperature increases with the increases in the hydrogen concentration.

The substrate temperature and the absorbed microwave power versus hydrogen ﬂow rate

are Shown in Figure 4.18. The uncertainty of the gas temperature is 100K.

2800

1

1

 

1

1

1

2700 _

0

O

Q 2600 r

O

I:

§

8 2500 ~
o.

15.
1-1

3

O 2400 -

 

2300 -

2200<>

0

O

O

l

l

l

1

l

-

-

-

1

r

 

 

0

2

4

6

8

1 0

H2 (sccm)

FIGURE 6.8 : GAS TEMPERATURE VS. H2 FLOW RATE

Pressure=120 Torr, Ar=100 sccm and CH4=1 sccm by system II.

284

Figure 6.9 displays growth rate plotted as a function of gas temperature with an

argon ﬂow rate of 100 sccm, a methane ﬂow rate of l sccm, t=8 hours. Hydrogen concen-

tration increased from 0-9% (See Figure 6.8). The data Shown in Figure 6.9 indicates that

as the gas temperature increases from 2215-2745 K, the growth rate increases.

-
7

m
.
.
m
_

6

1

1

1

1

1

0.375

 

5 r

’1':
g 4 -

E”
v

a)
.5

9‘

'5

E

65

3 -

o

0

E 2 _ O

12

O

O

O

0

10.3125

‘025

E‘.

8

9"
C)

8

2

. 0.1875 8'

70

1%

e

.3:

r O. 125

1 r

 

- 0.0625

 

0

4

r

1

r

1

00

 

2200

2300

2400

2500

2600

2700

2800

Gas Temperature (K)

FIGURE 6.9 : GROWTH RATE VS. GAS TEMPERATURE

Ar=100 sccm, CH4=1 sccm, t=8 hours by system 11.

285

Figure 6.10 displays the gas temperature plotted as a function of total gas ﬂow rate,

ft, with Ar:H2:CH4=100:4:1, ft=105-630 sccm and pressure at 120 Torr. The gas tempera-

ture ﬂuctuates around 2330 K. As Shown in Figure 4.21, the absorbed microwave power

and substrate temperature versus total gas ﬂow rate are constant under the same experi-

mental conditions displayed in Figure 6.10. The uncertainty of the gas temperature is

100K.

2400

I

I

I

I

I

I

 

2380

I

2360

I

2340 -

2320

I

2300

I

2280

I

2260

I

)
K
(

e
r
u
t
a
r
e
p
m
e
T

s
a
G

<>

I

l

_

l

1

_
J

l

0

 
  

_
J

I

l

l

1

l

l

l

2240 I

I

2220

22 0

O 0

100

200

300

400

500

600

700

Total Gas Flow Rate (sccm)

FIGURE 6.10 : GAS TEMPERATURE VS. TOTAL GAS FLOW RATE

Pressure=120 Torr and Ar:H2:CH4=100:4:1, ft=105 sccm by system 11.

286

Figure 6.11 and 6.12 display the gas temperature plotted as a function of pressure,

p, and absorbed microwave power, Pabs, respectively, with argon ﬂow rate equal to 100

sccm and methane ﬂow rate equal to 1 seem. As shown in Figure 6.11, the gas temperature

increases with the increases in pressure. At constant pressure, the higher hydrogen con-

centrations generally have higher gas temperature. This result is consistent with the data in

Figure 6.8. Above 80 Torr, the uncertainty of the gas temperature is 100K. At 60 and 80

Torr, the uncertainty of the gas temperature is 150K.

In Figure 6.12, the gas temperature increases with increases in absorbed micro-

wave power. The uncertainty of the gas temperature is 100K. The substrate temperature

increases with increases in either pressure or absorbed microwave power is shown in Fig-

ure 3.12 and Figure 3.13. Above 800 W, the uncertainty of the gas temperature is 100K.

Below 800 W, the uncertainty of the gas temperature is 150K.

287

3000 

 

H2=4 sccm

2900 ~

H2=2 sccm

H2=1 sccm

H2=0 sccm

 

 

O

o

It

o

ﬂ

c»

V

-

‘

1

-

-

-

-

-

-

.

 

 

2800 _

2700 —

2600 -

2500 -

2400 ~

2300 -

2200 -

)
K
(

e
r
u
t
a
r
e
p
m
e
T

s
a
G

 

2100 -

2000 -

1900

9

“

V

j?

0

o

*

v

 

1

‘

‘

50

100

150

200

250

Pressure (Torr)

FIGURE 6.11 : GAS TEMPERATURE VS. PRESSURE

AFIOO sccm and CH4=1 sccm by system II.

288

3000

 

 

H2=4 sccm

2900 -

H2=2 sccm

H2=1 sccm

H2=O sccm

 

 

 

2800 -

2700-

2600 —

2500-

2400—

2300

)
K
(

e
r
u
t
a
r
e
p
m
e
T

s
a
G

2200 -

If? 0

o

O,

c

V

o 11

11,, V

V

m 07

"

-

-«

-

1

-

1

.

-

 

2100 -

2000 -

1900

~1j> ‘0

_1

l

4

l

a

.

 

 

600

800

1000

1200

1400

1 600

Absorbed Microwave Power (W)

FIGURE 6.12 : GAS TEMPERATURE VS. MICROWAVE POWER

Ar=100 sccm and CH4=1 sccm by system II.

289

Figure 6.13 displays the growth rate plotted as a function of gas temperature with

an argon ﬂow rate of 100 seem, a methane ﬂow rate of 1 sccm, t: 8 hours. The same

experimental data points that are Shown in Figure 6.13 are also displayed in Figures 6.11

and 6.12. Thus, it is important to note that as gas temperature increases, the pressure and

microwave power are also increasing. Figure 6.13 shows that growth rate increases as gas

temperature increases.

290

A
g 10 t

°°
E

3‘1;
a:

5

o

E
11.5

5 -

 

 

o

*
,1

'V

r

V

V

09

8

.

1‘:

Q7.

O

o

0

E‘.
“0.628 R

a:
9

2
s-

70

42

0.314 g

 

-
I
I

0.942

15

1

1

1

1

 

 

o

r

H2=4 sccm

H2=2 sccm

H2=1 sccm

v H2=O sccm

 

1

 

0

v V.

O *1 ‘7

1

1

1

1

 

0.000

2000

2200

2400

2600

2800

3000

Gas Temperature (K)

FIGURE 6.13 : GROWTH RATE VS. GAS TEMPERATURE

Ar=100 sccm, CH4=1 sccm, and t=8 hours by system 11.

291

-
V
”
a
-
“
1
‘
-
'
.
.

Figure 6.14 displays the growth rate plotted as a function of substrate temperature

with an argon ﬂow rate of 100 sccm, a methane ﬂow rate of 1 sccm, t: 8 hours. It is impor-

tant to note that as substrate temperature increases, the pressure and microwave power are

also increasing. Figure 6.14 shows that growth rate increases as substrate temperature

increases. Between 60-100 Torr, the substrate temperatures were to low to be detected by

the back-Side temperature measurement using a pyrometer.

292

 

15

1

1

1

1

1

-
'
M
—
m

.942

A 10 ”

e

£050

z:

.5

o:

5

E

O
'5

13

 

o

1:

0

H2=4 sccm

H2=2 sccm

H2=1 sccm

o O

o

  
 

H2=0 sccm

v

1t

11'

o

1.

,

w

o

r

10.628 I"

a

8

e

S-

a?

1

_

5

7
0

4

A
0.314 =3

 
  

1

1

1

1

0

1

CL

.m0

500

550

600

650

700

750

800

Substrate Temperature (°C)

FIGURE 6.14 : GROWTH RATE VS. SUBSTRATE TEMPERATURE

Ar=100 sccm, CH4=1 sccm, and t=8 hours by system 11.

293

6.5.3 CN Emissions versus Controlled Nitrogen Impurity

Figure 6.15 displays the CN emission intensity plotted as a function of nitrogen

impurities with pressure at 120 Torr and gas chemistry ﬁxed at Ar:H2:CH4=100:4:1,

ft=105 sccm. As shown, the CN emission intensity Strongly increases as the controlled

nitrogen impurity increases from a few ppm to,1500 ppm.

30

I

I

I

 

I

I

I

I

o

o

o

25~

T;

g 20”

a.»

é

E 15~

<>

c

.2

.ﬂ

E1 10~

5

5—

  

0

o

0

01:00

1

l

1

1

r

1

1

-

T

-

—

l

 

 

0

200

400

600

800

1000

1200

1400

1600

N2(PPm)

FIGURE 6.15 : CN EMISSION INTENSITY VS. NITROGEN IMPURITIES

Pressure=120 Torr, Ar:H2:CH4=100:4:1, ft=105 sccm by system H.

294

-
'
1

"
"
"
"
“
r

6.6 Summary

The results of optical emission spectroscopy Show that the plasma properties are

inﬂuenced by the input variables: Ar/Hz/CH4 gas chemistry, deposition pressure, p, and

absorbed microwave power, Pabs. Figure 6.3 and Figure 6.4 Show that the experimentally

measured C2 emission intensity generally decreases as hydrogen concentration increases,

and is essentially constant as total gas ﬂow rate increases. As shown in Figure 6.5 and Fig-

ure 6.6, the C2 emission intensity increases as pressure and power increase.

Gruen et a1. believe that there is a close relationship between total growth rate and

the concentration of C2 dimer in the plasma and conﬁrm that C2 dimer is the growth Spe-

cies of the ultra-nanocrystalline diamond [Zhou 1998]. Based upon the comparison of

Figure 6.3 with Figure 4.17, and the results Shown in Figure 6.7, total growth rate is usu-

ally higher when the C2 emission intensity is higher except when the hydrogen

concentration is increased. These measurements indirectly support that C2 may be an

important growth species. This is no longer valid at higher hydrogen concentration. The

plasma deposition chemistry may change and maybe other radicals such as CH3 are also

present and inﬂuence the growth mechanism.

AS shown in the results of Figure 6.8, Figure 6.11, and Figure 6.12, the gas temper-

ature increases as hydrogen concentration increases, and increases as pressure and

absorbed microwave power increases. The comparisons of Figure 6.8 with Figure 4.18,

Figures 6.11 and 6.12 with Figures 3.12 and 3.13 Show that substrate temperature varies

with gas temperature in a manner similar to the measured gas temperature. Based upon the

295

-
'
“
‘
T
-
r

results shown in Figures 6.9 and 6.13, total growth rate is higher when the gas temperature

is higher. AS Shown in Figure 6.14, total growth rate is higher when the substrate tempera-

ture is higher. The gas temperature measured for the system II (1975K to 2960K between

60-240 Torr pressure regime) is higher than 1300 0C (1573 K, determined by Goyette et

al.). The different gas temperature may be caused by: (1) the reactor geometry difference

of these two systems (MSU-MPACVD system 11 versus ASTeX PDS-17 MPECVD sys-

tem), and (2) the higher temperature observed in MSU-MPACVD system II Should lead to

higher chemically active species concentrations and this reactor can be more efﬁcient than

the commercial ASTeX PDS-17 MPECVD system. That is increases in gas temperature

are expected to increase the dissociation of neutral species and produce higher chemically

active Species concentrations. The measured temperature (gas temperature and substrate

temperature) at the higher pressure regime probably explains why the ﬁlm growth rates are

higher when the gas temperature is higher.

The idea that molecular nitrogen becomes chemically active in the plasma dis-

charge is proven by the presence of the CN emission in Figure 6.15. It also indicates that

the molecular nitrogen interacts with methane in the gas phase inside the plasma dis-

charges. Therefore, nitrogen does change the bulk plasma properties.

296

-
"

7

Ultra-nanocrystalline Diamond Film Application

7.1

Introduction

Diamond’s marvelous physical, chemical, and electronic properties have attracted

researchers throughout the world. For instance, (1) its transparency over a wide wave-

length range and capability of withstanding high electromagnetic radiation power ﬂuxes

from laser beams and x-ray sources make it a good candidate as a wear-resistant, transpar-

ent, and protective coating for optical components and x-ray lithography masks [Zhu

1991] that can be operated in hostile environments; (2) its extreme hardness, Strength,

chemical resistance, low coefﬁcient of friction, high wear resistance, and excellent chemi-

cal and thermal Stability make diamond an excellent coating material for abrasive and

wear-resistant applications; (3) its high thermal conductivity, high hole mobility, and wide

band gap have the potential to fabricate active semiconductor diamond elements for high

power/high frequency and high frequency/high packaging density electronic devices that

can be utilized in high temperature and chemically harsh environments; (4) its biocompat-

ibility and high wear resistance indicates that diamond ﬁlms have potential for use in the

biomedical implantation and biosensors applications.

Nevertheless, diamond grown by the traditional Hz-CH4 CVD method usually has

very rough as-deposited surfaces. The rough surface is problematic in several ways: too

much light scattering for optical coating, increased coefﬁcient of friction for diamond

coated tools, and higher packaging costs for sensors. Moreover, post processing is time

consuming, expensive, and not reliable.

297

Thus, smooth nanocrystalline diamond may be a very desirable material that

enables many applications of CVD diamond. Recent years, an increasing amount of

research has been devoted to the applications of nanocrystalline diamond: (1) RF MEMS

and retinal implants by J.A. Carlisle et al.; (2) N-type conduction for nitrogen-doped

nanocrystalline diamond ﬁlms by CA. Williams et al., by S. Bhattacharyya et al., and by

Y. Dai et al.; (3) tool coatings by N. Salk et al. and by RH. Sun et al.; (4) optical coatings

by K. Yoshinori et al.; (5) SAW (Surface Acoustic Wave) devices by Y.K. Liu et al.; (6)

biomedical applications by A.N. Jones et al., by M. Amaral et al., and by Y. Kousar et al.;

and so on [Diam 2003].

In this chapter, three ultra-nanocrystalline diamond applications are brieﬂy

explored: (1) optical coatings for silicon substrates (Section 7.2), (2) tribological coatings

for SiC chemical process pump seals (Section 7.3), and (3) SAW device applications (Sec-

tion 7.4). The experimental fabrication, experimental measurements, results, and

summaries for each application are presented in the sections below.

7 .2 IR Optical Transmission Window

7.2.1 Introduction

Chemical resistance, wide optical transmission band, high resistance to abrasion

and high index of refraction are some of diamond’s unique properties. These properties

make diamond of interest as an optical coating for semiconductors, both for the purpose of

protection and for modifying the reﬂectivity of semiconductor materials. Silicon is used as

a semiconductor substrate material for a variety of optical applications including infrared

298

-
'
m
‘
“
"
-
I

(IR) windows. At an air/silicon interface, the reﬂective loss for photons ranges from

approximately 30% at long (IR) wavelengths to 50% at Short (blue-visible) wavelengths.

The loss is compounded due to reﬂection at front and back surfaces [Rein 2000].

High-quality single—crystalline diamond is transparent from ultraviolet, through visible

and infrared, and into the microwave portion of the electromagnetic spectrum. There’s a

slight absorption in the infrared, due to C-C bond vibrations, from 2.5 pm to 6 um (with

the largest peak at 5 pm wavelength corresponding to an absorption coefﬁcient of approx-

imately 12 cm'1 [Zait 1998] and [Thom 1993]), which is weak and maybe negligible if the

ﬁlm is thinner than 250 um [Ulcz 1998]. Therefore, diamond ﬁlm coatings are good candi-

dates for optical applications that require the good chemical and mechanical surface

properties of diamond and with the appropriate thickness of the coatings, it can even

improve the optical performance.

The application of polycrystalline diamond ﬁlms synthesized with good optical

quality by regular CH4/H2 CVD has proven difﬁcult [Angu 1998], [Ong 1989], and [Wang

1990]. The difﬁculties arise due to scattering from the rough surface of the polycrystalline

diamond ﬁlms and the absorption from the defects and impurities inside the bulk of the

crystals. However, the work of Ulczynski and Reinhard et a1. [Ulcz 1998], [Rein 2000],

who make use of the traditional CH4/Hz (sometimes with additional 02) polycrystalline

diamond growth has achieved as-deposited, non-polished CVD polycrystalline thin dia-

mond ﬁlms with near ideal optical transmission throughout the visible and near— to mid-

infrared. This thesis work investigates the use of UNCD for a Similar optical application,

i.e. it takes advantage of the smooth ultra-nanocrystalline diamond surfaces for making

299

optical coatings on silicon substrates with as grown ultra-nanocrystalline diamond thin

ﬁlms. The hydrogen free environment has also made it a clean and safe deposition proce-

dure to produce diamond coated optical components.

The goal of this UNCD application investigation is to study the optical properties

of these ﬁlms. This portion of Chapter 7 addresses the results of the investigation of the

optical properties of ultra-nanocrystalline diamond. The methods for producing the sam-

ples utilized in the research and the experimental conditions are brieﬂy described in

Section 7.2.2. The theoretical model and IR transmission measurements on the samples

are described in Sections 7.2.3.1 and 7.2.3.2, respectively. Experimental results in the

infrared, which are compared with theoretical calculations for the as grown single layer

(diamond) and two layers (diamond/SiOz) and for one-side and both-side optical coatings

on silicon substrates are presented in Section 7.2.4. Finally, Section 7.2.5 summarizes the

results of the ultra-nanocrystalline diamond IR window application.

7 .2.2 Experimental Method

Infrared transmission through single-layer diamond coated and two-layer dia-

mond/SiOz coated silicon substrates was experimentally investigated. A sketch of each of

the different UNCD coated substrates is shown in Figure 7.1. Substrates are 3” diameter,

254-305 pm thick, N/phos-type, (100) Silicon wafers, both side polished, with a resistivity

of approximately 10 S2 -cm. A SiOz layer was thermally grown in a furnace with high

purity dry oxygen at 1100 °C. The oxide thickness was measured by ellipsometry. Prior to

diamond deposition, either directly on silicon or on a silicon dioxide layer, substrates were

prepared by scratch seeding (Procedure is described in Section 3.3.1.2.). Ultra-nanocrys-

300

talline diamond ﬁlms for optical coatings were grown using system 11 with argon,

hydrogen, and methane gas mixtures. The deposition details were described in Sections

3.2.2.2, 3.2.3.2, 3.2.4.1, and 3.3.2.2. Although the reactor is capable of operating in a wide

experimental ﬁeld map (see Section 3.2.5), in this study, the experimental deposition

parameters were conﬁned to the following regime: the deposition pressure was 120 Torr

and 160 Torr, the microwave power was between 910-1314 W, the gas chemistry was

Ar/HZ/CH4=(99-95)/(0-4)/1 (%). The resulting ultra-nanocrystalline diamond ﬁlms have

RMS roughness of 7-18 mm depending on the deposition conditions.

301

Diamond

Single layer one side

(SLOS)

Diamond

Diamond

Single layer both sides

(SLBS)

Diamond

Q. Two layer one Side

.

(TLOS)

Diamond

    

1

Two layer both sides

Diamond

Si

   

(TLBS)

FIGURE 7.1 : A SKETCH OF CROSS-SECTION VIEW OF UNCD COATED

SI SUBSTRATES

  
 

302

7.2.3 Theoretical Calculation and Transmission Measurement

7.2.3.1

Theoretical Calculations

The theoretical calculations were based on the following assumptions [Huan

2001].

0 A planar, lossless diamond ﬁlm of uniform thickness

0 A normally incident EM wave

0

-

Transmission is only limited by specular reﬂection

Surfaces of the ﬁlms are smooth.

The surface roughness (7 nm to 18 nm) of the ﬁlms studied here is ignored in the

theoretical calculations because the infrared wavelength (2.5 pm to 14.3 um) utilized in

the measurements is much larger than the surface roughness. Therefore, the assumption of

specular reﬂection is meaningful. When using the algebraic equations [Ulcz 1998], the

N-layer problem is complicated and difﬁcult to carry out, because it involves summing up

N inﬁnite series, each term of which is itself a series. Another simpler method of analysis,

called Matrix Approach, was invented by Rouard in 1937 [Nuss 1976]. Rouard discovered

that a multiple-layer ﬁlm could be analyzed by representing each layer by a 2x2 matrix

Mj. This is the preferred method and is used in this research.

The matrix, M, is the characteristic matrix of the coating layer and is only a func-

tion of the layer parameters: the refractive index, n, the phase shift, 8, and the layer

thickness, t. That is

303

-
'
ﬂ
i
t
‘
é
ﬁ
l

I—

'1

sin8-

 c035.

i

Mj =

J

Ilj

 

 .J

The phase shift is deﬁned as:

21:

5]. — Tnjtj cos 0j

For normal incidence, cos0j = 1 .

21:

J

A “I J

When there is more than one layer, the matrix has the following form:

1

M = ['[Mi

i

(7.2)

(7.3)

(

)

(7.5)

If we simply take the product of the matrices representing each layer, then the effect of a

combination of layers is obtained. Special matrices used are

n

—l

1

[ O

Ln
L o

and [n] which represent top and bottom matrices;

1

f

n

—1 ,

O

is the top media matrix, where 110 is the index of refraction of air; and

n

l

_ 0

.1

P

] is the bottom media matrix, where nf is the index of refraction of the substrate.

 

n
L f

304

The reﬂectivity of a substrate with coating layers is given by [Nuss 1976]:

2

 

a and b are determined by the matrix method.

11

—1

[a] =

0

-M- l

b

n0 1

nf

(7.6)

(7.7)

With all the assumptions described at the beginning of this session, the ideal

smooth, no-loss ﬁlm transmission calculation is then given by:

T = l—R

(7.8)

When the light is not sufﬁciently monochromatic to produce interference effects

over the thickness of the semiconductor substrates, the through transmission, TT, is given

as:

_(1-RF><1-RB>

T _

 

T

(l—RR

FB)

(8.10)

RF and RB are the reﬂections corresponding to the front and back surfaces respectively.

Equation 8.10 was the equation that was used for the theoretical calculation in this thesis.

 

305

-
'
“
‘
7
.

.
r

 

7.2.3.2 IR TIansmission Measurements

A Beckman Spectrophotometer model IR4220 was used to perform the transmis—

sion measurements in this thesis research. The IR4220 Infrared Spectrophotometer has a

range of 4000 cm'1 to 250 cm'1 and is equipped with a nichrome source and a thermocou-

ple detector. The beam dimension is about 10 mm x 14 mm. The grating in the IR4220

changes at 650 cm'1 and mostly, it has higher dispersion values between 250 cm”l and 650

cm]. Therefore, the measurements performed here were from 4000 cm'1 to 700 cm], (2.5

pm to 14.3 pm). The measurements were made with a double beam at a scanning speed of

150 cm'llmin. The instrument response time was set at 1 second. The measured transmis-

sion data of the ﬁlms were normalized by the transmission value of air to obtain the

percentage of the transmission. The spike around 2333 cm'1 («429 11m) in the measure-

ments is random and has been conﬁrmed with other users that it’s caused by the

instrument.

Equation 8.11, the Sellmeier equation is used to calculate the wavelength depen-

dence of the refractive index for diamond [Klei 1999].

"d2“ =

4.33567.

2

+ 03306-7.

2

  

12 — (0.1060)2

112 — (0.1750)2

(8.11)

Silicon dioxide refractive index is assumed to be a constant of 1.46. Silicon refrac-

tive index is 3.45. Air refractive index is 1. An uncoated, both-side-polished Si wafer has

the transmission value around 53%.

306

7.2.4 Experimental Results

7.2.4.1

Introduction

The following experimental results are divided into four groups: (1) single layer

one side (SLOS) (Section 7.2.4.2), (2) single layer both sides (SLBS) (Section 7.2.4.3), (3)

two layer one side (TLOS) (Section 7.2.4.4), and (4) two layer both sides (TLBS) (Section

7.2.4.5). A sketch of each of these structures is shown in Figure 7.1. The transmission

measurements were compared to the theoretical calculations. From Section 7.2.4.2

through Section 7.2.4.5, the dotted curves represent the experimental measurements and

solid curves represent the theoretical calculations determined from the equation 8.10.

7 .2.4.2

Single Layer One Side With Ultra-nanocrystalline Diamond Optical

Coatings

Figure 7 .2 displays the optical transmission measurements compared to the theo-

retical calculations for two different deposition pressures. The deposition conditions were:

(1) gas chemistry of Ar/HZ/CH4= 100/4/1 sccm, (2) ﬁlm thickness=850 nm, and (3) depo-

sition pressure of 120 Torr shown in Figure 7.2(a) and 160 Torr shown in Figure 7.2(b).

As shown in Figure 7.2, there is no difference in the optical transmission between

the sample deposited by 120 Torr and the other by 160 Torr. The wavenumbers to obtain

maximum and minimum transmission are around the same places when ﬁlm thickness is

about the same. The theoretical calculations and experimental results are in good

agreement.

307

 

j

120Torr

I

.

I

I

‘

I

7

I

g

E

‘3. .1 ............................ _

:5; :22

t

I ‘1 1‘;

u:

-: ti»: :

g... ..’

-

60.

:$..V.. ..

.. 9

“1....

a D32

0‘

.
:

..

...................................

1'

..

'1
.

if? 1

‘

3

EL- .

:4NC;

100

90

80

I

7O

'.

.

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

50

4O

3O

20

10

 

40

3O

20

1O

 

 

 

5

;

;

RMS Roughness=15 run

 

1000

1500

2000

2500

3000

3500

4000

 

I

160Torr

I

:

I

I

'

I

I

r

100

(1))

90

80

80

a ........

.

70'

‘

’3

'.§..g‘9‘ 'ﬁ: ..................................................................

’ ‘0

.

..

i

e—I

, .w» - “1115-.

0‘ ’

-

I

s

o - "5"
.

' ’

60

3,.. .3.; ............... {:3

...................................................

$3
..‘

‘

..

_.

a”

.

50

'%I

:

..‘.
g.

.

h
'

.

.

.‘3-1 ......................................... q

9.:

.

.

.

'

~

..111....--' ...... . .......................................... J

RMS Roughness=14 nm

 

1000

1500

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.2 : OPTICAL TRANSMISSION OF SINGLE LAYER ONE SIDE

Ar/Hz/CH4=100/4/1 sccm and Film=850 nm by system 11.

308

 

Figure 7.3 displays the optical transmission measurements and associated theoreti-

cal calculations as the H2 vol% was varied from 0-4%. The deposition conditions were:

(1) pressure at 120 Torr, (2) Ar/Hz/CH4=100/4/l sccm, Ar/Hz/CH4=100/1/1 sccm,

Ar/Hz/CH4=100/0/1 sccm, and (3) ﬁlm thickness=657 nm.

Figure 7.4 displays the optical transmission measurements and associated model

calculations as the H2 vol% was varied from 1-2%. The deposition conditions were: (1)

pressure at 160 Torr, (2) ArﬂI2/CH4=100/2/l sccm, Ar/HZ/CH4=100/1/1 sccm, and (3)

ﬁlm thickness=480 nm.

Figure 7.3 and 7.4 indicate that with Ar=100 sccm and CH4=1 sccm, varying the

hydrogen ﬂow from 0-4 seem, the ﬁlms’ optical properties do not vary. The wavenumbers

to obtain maximum and minimum transmission are around the same places when the ﬁlm

thickness is about the same. Again. the experimental measurements are in excellent agree-

ment with the theoretical calculations.

309

 

J
_
.
.
.

1 00

 

1

f

1

(a)

Ar/H2/CH4leO/4/ 1 sccm

1

90-

 

30 —

20 ~

’6‘
E
c:

.2

(I)

ﬂ

E

(D

5

I...
1-1

...........................................

_

-

1 o ._

......................................................

RMS Roughness=13 nm

 

0

1 000

1 500

2000

3000

3500

4000

1 00

I

I

(b)

90 Ar/H2/CH4=100/1/1 sccm

Ti

.

V

 

$3

v

g

.2

.2

g

§

11.4
I-*

80 ..

,

!

.......

__,

70533.}.12 2.2 "

........................................................... _

:‘E 3

r.‘

.

: if

;

..'

60'. .‘.;1¥»

:

.............................. 2.

—

5' j' g

.

50 L_'.

40 L

30 ’

20 _

1 O ,_

.

‘

.

.

.

.

.

..................................................................... .1

  
 

RMS Roughness=1l nm

................................................... _

...............................................

.........................................................

_

.1

,

0

1 000

1 500

2000

3000

3500

4000

1 oo

 

1

1

(C)

901— kHz/CH4: Im/Oll .Sccm

................................................

80 ,_

.

.

........... ..1

A

Q

E

70 ..‘”..‘1

30.5-3‘ ‘..

........................................................ _

3

3

e

E

- °

,

E

‘34 " ‘3

‘

.

is ‘ 2‘ “at”, ‘

’_.

c

60 _

'3‘ 3 3'.

.

.

2

.................................

.9.

.2

15.;

c:

e

F

313‘

.

50 ..‘

4o _

.

.

.

.

.................................................... .1

............................................................ _

  
 

............................................................. .1

RMS RoughneSS=11 nm

1 ...................................... T

............

. .1

3

.

.

.

.

.

.

,

.

.

.

30 ._ .

L20

1 0 ._

0

1 000

1 500

2000

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.3.' OPTICAL TRANSMISSION OF SINGLE LAYER ONE SIDE

Pressure=120 Torr and Film:675 nm by system 11.

 

310

 

 

 

100 ﬁ

I

(a)

/H2/CH4=100/2/1 sccm

r

f

I

‘

l.

'

I

'

90,_.,.

.

.....

...........

......................................................... _

80-

-

g3

E“
.2

.2
E

W

C‘.

E

E“

30-

.............................................................. -

20-

10...

 
  

RMS Roughness=10 nm

............. ..J

-

;

1000

1500

2000

2500

3000

3500

4000

 

100

I

I

W

(b)

Ar/I-I2/CH4=100/;1/1 sccm

I

:

I

T

90.. ....... g .............. ......................................................................... _.

SR

‘5
.2

.."re’
E

U)

c:

E
F<

 

 
' RMS Rdﬁghﬁésggé'ﬁih'   

10” 

20 l

c-d

‘

1000

1500

2000

2500

3000

3500

4000

Wavenumber (cm’l)

FIGURE 7.4 : OPTICAL TRANSMISSION OF SINGLE LAYER ONE SIDE

Pressure=160 Torr and Film=480 nm by system H.

 

311

Figure 7.5 displays the optical transmission measurements compared to theoretical

calculations for two different ﬁlm thickness. The deposition conditions were: (1) pressure

at 120 Torr, (2) Ar/H2/CH4= 100/2/1 sccm, and (3) the ﬁlm thickness=870 nm and 620

nm.

Figure 7.6 displays the optical transmission measurements compared to theoretical

calculations for two different ﬁlm thickness. The deposition conditions were: (1) pressure

at 120 Torr, (2) Ar/H2/CH4= lOO/l/l, and (3) the ﬁlm thickness=430 nm shown in Figure

7.6(a) and 675 nm.

As shown in Figure 7.5 and Figure 7.6, ﬁlms produced with the same deposition

pressure and gas chemistry have a similar optical transmission behavior. As expected from

the theory, the optical transmission performance mainly depends on the ultra-nanocrystal-

line diamond ﬁlm thickness.

312

  __.. Y .... l.

_

-

.

_

, ..

-L

-«. .

m~__nh’._-c;".;..1_:_._.‘.’1__’... ‘

..

‘

100

r

 q

I
-
c

1
.

(a)

870nm

;

.

2

5

90-.....gw

............... ............. .1

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

703,123.

60 .- ..... g .....' ......... an:

"

..1 .........

...........................................

50.. 7

........: .....

.....q

40

3o__......:.4.,4

,,,,,

.

............. _

20_ .1

q

l

....ﬁ. H..1H§.....HM.§

......... _

W ...... g1

is

j

T

.....

1

;

'

.............. _

RMS Roughness=12 nm

 

1000

1500

2000

2500

3000

3500

4000

100

T

(b)

620nm

F
».

.
g

l
_

'.
2

 

90...‘ ...... ...........

......

............. ..

w...

_

............. M

70_ ..... zq ..... gﬁj'gigM..'?;.;2...... ..‘..TH.HH.HH.9H.HH.HH_% ............. _

60 9'1.

;l!.1HM§aHH.”_ _§ ..... iQ.f M

:

..H“.€”H.HH.HH§ ....... _ ......

50.; ______

.

l

.

............ J

40L. ......

,

.Q

.A

......... Y .......

............. _

30....

l.

..

1,,

H

......... ..

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

 

 

‘~

"

.
3

Y
g

 

 

1O,.......L...W

..... .. ............. ..l

,.

'

'

RMS Roughness=11 nm

 

1000

1500

2000

2500

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.5 : OPTICAL TRANSMISSION OF SINGLE LAYER ONE SIDE

Pressure=120 Torr and Ar/H2/CH4=IOO/2/1 sccm by system 11.

313

 -_—————

A

A“

-

 

1oo

.

a

j

.

,

,L

 

(a)

430nm

:

‘

:

90......

.....................................................................

...

80... _

.
‘

E

i;
‘IA

.i
.

........................................................

..i

Z
-

i
‘

70.... . ..’. .

’

...)... .33..' I: ........

..............................................

.—

’

,n

.‘ Lt‘

:3 a! i .' “ '

*

603:9";

'. v ,2." .§ ...................................................

 

= 3; i

9‘:

,

50_ :..‘

:

3.

.............

c

'

.1

..4

A

§

x;

O

a

g

e

E

g

F
f.

40,. ...............................................................................

_

.

30. ‘

......................................................... _

20_

10_

  
 

. a.”..H..SUN.“HH.HLH.HHH.HQH.HHH.HH% ............. _

......................... .a

RMS Roughness=9 nm

J_

l

l

‘

;

0

1500

3000

3500

4000

 100

I

I

I

i
;

I
;

0))

675nm

90....

.................................................................... —

80.. ..... :.

...................................................................... _‘

s

5‘

.,

'54 i:

70"i";""" ." .,

...'

........................................................... ..

~ g :

. ‘ :¢,c~ m

;~,~n

‘$ 3‘ .0

33

:°.

:

60' ..‘

. .....H: ............

‘ ..................................... ...

‘4

“~:

$33

g

'

3

.

50- ,

..

..................................................... .-

40.... ..

.

...

................................................................... -

30...

.......................................................... _

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

20L-

10....

  
 

RMS Roughnoss=ll nm

............. _

.......

_

;

1000

1500

2000

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.6 : OPTICAL TRANSMISSION OF SINGLE LAYER ONE SIDE

Pressure=120 Torr and Ar/H2/CH4=lOO/1/l sccm by system II.

314

7.2.4.3

Single Layer Both Side With Ultra-nanocrystalline Diamond Optical

Coatings

Figure 7.7 displays the optical transmission measurements compared to theoretical

calculations for SLBS and SLOS substrates. The deposition conditions were pressure:

120 Torr and Ar/HZ/CH42 100/2/1 sccm. As predicted from the theoretical calculations,

the SLBS substrate improves the optical transmission by 33% over the SLOS substrate

around 6.15 pm wavelength.

Figure 7.8 displays the optical transmission measurements compared to theoretical

calculations for three different SLBS substrates. The deposition conditions were pressure

at 120 Torr and Ar/HZ/CH4= 100/4/1 sccm. Sample(a) ﬁlm thicknes=450 nm and 485 nm,

sample(b) ﬁlm thicknes=675 nm & 590 nm, and sample(c) ﬁlm thicknes=895 nm & 405

nm. Theoretical calculations and experimental results again match well.

315

 

 

620 nm & 685 nfn; RMS Roughness=ll and 13 nm

my layer both sides

.
f

100

90

 

 

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

 

4O ........................................................... _

30

........................................................... -

20W

.............. -

10- ............................................................................................... 4

 

0

 

 

1000

1500

2000

2500

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.7 : OPTICAL TRANSMISSION OF SINGLE LAYER

Pressure=120 Torr and Ar/H2/CH4=100/2/1 sccm by system H.

316

100 

450nm & 485 nm

3°,

3

3

r

90- ....... . ................... 3...............

1................................................ ..

E

.

t

:

_.§. . .3.

...._

O...

'. Ox 1 ............. ...

80

2}

S.

43: -: 33 "

g“

..

701‘9;‘°-'“"§ .

.................. _.

-95.

._

‘1

8:

'

a;

.:_.

.......................................

_

—

—

—

 

60 ‘-..:

5013'?

40—

30E“

3v
A 9

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

A 0v

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

20...

10"“

 
  

ERMS Roughness=12 and 15 nm

......................................... .—

...................................................

.

0

1 000

1 500

2000

2500

3000

3500

4000

100

" v
A 0

 

r

r

r

v

675 nm&590 nm

5

5

~

r

 
 
  

ERMS Roughness=13and 13 nm

10E .

I

1

‘

0

1 000

1 500

2000

2500

3000

3500

4000

 

100

V

V

I

ﬂ

895 nm&405 nm

:

T

-

V

.

90,. ;

I ,. , .............. .-

80...!

£3; ..':

................................. _.

S

7o ',:-

p".8..... {.1 ..... .

f

.

......... .....

“‘3 ...........................

,

.

............ _

‘1 a; r ‘7;

60 3:." .

.

3

.3.

E

~

.

..

,.

'i

40...

............................................................ —.

30..

...................................................... J

20..

10..

 
  

RMS Roughness=15 and 18 nm

...............................................

....................

.__E

0

1 000

1 500

2000

2500

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.8 : OPTICAL TRANSMISSION OF SINGLE LAYER BOTH SIDES

Pressure=120 Torr and Ar/HZ/CH4=1OO/4/1 sccm by system II.

317

7.2.4.4 Two Layer One Side Diamond/SiOZISi/SiOz Optical Coatings

Figure 7.9 displays the optical transmission measurements compared to theoretical

calculations for TLOS UNCD coated substrates and SiOz/Si substrate. The deposition

conditions for the ﬁlm transmission measurement shown in Figure 7.9(a) were: pressure:

120 Torr and Ar/H2/CH4= 100/1/1 sccm on SiOzlSi substrate with Si02 layer thick-

ness=74.4 nm. The diamond ﬁlm thickness is 650 nm. The SiOz layered on Si wafer was

thermally grown in a furnace with high purity dry oxygen at 1100 °C. The transmission

measurement of SiOZISi/SiOZ substrate (before diamond deposition) was displayed in Fig-

ure 7.9(b).

The optical transmission measurements again agree with theoretical calculations.

The transmission values are relatively low because neither the oxide or diamond thickness

have been optimized for anti-reﬂection performance in this spectral range. The absorption

around 1100 cm”1 (9 mm) displays an oxide absorption, due to the stretching and bending

mode excitations in the oxide.

318

100 

Si02=E74.4 nm & biamond=650 nm

I

I

'

 

1

i

1

1

RMS Roughness=9 nm

 

 

1000

1500

2000

2500

3000

3500

4000

 

100 3102;744 nm 7

f

f

T

T;

80""“" .............. :.................................................................... _

.
60 ...! {~

,

1:

5°

“I“; :91. .......... ..................................... .

40- ...................l ........................................... -

A N V

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

A U v

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

20_ ........................................... _

 

 

 

1000

1500

2000

2500

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.9 : OPTICAL TRANSMISSION OF TWO LAYERS ONE SIDE

Top: Pressure=120 Torr, Ar/l-12/CH4leO/1/1 sccm by system H.

Bottom: Thermally grown with high purity dry oxygen at 1100 °C in a furnace.

 

319

7.2.4.5 Two Layer Both Side Diamond/SiOZISi/SiOleiamond Optical Coatings

Figure 7.10 displays optical transmission measurements compared to theoretical

calculations for TLBS substrate and TLOS substrate. The deposition conditions were:

pressure at 120 Torr and Ar/H2/CH4= 100/1/1 sccm on SiOZISi/SiOz substrate with Si02

thickness=74.4 nm. The diamond ﬁlm thickness is 650 nm and 340 nm. As expected, the

transmission measurement of the TLBS, Diamond/SiOZISi/SiOleiamond, sample dis-

plays much improved transmission when compared to the TLOS, Diamond/SiOZ/Si/SiOz,

substrate.

The optical transmission measurements agree well with theoretical calculations.

The optical transmission is higher than that of the two-layer-one-side Diamond/5102 opti-

cal coatings but generally lower than those of single-layer—both-side ultra-nanocrystalline

diamond coated ﬁlms (Compare Figure 7.10 with Figure 7.8(a) and Figure 7.8(b).). It is

because neither the oxide or diamond thickness have been optimized for anti-reﬂection

performance in this spectral range. The Si02 layer is not a good Fabry-Perot medium if the

window is designed for a wide range of infrared wavelength because Si02 absorbs heavily

in portions of IR wavelength.

320

100 

l

Si02;74.4 and 711.4 nm & D1amond=65b and 340 nln

Two layer both sides

3

RMS Roughness=9 and 7 nm

 

 

)
%
(

n
o
i
s
s
i
m
s
n
a
r
T

30-

 
 

1o- 

20-

 

1000

1500

2000

2500

3000

3500

4000

Wavenumber (cm'l)

FIGURE 7.10 : OPTICAL TRANSMISSION OF TWO LAYERS

Pressure=120 Torr and ArﬂI2/CH4=100/1/1 sccm by system 11.

321

 

 

—
'
ﬂ
“
w

7.2.5 Summary

Ultra-nanocrystalline diamond coatings applied as a IR optical transmission win-

dow have demonstrated that the ultra-nanocrystalline diamond has similar optical

properties to high-quality diamond over 2.5-14.3 um wavelength. This is demonstrated by

the fact that in all cases, the experimental measurements ﬁt well with the theoretical calcu-

lations. In the theoretical model, a diamond slab (a planar, lossless diamond ﬁlm of

uniform thickness) was assumed. Thus, grain boundaries and other possible losses in the

ﬁlms were ignored. It’s shown in the experimental results that the many ultra-nanocrystal-

line diamond grain boundaries have little effect on the experimental performance.

As expected, the IR optical transmission is sensitive to the thickness of the coat-

ings and the differences in thickness of both sides. With the aid of theoretical calculations,

the design and fabrication of ultra-nanocrystalline diamond IR optical windows can be

achieved. Over a wide range of experimental deposition conditions, i.e. argon ﬂow rate

ﬁxed at 100 sccm, methane flow rate ﬁxed at 1 seem, and varying the hydrogen ﬂow from

O to 4 sccm at 120 Torr or 160 Tor, there is no noticeable effect in the IR optical transmis-

sion. The upper limit for optical absorption is about 494 cm’l, which is determined with

the ﬁlm thickness of 405 nm and approximately 2% difference between the experimental

measurements and the theoretical calculations.

It was demonstrated that the both-side coated ultra-nanocrystalline diamond IR

windows have higher transmission than those of one-side coatings. In this research, two

layered coatings do not have higher transmission than single layer because the thickness of

diamond ﬁlms and SiOz layers were not optimally designed. For operating over a wide

322

l

J
.
.
.
.
.
.
.
.
2

span of infrared wavelength, Si02 probably is not a good choice. The dip of the transmis-

sion in two-layered coatings around llOOcm'1 (9 pm) is an oxide absorption, due to the

stretching and bending mode excitations in the oxide.

There is no visible IR—active single phonon absorption (1000 cm'1 to 1500 cm’l) in

the measurements, therefore the ﬁlms are absent of signiﬁcant defects or impurities. The

intrinsic optical properties of diamond are associated with the arrangement of the carbon

atoms on a diamond lattice. Since the thickness of the ﬁlms is very thin, the multi-phonon

absorption (2 pm to 7 pm, about 1428 cm"1 to 5000 cm‘l) is not seen in the measurements

because the weakness of the absorption. However, the absorption around 2830 cm'

1to

2925 cm], due to CHx has been observed in several ﬁlms. This indicates that there may be

hydrogen exists in the ﬁlms, which is usually predominantly bonded to sp3-carbon in the

grain boundaries or at the surface.

7.3

Tribological Application-Seal Coatings

7.3.1 Introduction

Diamond ﬁlms grown by chemical vapor deposition have attracted considerable

scientiﬁc interests due to their excellent properties, like high hardness, high wear resis-

tance, and good tribological properties [Asma 1999], [Asma 2001], [Gabl 2000], [Lee

1999], and [Neer 1998]. A wide range of research on the engineering applications of dia-

mond coating, including oiless bearings, seals, and cutting tools has been generated.

Having a smooth coated surface in tribological applications is extremely important for dia-

mond coatings in sliding contact with metals because the friction decreases as the surface

323

J
'

'

f
‘

;
”
t
a
‘
n
a
F

roughness decreases [H01] 1994], [Kohz 1994], and [Bogl 1995]. A reduction in wear rate

was found when smoother diamond coating was sliding with mating AISI 52100 steel ball

[Kohz 1994]. Casey M. and Wilks J. developed theories that explained the friction mecha-

nism of diamond versus the surface roughness. They interpreted the friction in terms of the

force required for the asperities on one surface to slide over those of the other [Case 1973].

Thus, very smooth ﬁlms are desirable. During the last few years, developing diamond ﬁlm

coatings with a smooth surface for tribological applications has been attempted. The atten-

tion has been paid to the nanocrystalline diamond ﬁlms [Holl 1998] and [Erde 1999].

The aim of this research is to explore the applications to ultra-nanocrystalline dia-

mond coatings on SiC chemical process pump seals. The experimental method and

measurements utilized in this research are brieﬂy described in Section 7.3.2. The primarily

experimental results are presented in Section 7.3.3. Section 7.3.4 discusses and summa-

rizes the work of ultra-nanocrystalline diamond coatings on SiC seals.

This application investigation was the results of the collaboration between Mate-

rial Science Chemistry Division of Argonne National Laboratory (ANL), Flow Serve

Company (FSC), and the author. ANL and FSC supplied the SiC seals and agreed to per-

form the experimental measurements and mechanical tests. The author deposited UNCD

ﬁlms on the SiC seals.

324

7 .3.2 Experimental Deposition Method

The substrates used for ultra-nanocrystalline diamond coatings were 30 um lapped

SiC seals with an outside diameter of 4.9 cm, inside diameter of 3.55 cm, and height of

1.55 cm. The thermally ﬂoating substrate holder set-up is shown in Figure 7.11. As shown

in Figure 7.11(a), a moly substrate holder was designed and built by MSU researchers. It

was designed to allow the pump seal, shown in Figure 7.12 to be placed inside the cut-out

region, i.e. the grey area of Figure 7.1 1(a). Once placed in the moly holder, only the por-

tion of the SiC seal that was to be coated was exposed to the deposition discharge. See

Figure 7.11(b). After deposition, the seal was removed from the substrate holder and only

the top surface of the seal shown in Figure 7.12 was coated with UNCD.

J
V

"
1
-
‘
“
“
"
"
l

 

325

(a)

Top view of the seal substrate holder.

A SiC seal was inserted into the grey area.

(b)

Plasma

Quartz Dome

SiC Seal

    
  

Baseplate

Gas Inlet

 

 

  

Holder Baseplate

V WE

 

Quartz Tube A Metal Tube

Substrate Holder

1

FIGURE 7.11 : SEAL COATING SUBSTRATE HOLDER SET UP

SiC seal coating thermal ﬂoating substrate holder set-up.

 

326

T— - -.-. .--” __- ﬂ ‘

 

Prior to diamond deposition, the top surface of the SiC substrate was prepared by

scratch seeding (Procedure is described in Section 3.3.1.2.) Ultra-nanocrystalline diamond

coatings for SiC seals were deposited by MSU-MPACVD reactor system 11 with

Ar—Hz-CH4 plasma discharges. The details were preciously described in Sections 3.2.2.2,

3.2.3.2, 3.2.4.1, and 3.3.2.2. These exploratory experiments were performed under the fol~

lowing deposition conditions: pressure of 160 Torr, microwave power between 1108-1226

W, gas chemistry conﬁned within Ar/H2/CH4=(99-95)/(O-4)/1%, the total gas ﬂow rate

between 101-105 sccm, and the deposition time between 8—24 hours. Only three SiC seals

were coated for this investigation.

7 .3.3 Experimental Measurements and Results

The Raman spectroscopy characterization indicated that the deposited ﬁlms on the

30 um lapped SiC seals were ultra-nanocrystalline diamond. Before deposition and seed-

ing, the typical surface roughness of each seal was on the order of a few to 10’s of

micrometers. A typical ultra-nanocrystalline diamond coated SiC seal is shown in Figure

7.12. The visual appearance and the inspection under optical microscopy suggest a very

uniform coating.

327

7

-

 

 

FIGURE 7.12 : A ULTRA-NANOCRYSTALLINE DIAMOND COATED SIC SEAL

A proﬁlometer (interferometric imaging), at MSC Division of ANL was used to

perform the surface roughness characterization. A typical surface proﬁle of the diamond

coated SiC seal is shown in Figure 7.13. The scanned area is 5252x3985 um2 (~20 mmz).

As shown in Figure 7.13, the surface roughness is within i 0.63 pm, even though the ini-

tial surface roughness of SiC substrate was on the order of a few to 10’s of micrometers.

 

328

I
"
m
“
‘
“
“
F

 

 

_-=..—.--—- h... .__. '

3.14

1.88

0.63

-O.63

-1.88

-3.14

3.14

1.88

0.63

-0.63

-l .88

-3. 14

)
m
u
(

t
h
g
i
e
H

e
c
a
f
r
u
S

X Proﬁle

 

 

1050.

2101.

3151.

4201.

5252.

 

Y Proﬁle

1

 

  

7917.

1594.

2391.

3188.

3985.

Position (um)

FIGURE 7.13 : SURFACE PROFILE INSPECTED BY INTERFEROMETRIC

Pressure=160 Torr, Ar/Hz/CH4=100/2/1 sccm and deposition time=8 hours

IMAGING [HUAN 2001]

by system II.

The ultra-nanocrystalline diamond ﬁlms deposited by MSU—MPACVD reactor sys-

tem II with deposition pressure of 160 Torr, Ar/HZICH4=100/2/l sccm, and depsoition

time of 8 hours passed the adhesion test but did not perform well in the wear durability test

on the water pump at ANL (The wear resistance test was taken against SiC counterface).

329

—7‘nﬁ '- ~L_.

”WA, ._____,.___, ,_ 

1

As we have learned from Figure 4.22 and Figure 5.20, Ar/HZ/CH4=100/4/l sccm

gas chemistry yields 1.59 times higher growth rate than Ar/HZ/CH4=100/0/l sccm gas

chemistry but the RMS surface roughness is 2.36 times worse. As supplied by ANL, the

SiC seals were 30 um lapped and thus had rough deposition surfaces. From the viewpoint

of ultra-nanocrystalline diamond, there are lots of big potholes and deep ditches on the

surface of the seal. In order to achieve a smooth surface, one must ﬁll the “holes” and

“ditches” on the substrates. To ﬁll them up quickly and still have a smooth surface, a dif-

ferent deposition process chemistry was developed. The ultra-nanocrystalline diamond

deposition process started with a high-growth-rate gas chemistry ratio and ended with a

low-surface-roughness gas chemistry ratio. It was hoped that the process would substan-

tially reduce the deposition time and the surface roughness and also achieve a continuous

ﬁlm.

Two more seals

were coated

with UNCD. One was deposited

with

Ar:H2:CH4=100:2:1 gas chemistry ratio, ft=103 sccm for 24 hours. The other was ﬁrst

deposited with Ar:H2:CH4=100:4:1 gas chemistryratio, ft=105 sccm for 8 hours and fol-

lowed by ArzCH4=100:1 gas chemistry ratio, ft=101 sccm for 16 hours [Suma 2002].

Although the surface proﬁle of the seals could not be carried out at MSU because the lack

of the appropriate equipment, one was able to compare the optical microscopy images of

the coated seals (See Figure 7.14.). The seal deposited with the two different gas chemis-

try ratio method is much smoother (See Figure 7.14(b).). The collaboration with ANL and

FSC was ended before these seals could be tested. Thus, the author was not able to deter-

mine the mechanical properties of these coated seals. Even though the ﬁnal test results

330

I
.
-
.
.
.
.

 

_. -_ : ...—oun. “~_..r

were not available to MSU researchers, from the work reported above, one may draw the

conclusion that the ultra-nanocrystalline diamond coating for the SiC seal had lower sur-

face roughness when lower hydrogen concentrations are present in the deposition process.

As a result, the process may lead to lower friction coefﬁcient and smaller wear coefﬁcient.

Therefore, it may have potential to pass the long term durability test in the water pump.

33]

(a)

(b)

. w n’7‘619.‘r5‘7.v'- r 5 

 

FIGURE 7.14 : SURFACE PROFILE INSPECTED BY OPTICAL MICROSCOPY

(a): Pressure=160 Torr, Ar/HZICH4=100/2/l seem, and t=24 hours (26 um x 20 um).

(b): Pressure=l60 Torr, Ar/HZ/CH4=100/4/1 sccm, and t=8 hours then followed by

Ar/CH4=100/l sccm, and t=16 hours (13 um x 10 pm) by system H.

332

 —

-

A‘ use- -" .-_-"". 1W '

7 .3.4 Summary

Ultra-nanocrystalline diamond coatings for SiC seals was experimentally

explored. The SiC seals were successfully coated with UNCD. The ﬁlms were uniformly

coated and adhered well on the SiC substrates. The reason why the ﬁrst UNCD coated SiC

seal did not perform well in the wear—resistant test probably was caused by the initially

very rough SiC substrate surface (30 pm). If the SiC seals were initially mechanically pol-

ished to achieve excellent smoothness and then the deposition methodologies developed

and described in Chapters 3 and 4 of this thesis were employed for the coatings. It appears

possible to uniformly coat UNCD on SiC chemical process pump seals.

An additional improvement may also be carried out for the ultra-nanocrystalline

diamond coatings on the SiC chemical process pump seals. That is, if the SiC seal surface

must be rough, then one can use a combined gas chemistry growth step process described

in Section 7.3.3 to (1) quickly ﬁll up the big and deep tracks left on the 30 um lapped sur-

face, (2) deposit the coatings in a clean environment, and (3) synthesize smooth ﬁlms

which deduce low friction coefﬁcient, low wear rate, and low counterface coefﬁcient.

Therefore, prolong the mean life time of the pumps, reduce leakage of hazardous gases,

and lower the friction energy loss.

333

 

 

7 .4 SAW Devices Based on Ultra-nanocrystalline Diamond

7.4.1 Introduction

Filters Operating at high frequencies up to many gigahertz are required in telecom-

munications to speed up data transmission. Since diamond ﬁlms have a large SAW

(surface acoustic wave) velocity ~1x104 mlsec, their use as a SAW substrate can provide

great advantages in the fabrication of high frequency SAW devices. However, diamond is

not piezoelectric. It must be combined with a piezoelectric thin ﬁlm, such as ZnO and

AlN.

Diamond ﬁlms used as surface acoustic wave substrate materials have attracted

researchers’ attention for relaxing the lithographic criteria [Yama 1989] and [Naka 1998].

The need for a piezoelectric overlayer, often ZnO, has introduced additional complexities:

(1) velocity dispersion of the layered structure and (2) allowance of multiple modes. Nev-

ertheless, there are successful devices composed of ZnO, polycrystalline diamond, and Si

layered structures and they have been fabricated through different diamond synthesis

methodologies [Naka 1994], [Naka 1995], [Higa 1997], and [A530 2001]. However, quite

often, the micron sized polycrystalline diamond ﬁlms on Si Substrates are rough and

require post processing, like mechanical polishing or etching during the device fabrication

processing. Moreover, the intergrain phonon scattering in polycrystalline diamond can

affect the device performance if the lateral grain dimensions are on the length scales

between acoustic wavelengths and SAW device aperture and transducer separations [Bi

2002]. One of the properties of ultra-nanocrystalline diamond that has received much

attention is its smooth surfaces that are the result of by the ﬁlms’ small crystal size. The

 

334

A
V
I
ﬂ
T
.
.
.
“
.
.
.

 

 

 

 

utility of ultra-nanocrystalline diamond may solve the above fabrication problems if the

elastic properties of ultra-nanocrystalline diamond do not differ from those of their

high-quality, larger grained counterparts.

The motivation of this application investigation were: (1) to study the elastic

response of ultra-nanocrystalline diamond in the 1 GHz regime, and (2) to investigate the

possibility of ultra-nanocrystalline diamond as a SAW substrate material. The UNCD

samples produced for this research and deposition conditions are brieﬂy described in Sec-

tion 7.4.2. The theoretical calculations and measurements on the samples are described in

Section 7.4.3. Experimental results compared with theoretical calculations for high-qual-

ity polycrystalline diamond are presented in Section 7.4.4. Finally, Section 7.4.5

summarizes the research of ultra—nanocrystalline diamond SAW devices.

This application investigation was the result of a collaborative effort between the

Physics Department at Michigan State University and the author. The UNCD ﬁlm used in

the SAW-device fabrication was synthesized by the author. The SAW device design, fabri-

cation, and testing were carried out by Dr. Bi and Dr. Golding.

7.4.2 Experimental Method

Ultra—nanocrystalline diamond ﬁlm for a SAW-device substrate was grown on a

3”(100) silicon wafer by MSU-MPACVD reactor system 11 with a mixture of argon,

hydrogen, and methane gases. The deposition details have been previously described in

Sections 3.2.2.2, 3.2.3.2, 3.2.4.1, and 3.3.2.2. Prior to the deposition, the Si wafer was pre-

pared by scratch seeding (Procedure is described in Section 3.3.1.2.) The deposition

335

 

 

 

l

conditions were: the pressure was 160 Torr, the microwave power was 1.1 kW, and the gas

chemistry was Ar/HZ/CH4=100/4/1 sccm. The resulting ultra-nanocrystalline ﬁlm, shown

in Figure 7.15, had a RMS surface roughness of 50 nm. The upper limit on the average

grain size was estimated to be approximately 30 nm, shown in Figure 7.15(a) [Bi 2002].

As shown in Figure 7.15(b), the SAW device was fabricated on a 23-um-thick ultra-nanoc-

rystalline diamond ﬁlm which was deposited uniformly over three inch Si substrate. The

photolithographic IDTs (interdigital transducers) consisted of 100-nm-thick Al electrodes.

The Al electrodes were prepared on the UNCD as-grown surfaces by thermal evaporation

which were then followed by lift-off processing. A ZnO layer with thickness between 1.0

and 2.5 pm, served as the piezoelectric layer, and was sputtered on the ultra-nanocrystal-

line diamond growth surface (DC triode sputtering in 85% of argon and 15% of oxygen at

5 mTorr) after IDT preparation [Bi 2002].

336

:
.
-
.
2

 

 

(b)

 15.0kV 9.2mm )QS-Ol‘.

 

FIGURE 7.15 : SEM IMAGES OF UNCD ON A SI SUBSTRATE [B12002]

(a) Plan view image of as grown UNCD ﬁlm.

(b) Cross—section image of UNCD grown on a Si substrate.

Pressure: 160 Torr, Ar/H2/CH4=100/4/1 sccm by systemII

337

J
V

’
ﬂ
‘
d
‘
-
r
'

 

 

7.4.3 Theoretical Calculations and Measurements

The theoretical calculations and measurements were carried out by Dr. Bi and Dr.

Golding at Physics Department of Michigan State University. As described in [Bi 2002]:

“The modal velocities

of ZnO/Diamond/Si multilayers were calculated with a

matrix-based approach and the Si(100) wafer was treated as an inﬁnite half-space. The

surface wave phase velocity is a function of khdiamond and kth0. Here k =

, 2» is the

2_1t

A

acoustic wavelength. If khdiamond > 4, then Si substrate is insigniﬁcant. That satisﬁed the

experimental conditions. Therefore, khdiamond equal to 4 was set for all numerical calcula-

tions. The nanocrystalline diamond layer was treated either as an isotropic medium or a

(100) oriented crystalline diamond layer. For these two cases in the measurement range

studies, the dispersion curves for the two lowest modes were indistinguishable. The den-

sity of nanocrystalline diamond was taken as 3.5 g/cm‘3. The calculations agree well with

the published results by H. Nakahata et al. [Naka 1995].

Saw-device experimental response was studied with frequency and time domain

methods. Phase velocities were measured with a network analer and the group velocities

were measured by time-of-ﬂight techniques. The measurements were done at room tem-

perature without employing any impedance matching structures.”

7 .4.4 Experimental Results

As described in [Bi 2002]: “For a given value of kth0, the phase velocities of sev-

eral modes are found and the modes are highly dispersive at small values of kth0. At

large kth, the lowest order mode tends toward the phase velocity of the Rayleigh wave

on ZnO but as kthO —> 0, it approaches the diamond Rayleigh wave velocity. The

338

 

experimental data agree well with the control device (SAW devices built on a

300-um-thick polycrystalline diamond substrates supplied by Diamonex Corp.) and the

theoretical calculations.”

7 .4.5 Summary

As described in [Bi 2002]: “The utility of as grown ultra-nanocrystalline diamond

as a high frequency SAW device substrate material is demonstrated. The elastic properties

of ultra-nanocrystalline diamond and polycrystalline diamond are indistinguishable in this

study. SAW devices based on ultra—nanocrystalline diamond and on polycrystalline dia-

mond substrates are indistinguishable (within combined uncertainties) and both agree well

with the theoretical calculated dispersion curves which are based on elastically isotropic

or single crystal diamond. (In this study, the uncertainty of the measurements of kth0 is

about i10%.) The merits of smooth surface, absence of intergrain scattering, and elastic

isotropy have put ultra-nanocrystalline diamond in the advantages over the polycrystalline

diamond. Another achievement is that the surface acoustic wave velocities over 1x104 m/s

have been realized on the as grown ultra-nanocrystalline diamond surface.”

 

339

 

   

8

Summary

8.1

Introduction

Research activities for the thesis research reported here were initiated in 1997. At

that time, microwave plasma assisted ultra-nanocrystalline diamond CVD was a recent

development. Using Ar/Hz/CH4 or Ar/HZ/C6O chemistries, ultra-nanocrystalline diamond

ﬁlm synthesis had already been proposed and had been demonstrated over a limited exper-

imental regime, i.e. pressure, substrate size, etc. For example, Gruen et al [Zuik 1995],

[Erde 1996] reported on the growth of ultra-nanocrystalline diamond ﬁlms with gas mix-

tures of argon ( 297%), methane (1%) or C60 (~600°C), and hydrogen (l-2%).

Microwave discharges were created over limited substrate areas, < 20 cm

2, and from

80-100 Torr. The best ﬁlms were produced when the input hydrogen gas content was

reduced to 1% or less, and the deposition rate increased as the pressure increased. How-

ever, these mostly argon microwave discharges were difﬁcult to control as pressure

increased or as hydrogen content decreased and the initial experiments suffered from the

lack of reactor and process versatility. Thus in 1997 ultra-nanocrystalline diamond ﬁlm

deposition technologies and associated deposition experience were very limited.

In order to apply and utilize the many beneﬁcial properties of ultra-nanocrystalline

diamond material in 1997, there was the need to develop process technologies and meth-

odologies that enabled the routine and repeatable synthesis of smooth and uniform

ultra-nanocrystalline diamond ﬁlms over large substrate surfaces and with improved depo-

sition rates.

340

J
'

‘
1
"
—
r
J
“

.
.
.
n

 

 

Thus an important initial objective of this thesis research was to utilize previously

developed MSU microwave plasma reactor technology and to explore new process tech-

nologies and methodologies that enabled smooth and uniform ultra-nanocrystalline

diamond synthesis over a large experimental parameter space. Initial objectives were to

develop techniques to deposit ultra—nanocrystalline diamond ﬁlms uniformly over three

inch silicon substrates, over a 80-150 Torr pressure regime, and in an almost “hydrogen

free” (1% or less) gaseous environment. The research activities summarized in this thesis

have achieved these objectives. The methodologies and procedures that were developed

utilize technologies that can be readily extended to larger diameter substrates by employ-

ing a similar, scaled up 915MHz reactor design. Thus it is expected that the results of this

thesis can be easily scaled up to 6—8 inch substrates.

In particular, the research activities of this thesis have been demonstrated that in

order to routinely synthesize high quality ultra-nanocrystalline diamond ﬁlms the reactor

deposition process must be performed in a high purity, i.e., base pressure: 10’7~10'6 Torr,

deposition environment. This was observed by comparing the experimental performance

between System I and System H (See Chapter 4 and Chapter 5). Under the high vacuum

conditions of System 11, the ﬁlm deposition is more repeatable, smoother, and more uni-

form and can be performed over a much larger experimental parameter space (60—240

Torr), and deposition can take place without any input hydrogen gas. Using (0-1%) H2,

large, uniform, intense, greenish-white, hydrogen deﬁcient, argon discharges were sus-

tained in contact with three inch silicon substrates over a 60-200 Torr pressure regime.

Using (2-4%)% H2 high quality ﬁlm deposition was extended to 240 Torr. At a given pres-

341

J
‘
V
T
H

:
“
‘
r
I

 

 

 

 

sure, uniform ﬁlm deposition was achieved by optimizing substrate holder geometry,

substrate position, microwave power, gas chemistry and total gas ﬂow rate. Film crystal

sizes ranged from 3-30 nm and ﬁlm roughness from 11-50 nm. Film uniformities of

almost 100% were achieved over three inch silicon wafers. At a pressure of 220 Torr with

CH4 ~1% and H2 ~4%, linear growth rates as high as ~ 0.9 pm (total growth rate 14.25

mg/hr) were achieved. It is not possible to directly compare these growth rates with those

of other research groups [Zuik 1995], [McCa 1998], [Zhou 1998], because of major differ-

ences in

reactor geometry, substrate

heating, deposition area, and growth rate

measurement techniques. However, it is clear that the maximum total growth rate reported

is the highest reported for ultra-nanocrystalline diamond ﬁlm synthesis.

8.2 The Performance Comparison Between Systems I and II

The initial ﬁlm synthesis experiments were performed using system I. This system

did not have high vacuum capability, i.e. it was only pumped down to ~2x10’2 Torr before

each experimental run and the leak rate was ~12.6 mTorr/hr. However, it was still useful

for the initial exploratory experiments presented in this thesis. The initial experimental

process methodology, such as discharge start-up, discharge positioning over the substrate,

discharge stabilization, etc. were developed with the system I reactor. Once the construc-

tion of system 11 was completed, system II was used. The initial experiments with system

11 indicated that the ﬁlm quality improved considerably. Thus, the remaining experimental

results were performed with reactor system II.

342

  

.
_
_
_
r
.

 

Since this thesis utilized both reactor systems, it is useful to compare their perfor-

mance here. Table 8.1 summarizes the performance of these reactors. Besides the high

vacuum capability (base pressure 10’7 Torr, leak rate 4 mTorr/hr), a minor difference

between systems I and II is that system II was less energy efﬁcient than system I (see

Table 4.3). This may have been caused by the differences in reactor cooling. To improve

the energy efﬁciency of system H, the reactor operator can increase the temperature set

point of the Neslab chiller from 15 0C to a higher degree, i.e. lowering the cooling

efﬁciency.

TABLE 8.1: EXPERIMENTAL PERFORMANCE OF SYSTEMS I & II

 

System

Reactor

Film Characteristics

Performance

 

I

GT: 0.163-12.963mg/hr

Film texture: (1 l l)or(220) or No preferred

Gave: 0.010—0.809 um/hr

I(220)/I(111)=0.06-4-04

Sy: 4411-69 kW*hr/g

Film roughness: 14-149 nm

C: 0.504-33.382%

Morphology: mostly (111)tr1angular

 

11

GT: 0175-1425 mg/hr

Film texture: mostly (220)

Crystal srze: 30—2700 nm

Gave: 00] 1-0.889 um/hr

I(220)/I(1 l 1)=O.13-9.31

 

 

Syi 4229_99 kW*hr/g

.

C' 054234409470

Frlm roughness: 1 1-50 nm

Morphology: featureless

Crystal size: 3-30 nm

 

 

 

As shown in Table 8.1, the growth rate, speciﬁc yield, and carbon conversion efﬁ-

ciency for the two systems are similar. However, system H allows a stabilized and ﬂame

free discharge operation over a larger range of pressures, total gas ﬂow rates, and H2/CH4

ratios, i.e. system II allowed deposition to occur over a larger input experimental parame-

343

 

ter space. For system I, the plasma could not be stabilized and/or ﬂame free if the

hydrogen to methane ratio was under three. On the other hand, experiments that utilized

system 11 were performed with hydrogen to methane ratios as low as zero.

A very important difference between the two reactors was the ﬁlm quality. When

comparing the results under the same input experimental conditions, it was observed that

system 11 has lower ﬁlm roughness, smaller crystal size, much better ﬁlm uniformity.

Experiments were repeatable. Thus the best over all deposition performance was achieved

with the high-purity system 11.

8.3 Summary of Experimental Performance

Since system II had the best overall operational performance, this section only

summarizes system 11 experimental data. With system 11, the input parameters were per-

formed in the following regime: the deposition pressure was between 60-240 Torr, the

microwave

power

was

between

732-1518

W,

the

gas

chemistry

was:

Ar/HZ/CH4=(99-9O)/(O-9)/1 (%), total gas ﬂow rate was between 101-630 sccm. The ﬁlms

were deposited on 3” diameter Si substrates. 100% uniformity was achieved at pressures

of 60-80 Torr.

344

 

 

 

Growth Rate (See Figures 4.22 and 4.23)

a) Growth rate increased as pressure and microwave power increased. The highest

growth rate of 14.25 mg/hr (~0.9 um/hr) was achieved at 220 Torr, with

Ar:H2:CH4=100:4:1, f =105 sccm, and 1.4 kW of input power.

b) Growth rate was independent of total gas ﬂow rate.

0) Growth rate was constant as the deposition time was varied from 8 to 25 hours.

(1) Growth rate increased dramatically as H2 concentration increased from 0—9%.

Carbon Conversion Efﬁciency (See Figures 4.32, and 4.33)

a) Carbon conversion efﬁciency increased as total gas ﬂow rate decreased.

b) Carbon conversion efﬁciency as high as 44% was achieved under the conditions:

220 Torr, 1.4kW, Ar:H2:CH4=100:4:1, ft=105 sccm.

Film Roughness (See Figure 5.20)

a) Film roughness increases as pressure increases; for example, surface roughness of

12 nm (with average ﬁlm thickness 0.611 pm) at 100 Torr versus surface roughness

of 26 nm (with average ﬁlm thickness of 5.284 mm) at 200 Torr. Film roughness of

11 nm was achieved under the conditions: 120 Torr, 898W, ArzCH4=100:1, f :10]

sccm, and average ﬁlm thickness=l.029 urn.

b) Film roughness increases dramatically as H2 concentration is increased (11 nm at 0

vol% versus 44 nm at 9%).

c) Film roughness increases as ﬂow rate is increased.

Film characterization by XRD, Raman (Figure 5.22), and TEM (Figure 5.36), show

deposited ﬁlms are diamond for all the ﬁlms.

Film Morphology is featureless (See Figures 2.6, 5.28, and 7.15(a)).

345

J

l
?
m
T
.

“
‘
1
-
I
H
P

 

 

 

-

The crystal sizes range from 3-30 nm with an upper bound of 30 nm for the ﬁlms pro-

duced in lower leak rate system, systemII (See Figure 5.34 for representative ﬁlm crys-

tal size distribution).

-

Cross-section TEM images show an interlayer between UNCD ﬁlms and Si substrates

in most cases.

As shown in Figures 5.17 through 5.20, ﬁlms with roughness of less than 50 nm

and with crystal size of less than 30 nm can be achieved under the following experimental

conditions: deposition pressure between 60-240 Torr, microwave power between

732-1518 W, gas chemistry conﬁned within Ar/HZ/CH4=(99-90)/(0-9)/1 (%), total gas

ﬂow rate between 101-111 sccm, and deposition time up to 25 hours. If ﬁlm roughness <

30 nm is required, then the deposition conditions are limited to: pressure between 60-200

Torr, microwave power between 732-1310 W, gas chemistry conﬁned within:

Ar/HZICH4=(99-95)/(O—4)/ 1 (%), total gas ﬂow rate between 101-105 sccm, and deposi-

tion time up to 8 hours.

It was observed that the deposition operational ﬁeld map that developed the rela-

tionship between substrate temperature, TS, pressure, p, and input power, Pabs, for

ultra—nanocrystalline diamond synthesis behaved differently from that of more common

H2/CH4 depsoition process. If the gas chemistry was held constant, then an unique MCPR

operational ﬁeld map of UNCD growth was established experimentally (See Figure 3.12).

However, when input gas mixtures are varied, the curves change. This is particularly true

for changes in input hydrogen concentrations. Thus, this feature of the ﬁeld map versus

 

 

346

 

 
 

input gas concentrations is an important difference from the conventional HZ/CH4 deposi-

tion process.

Table 8.2 shows a comparison of crystal sizes of ultra-nanocrystalline diamond.

The experiments performed by Csencsits et a1. [Csen 1996], Erdemir et a1. [Erde 1996],

Gruen et a1. [Grue 1996], Luo et a1. [Luo], and Qin et al. [Qin 1998] were using a micro-

wave plasma chemical vapor deposition reactor (ASTeX PDS-17) and heated substrate

holder.

TABLE 8.2: CRYSTAL SIZE COMPARISON OF ULTRA-NANOCRYSTALLINE

DIAMOND FILMS

 

Reference

Gas
.

Ch

Flow
Rate

P

emlstry

(seem)

(Tom)

M P
'
(W)

'

Ts
o

( C)

tn

'm
(um)

Crystal Size

(um)

S R

'

'
(w)

 

 

[Csen 1996] NA

~100

100

1500

850

NA

ave.=7.7 (equiaxed)

30-50

110(dendritic)

 

[Csen 1996]

Ar:H2:C60

~100

100

1500

850

NA

ave.=14.9 (equiaxed)

NA

=98:2

 

[Erde 1996]

Ar:H2:CH4

101

100

1500

800- NA

10-30

20-40

=98:2:1

850

 

[Erde 1996]

Ar:H2:C6O

100

100

1500

800-

<
_ 6

ave.=15

20-50

=97:2:1

850

range: 3-110

 

[Grue 1998]

Ar:H2:CH4

100

100

1200

800

5

30—50

54.19

=90:9:1

 

[Grue 1998]

Ar:H2:CH4

100

100

1200

800

5

10—30

18.84

=97:2:1

 

=99:0:1

 

[Grue 1998]

Ar:H2:CH4

100

 
 
  

20 nm within an array

of larger crystals

Ar,H2, C60

850

NA

(~300 nm)

100

1200

800

5

 

NA

NA

1500

3-20

NA

Luo

NA

 

 

 

 

 

 

347

J
*

!
1
7
7
:
3

:
.
2
-
1
-
a
r
I

 

 

 

TABLE 8.2 (cont’d)

 

Reference

Ch

Gas
.

Flow
Rate

emlsmj

(seem)

P

T
( orr)

M P
W '
)

(

Ts
o

( C)

tF.

' m
(um)

Crystal Size

(um)

s R

'

'
(um)

 

 

[Qin 1998]

Ar,H2, C60

NA

100

800

800

NA

ave.=3-5

NA

occasionally see size

of 15 nm

 

 
  

20 nm has been seen

“1100/Ml

 

 

 

 

:19

 

 

 

 

M.P.: Microwave power

8.4

Plasma Diagnostics Using System 11

Goyette et al. [Goye 1998] observed a strong linear correlation between the exper-

imentally measured C2 emission intensity and the experimentally measured C2

concentrations. They observed the C2 densities were linearly related to C2 emission inten-

sity. Based on the work of Goyette et al., in this thesis, it is assumed that C2 emission

intensity measurements are linearly related to C2 concentrations in the discharge.

Experimentally measured C2 emission intensities increased as pressure or micro-

wave power increased. C2 emission intensity

generally decreased as hydrogen

concentration increases. Therefore, if the observed behavior by Goyette et al. also holds

for the MSU reactors, C2 concentration increased as pressure or microwave power

increased. C2 concentration generally decreased as hydrogen concentration increased. It

was observed that total growth rate is usually higher when the C2 emission intensity is

higher except when the hydrogen concentration is increased. Thus these measurements

indirectly support the hypothesis that C2 is an important growth species. This is no longer

valid at higher hydrogen concentrations. The plasma deposition chemistry may change

348

.
.
A
m
:
2
'
5

“
.
r
t

 

and maybe other radicals such as CH3 are also present and inﬂuence the growth mecha-

nism. It was also observed that the ﬁlm roughness increased as hydrogen percentage

increased. This suggests the discharge and deposition chemistry may change their behav-

ior towards H2/CH4 discharges.

The gas temperature ranges from 1975K to 2960K between 60-240 Torr pressure

regime and increases as pressure or microwave power increases. The gas temperature also

increases as hydrogen concentration increases. The gas temperature measured for the reac-

tor of system II is higher than the 1300 °C (1573 K) reported by Goyette et al. The

different temperature may be caused by: (1) the reactor geometry difference of these two

systems (MSU-MPACVD system II versus ASTeX PDS-l7 MPECVD system), and (2)

the higher temperature observed in MSU-MPACVD system H should lead to higher chem-

ically active species concentrations and the MSU system 11 reactor may be more efﬁcient.

That is increases in gas temperature are expected to increase the dissociation of neutral gas

species and produce higher chemically active species concentrations in the discharge. The

measured temperature at the higher pressure regime probably explains why the ﬁlm

growth rates are higher when the gas temperature is higher.

8.5

Controlled Nitrogen Impurity Study By System II

The addition of nitrogen impurities decrease the growth rate and the carbon con-

version efﬁciency approximately 38%, and increases the speciﬁc yield approximately 68%

as nitrogen impurity increases from a few ppm to 2500 ppm. Adding nitrogen impurity

into the deposition process has no signiﬁcant effect on ﬁlm roughness. The ﬁlms all have

349

 

featureless morphology and the ﬁlms begin to look opaque. Therefore, the optical proper-

ties of the ﬁlms have been changed. The molecular nitrogen becomes chemically active in

the plasma discharge and it interacts with methane in the gas phase inside the plasma

discharges.

8.6

Ultra-nanocrystalline Diamond Film Applications

The study results of ultra-nanocrystalline diamond coatings as the IR optical trans-

mission window have demonstrated that the ultra-nanocrystalline diamond has similar

optical properties as high-quality diamond in this region of infrared wavelength. There is

no visible IR-active single phonon absorption (1000 cm“1 to 1500 cm'l) in measurements,

therefore the ﬁlms are absent of signiﬁcant defects or impurities (The intrinsic optical

properties of diamond are associated with the arrangement of the carbon atoms on a dia-

mond lattice.)

A maximum of 89% is achieved by one layer both-side ultra-nanocrystalline dia-

mond coated silicon substrates. In this research, two layered coatings do not show higher

transmission than single layer because of the thickness of diamond ﬁlms and Si02 layers.

The IR optical transmission is sensitive to the thickness of the coatings and the differences

in thickness of both sides. With the aid of simulation by the theory, the design and fabrica-

tion of ultra-nanocrystalline diamond IR optical window at desired increment can be

achieved.

The utility of as grown ultra-nanocrystalline diamond as a high frequency SAW

device substrate material is demonstrated. The elastic properties of ultra-nanocrystalline

350

diamond and polycrystalline diamond are indistinguishable. The merits of a smooth sur-

face,

the

absence

of

intergrain

scattering,

and

the

elastic

isotropy

give

ultra-nanocrystalline diamond in the advantages over the polycrystalline diamond.

8.7

Recommendations for Future Research

(1) The experimental work in this thesis indicates that a maximum deposition rate

of 14.25 mg/hr (~O.9 um/hr) was achieved at 220 torr. These experiments were limited to

three inch diameter substrates. Thus, the process costs are still high for commercial appli-

cations of UNCD. It is useful to extend the process to larger substrate areas and higher

deposition rates. Thus, future work should investigate the potential scale up of this meth-

odology and process to six-eight inch diameter substrates and modify this process to

increase the deposition rates. One potential method of increasing deposition rates is to try

to deposit UNCD ﬁlms at higher pressures.

(2) From the results of this thesis research, growth rate is essentially independent

of total gas ﬂow rate. However, the low-total-gas-ﬂow-rate has higher carbon conversion

efﬁciency and lower ﬁlm roughness. More research in low or even lower total gas ﬂow rate

is recommended.

(3) In order to have a better understanding of the ultra-nanocrystalline diamond

synthesis. More study of the relationship between internal variables, such as electron tem-

perature, and input and output variables must be carried out.

35]

 

(4) Using a heater substrate holder set-up to (i) allow the independent control of

the substrate temperature, (ii) improve the growth rate, and (iii) improve the ﬁlm quality at

deposition pressure of 60-80 Torr. (At deposition pressure of 80 Torr, the testing results of

heater showed 77% increases in growth rate and ﬁlm roughness decreased from 26 nm to

12 nm.)

(5) A transitional interlayer exists between silicon substrate and ultra-nanocrystal-

line diamond ﬁlm. Further probing using EELS (Electron Energy Loss Spectroscopy) is

recommended to identify the nature of this interlayer.

(6) As the grain sizes decrease, there is an increasing fraction of the carbon atoms

with complex bonding conﬁguration resides in or near grain boundaries. It makes ﬁlm

structure quality determined by Raman ambiguous and undetermined most of the time.

Therefore, determine the ﬁlm quality of ultra-nanocrystalline diamond ﬁlms with x-ray

absorption spectroscopy is recommended.

(7) In addition to [Bi 2002], it is of interest to compare the propagation loss of

SAW devices of ultra-nanocrystalline diamond with that of polycrystalline diamond.

(8) Develop a combined (i) gas chemistry growth steps or (ii) deposition pressure

to synthesize smooth ultra-nanocrystalline diamond ﬁlms efﬁciently.

 

(9) The SAD patterns indicate that ultra-nanocrystalline diamond has random ori-

entation. However, the XRD results show mostly preferred (220) orientation. Neither SAD

352

 

pattern nor the X-ray two-theta scan is able to quantitatively determine the texture of the

ﬁlms. Therefore, to verify the texture of ultra-nanocrystalline diamond ﬁlms quantita-

tively, a more detail study by X-ray pole ﬁgures and inverse pole ﬁgures is recommended.

353

 

 

 

 

9

Appendix

9.1

System]

1‘: total gas ﬂow rate

p: pressure

t: time

Ts: substrate temperature

Pabs: microwave power

Gt: total growth rate

Gave: linear growth rate

S: speciﬁc yield

C: carbon conversion efﬁciency

RMS: root mean square surface roughness

I(220)
I_(_11—1): XRD ratio of integrated intensmes of diamond <220> and <111> peaks

..

.

.

.

.

Experimental data for the experiments performed in systems I.

TABLE 9.1: EXPERIMENT DATA FOR SYSTEM 1

 

Ar

%

H2

CH4

ft

p

t

Ts

Pabs

Gt

Gave

5

%

%

sccm

Torr

hr

0c W mg/hr

urn/hr

kWh/g

C

%

RMS

@9—

nm

[(111

 

 

98.68

0.99

0.33

304

120

8

794

717

0.238

0.015

3013

0.736 NA

0.06

 

98.52

0.99

0.49

203

120

8

815

719

0.163

0.010

4411

0.504

NA

0

 

98.04

1.47

0.49

204

120

8

810

756

0.463

0.029

1633

1.433

NA

0.12

 

98.04

1.47

0.49

408

120

8

870

867

0.325

0.020

2668

0.503

NA

0.25

 

97.56

1.95

0.49

102.5

120

8

788

832

0.588

0.037

1415

3.639

80

0.10

 

97.56

1.95

0.49

205

120

8

813

765

0.838

0.052

913

2.593

NA

0.16

 

1.95

0.49

205

120

8

97.56

97.56

 

  
 

1.142 NA

 

0.750

836

0.738

120

8

790

702

120

8

806

741

0.046

1.95

0.49

410

1.95

0.49

97.56

0.047

936

1.547

NA

0.725

0.045

1133

3075

1022

2.243

NA

0.12

 

 

0.14

NA

 

 

 

 

820

 

 

 

 

 

 

 

354

TABLE 9.1 (cont’d)

 

Ar

%

H2

%

CH4

f:

p

I

TS

Pabs

Gt

Gave

8

%

sccm

Torr

hr

0C W mg/hr

urn/hr

kWh/g

C

%

RMS

QC!)

nm

1( l l 1 ]

 

 

97.56

1.95

0.49

615

120

8

830

821

0.750

0.047

1095

0.774 NA

NA

 

97.56

1.95

0.49

615

120

8

807

839

0.763

0.048

1 100

0.787

30

0.18

 

97.56

1.95

0.49

615

120

43

815

839

1.004

0.063

836

1.036

149

0.12

 

97.56

1.95

0.49

615

120

24

850

821

0.950

0.059

864

0.980

71

0.15

 

97.56

1.95

0.49

615

120

4

810

821

0.700

0.044

1173

0.722

NA

NA

 

97.56

1.95

0.49

615

120

4

840

858

0.675

0.042

1271

0.696

NA

NA

 

97.56

1.95

0.49

615

120

4

850

728

0.550

0.034

1324

0.567

NA

NA

 

97.56

1.95

0.49

615

120

4

830

802

0.675

0.042

1 188

0.696

15

0.17

 

97.56

1.95

0.49

615

120

2

850

847

0.700

0.044

1210

0.722

14

0

 

97.56

1.95

0.49

615

120

1

815

847

0.500

0.031

1694

0.516 NA

NA

 

97.40

1.95

0.65

308

120

8

830

784

1.125

0.070

697

1.741

NA

0.23

 

97.32

1.95

0.73

6165

120

8

830

865

1.038

0.065

833

0.714 NA

NA

 

97.09

1.94

0.97

206

120

8

900

821

2.025

0.126

405

3.134 NA

0.38

 

96.39

2.89

0.72

622.5

120

8

850

895

2.163

0.135

414

1.487

NA

0.20

 

96.39

2.89

0.72

6225

120

8

850

899

2.313

0.144

389

1.590 NA

0.17

 

96.15

2.88

0.96

104

120

8

830

839

3.450

0.216

243

10.675

93

0.40

 

96.15

2.88

0.96

416

120

8

900

832

2.188

0.136

380

1.693

NA

0.26

 

96.15

2.88

0.96

624

120

8

823

821

2.838

0.177

289

1.464

90

0.24

 

95.24

3.81

0.95

105

120

8

844

858

4.625

0.289

186

14.311

NA

0.20

 

95.24

3.81

0.95

105

120

8

855

854

4.263

0.266

200

13.191 NA

0.10

 

95.24

3.81

0.95

105

120

8

850

858

4.625

0.289

186

‘

14.311

96

0.47

 

120

8

1.818

120

8

120

8

0.277

197

0.95

105

95.24

3.81

95.24

3.81

95.24

3.81

93.46

5.61

0.95

315

0.224

222

0.93

642

0.95

630

880

795

3.588

875

795

3.525

850

873

4.438

 

 

 

 

13.733 NA

  
 
 
  
 
 

 

 

 

20-056 NA

8.906 NA

17.522 NA

 

 

 

 

 

876

7.238

 

897

906

1 1513

0.220

226

890

925

1 1325

900

876

10950

 

850

888

12963

169‘“

NA

0.809

69

0.718

79

3.733

NA

92.59

5.56

92.59

5.56

92.59

5.56

1 19

0.20

0.707

82

1.85

216

0.683

80

1.85

108

7.27

1 .82

1 10

3.701

NA

0.452

121

1.85

108

120

8

845

3.93

2.59

0.08

0.16

 

1.04

2.74

4.04

120

8

120

8

120

8

120

8

90.91

355

 

TABLE 9.1 (cont’d)

 

Ar

H2

CH4

ft

p

1

T5

Pabs

Gt

Gave

5

%

%

%

sccm

Torr

hr

0c W mg/hr

tun/hr

kWh/g

C

%

RMS m

nm

1

1 1 1 1

 

 

99.17

0.66

0.17

605

140

8

755

705

0.188

0.012

3750

0.582

NA

NA

 

97.56

1.95

0.49

102.5

140

8

840

828

1.100

0.069

753

6.808

NA

0.22

 

97.56

1.95

0.49

205

140

8

840

820

l . 100

0.069

745

3.404 NA

0.17

 

97.56

1.95

0.49

307.5

140

8

845

816

1.025

0.064

796

2.1 14 NA

0. 16

 

97.56

1.95

0.49

410

140

8

850

809

1.188

0.074

681

1.838

NA

0. 17

 

97.56

1.95

0.49

615

140

8

830

784

1.238

0.077

633

1.277

43

0.19

 

97.56

1.95

0.49

615

140

24

825

705

1.625

0.101

434

1.676

94

NA

 

96.15

2.88

0.96

104

140

8

850

947

5.088

0.317

186

15.744 NA

1.16

 

96.15

2.88

0.96

208

140

8

904

956

4.888

0.305

196

7.563

NA

0.60

 

96.15

2.88

0.96

312

140

8

900

954

4.988

0.311

191

5.145

NA

0.46

 

96.15

2.88

0.96

416

140

8

950

932

4.888

0.305

191

3.781

NA

0.49

 

96.15

2.88

0.96

624

140

8

860

932

5.413

0.338

172

2.792 NA

0.34

 

95.24

3.81

0.95

105

140

8

867

956

7.163

0.447

133

22.165 NA

1.83

 

95.24

3.81

0.95

210

140

8

890

987

7.350

0.459

134

11.372 NA

1.99

 

95.24

3.81

0.95

315

140

8

888

965

6.950

0.434

139

7.169

NA

0.71

 

95.24

3.81

0.95

420

140

8

880

972

6.863

0.428

142

5.309 NA

0.63

0.443

136

97.56

1.95

98.36

1.31

0.33

305

0.95

630

95.24

3.81

885

965

7.100

 

 

 

 

  
95.24 
  
 

33.382 NA

3.249 NA

13.422 NA

 

160

20

860

 

 

 

 

898

2.21

1017

8.675

900

959

10.788

0.96

208

0.114

495

0.49

615

0.49

615

800

838

1.05

850

903

1.825

2.279

95

0.066

798

1.882

58

3.662

NA

0.138

406

96.15

2.88

3.81

0.95

0.673

89

97.56

1.95

160

8

910

0.541

117

105

160

8

3.62

3.02

 

 

0.40

NA

0.18

NA

 

 

 

140

8

160

8

160

8

 

 

356

 

9.2

System 11

Experimental data for the experiments reported in Section 4.3.

TABLE 9.2: EXPERIMENT DATA FOR SYSTEM II

 

Ar/Hz/CH4

p

sccm

Torr

1

hr

Ts

cc

Fab,

w

Gt

Gave

5

mg/hr

pm/hr

kWh/g

c

%

RMS

IELO)

nm

1( 1 1 1)

 

 

100/10/1

120

8

619

902

5.400

0.337

167

16.709

44

1.81

 

100/8/1

120

8

659

977

5.263

0.328

186

16.285

41

2.62

 

100/6/1

120

8

637

977

4.538

0.283

215

14.042

37

2.66

 

600/24/6

120

8

637

942

3.925

0.245

240

2.024

82

0.13

 

400/16/4

120

8

630

898

4.038

0.252

222

3.124

81

0.14

 

300/12/3

120

8

621

898

3.875

0.242

232

3.997

75

0.33

 

200/8/2

120

8

629

907

3.538

0.221

256

5.474

54

1.23

 

600/12/6

120

8

605

926

3.025

0. 189

306

1.560

34

0.64

 

400/8/4

120

8

607

871

3.038

0. 189

287

2.350

34

1.05

 

300/6/3

120

8

596

871

3.025

0.189

288

3.120

33

1.00

 

200/4/2

120

8

600

878

2.738

0.171

321

4.236

17

4.17

 

100/4/1

60

8

UFL

768

0.375

0.023

2048

1.160

13

0

 

100/4/1

80

8

UFL

800

0.875

0.055

914

2.708

26

0.24

 

100/4/1

80

8

726*

851

1.55

0.097

549

4.796

12

0.50

 

100/4/1

100

8

UFL

847

1.888

0.118

449

5.842

19

0.71

 

1.19

 

357

3.800

3 .663

100/4/1

120

100/4/1

120

100/4/1

120

626

945

650

973

 

0.237

249

1 1.758

24

 

 

 
 
 
  

1214

7.913

19.844

25

1 1.836

23

12.009

23

1053

6.413

24.485

32

1 1.335

25

0.242

249

0.239

252

0.228

266

16

642

16

629

0.400

164

0.494

153

 

965

3.825

965

3.881

100/4/1

120

100/4/1

140

 

100/4/1

160

709

8

8

8

8

NA

3.51

 

3.08

5.01

1.96

 

 

 

680

 

 

 

TABLE 9.2 (cont’d)

 

Ar/Hz/CH4

p

1

sccm

Torr

hr

Ts

°C

Fab,

Gt

Gave

3

W

mg/hr

urn/hr

kWh/g

c

%

RMS m

nm

1( 1 11)

 

 

100/4/1

160

8

697

1223

8.238

0.514

148

25.491 NA

NA

 

100/4/1

160

1 1

678

1246

7.873

0.491

158

24.362

36

NA

 

100/4/1

160

24

672

1139

8.658

0.540

132

26.791

50

NA

 

100/4/1

160

25

675

1181

7.936

0.495

149

24.557

47

NA

 

100/4/1

180

8

709

1361

9.850

0.614

138

30.479

33

2.69

 

100/4/1

200

8

770

1429

12.700

0.792

113

39.298

32

4.31

 

100/4/1

220

8

738

1405

14.250

0.889

99

44.094

47

3.85

 

100/4/1

240

8

730

1518

14.025

0.875

108

43.398

34

3.18

 

100/2/1

60

8

752

0.297

0.019

2532

0.919

24

0

 

100/2/1

80

8

UFL

772

0.395

0.025

1954

1.222

22

0.15

 

100/2/1

100

8

851

1.225

0.076

695

3.791

12

0.70

 

100/2/1

120

8

616

906

2.750

0.172

329

8.509

16

1.90

 

100/2/1

140

8

684

1017

5.813

0.363

175

17.987

21

6.42

 

100/2/1

160

8

696

1 147

8.063

0.503

142

24.950

25

6.26

 

100/2/1

160

8

706

1160

7.575

0.473

153

23.440 NA

NA

 

100/2/1

180

8

681

1211

8.438

0.526

144

26.110

25

5.01

 

100/2/1

200

8

707

1310

10.588

0.660

124

32.763

26

3.71

 

100/2/1

220

8

692

1322

1 1.525

0.719

115

35.662

35

4.62

 

100/1/1

60

8

732

0.075

0.005

9760

0.232

17

0

 

0

 

 

 

100/1/1

120

100/1/1

100

100/1/1

80

815

1.025

0.894

25

594

890

8

8

8

8

UFL

768

0.018

2657

0.064

795

  
 
 
 

 

 

1 199

7.175

22.202

22

1119

5.388

1005

3.725

0.232

270

0.336

208

0.140

399

 

 

0.148

375

16.672

19

11.526

17

 

8

8

8

0.448

167

601

894

3.172

13

7.349

13

 

6.925

13

NA

100/1/1

180

100/1/1

140

100/1/1

160

100/1/1

120

 

621

655

665

0.25

1.55

 

4.27

3.48

5.21

 

 

 

0.289

2.375

2.238

358

 

 

TABLE 9.2 (cont’d)

 

Ar/Hz/CH4

p

t

sccm

Torr

hr

Ts

0c

Pabs

Gt

Gave

S

W

mg/hr

urn/hr

kWh/g

C

%

RMS

I_(_2_2_(_)_)

nm

1( 1 1 1)

 

 

100/1/1

200

8

673

1251

8.825

0.551

142

27.307

22

5.58

 

100/1/1

200

8

677

1314

8.838

0.551

149

27.348

27

5.99

 

100/0/ 1

60

8

UFL

740

0. 175

0.01 l

4229

0.542

27

0

 

100/0/1

80

8

UFL

772

0.288

0.018

2681

0.891

23

0

 

lOO/O/l

100

8

UFL

803

0.813

0.051

988

2.516

16

0.19

 

100/0/1

120

8

587

898

2.063

0.129

435

6.384

11

1.40

 

100/0/1

140

8

583

946

3.700

0.231

256

11.449

14

8.31

 

100/0/1

160

8

608

1037

4.988

0.311

208

15.435

13

9.31

100/0/1

180

8

 

 

  
 
  

l 144

4.750

1220

6.038

1227

5.300

0.296

241

16.400

18

0.377

202

18.684

18

14.698

17

0.331

232

200

8

100/0/1

100/0/1

200

8

61 1

 

3.92

3.94

4.46

 

 

 

 

 

 

 

625

604

 

*: Temperature was measured by thermal couple and the experiment was performed with

heated stage.

359

 

 

TABLE 9.3: CONTROLLED N2 IMPURITY STUDY BY SYSTEM II

 

Armz’cm

f1

N2

sccm

seem

ppm

t

hr

Ts

°C

Pabs

Gt

Gave

S

w

mg/hr

11mm;

kWh/g

C

%

RMS

1_(2_20_)

nm

1(1 1 1 )

 

 

100/4/1

105

5

8

626

945

3.800

0.237

249

1 1.758

24

1.19

 

100/4/1

105

100

8

646

957

3.813

0.238

251

11.799

21

1.27

 

100/4/1

105

200

8

640

969

2.925

0.182

331

9.051

20

0.63

 

100/4/1

105

300

8

645

969

3.388

0.21 1

286

10.484

20

1.21

 

100/4/1

105

600

8

645

969

3.225

0.201

300

9.979

21

0.86

 

100/4/ 1

105

800

8

642

969

3.000

0.187

323

9.283

21

0.65

 

100/4/1

105

1200

8

645

957

2.775

0.173

345

8.587

24

0.36

 

105

105

100/4/1

1500

8

 

  
  

100/4/1

2000

8

2500

8

646

973

2.350

0.147

636

981

0.160

383

625

985

2.625

0.164

371

419

7.272

8.123

23

7.931

21

100/4/1

105

18

0.81

 

0.51

0.73

 

 

 

 

 

2.563

 

 

 

 

 

360

 

 

9.3 OES

Experimental data for the OES measurements reported in Chapter 6.

TABLE 9.4: OES DATA

 

Ar/Hz/CH4

Pressure

Pabs

C2

(sccm)

(TOIT)

(W)

(a.u.)

Tg

(K)

 

100/0/1

60

681

1.36

1975

 

100/0/1

80

744

2.54

2000

 

100/0/1

100

803

2.28

2150

 

100/0/1

120

831

3.69

2205

 

100/0// 1

140

946

9.61

2345

 

100/0/1

160

1041

24.17

2450

 

100/0/ 1

180

1 152

20.05

2540

 

100/0/1

200

1259

13.96

2565

 

100/1/1

60

721

1.92

2005

 

100/1/1

80

768

2.47

2140

 

100/1/1

100

831

4.26

2165

 

100/1/1

120

871

6.4

2260

 

100/ 1// l

140

986

21.63

2420

 

100/1/1

160

1080

46.58

2510

 

100/1/1

180

1164

78.14

2530

 

100/1/1

200

1259

7 88.44

2595

 

100/2/1

60

721

1.29

2100

 

100/2/1

80

772

1.99

2110

 

 

 

 

 

 

100/2/1

100/2/1

100/2/1

835

878

100

120

140

4.25

2205

8.86

2335

100/2/1

100/2//1

  
  
 

157.59

2730

125.66

2690

47.26

2525

19.87

2430

100/2/1

1 183

83.41

2595

100/2/1

160

180

200

1259

220

1334

994

l 104

 

 

 

361

 

TABLE 9.4 (cont’d)

 

Ar/112/CH4

Pressure

Pabs

C2

(sccm)

(TOIT)

(W)

(a.u.)

Tg

(K)

 

100/4/1

80

809

0.83

2140

 

100/4/ 1

100

869

2.86

2205

 

100/4/1

120

908

7.88

2355

 

100/4//1

140

1033

18.09

2445

 

100/4/1

160

1151

40.59

2595

 

100/4/ 1

180

1242

79.2

2685

 

100/4/1

200

1338

81.54

2785

 

100/4/1

220

1413

82.63

2845

 

100/4/1

240

1520

64.57

2960

 

100/0/1

120

851

2.75

2215

 

100/1/1

120

863

4.23

2220

 

100/2/ 1

120

871

5.74

2270

 

100/4/ 1

120

894

5 .6

2340

 

100/6/1

120

908

5.22

2585

 

100/8/ 1

120

922

4.46

2680

 

 

 

 

 

100/10/1

100/4/1

200/8/2

934

908

908

120

120

3.31

2745

  
 

400/ 16/4

600/24/6

4.64

2295

4.28

2330

3.87

2315

4.33

2375

120

120

120

120

908

4.5

908

908

300/12/3

 

 

2340

 

 

 

362

 

TABLE 9.5: CN STUDY

 

Ar/Hz/CH4

N2

Pressure

Pabs

CN

(Sccm)

(ppm)

(Torr)

(W)

 

100/4/1

0

120

908

0.18

 

100/4/ 1

25

120

908

0.28

 

100/4/ 1

50

120

908

1.05

 

100/4/1

100

120

908

3.16

 

100/4/1

200

120

908

6.52

 

100/4/ 1

300

120

908

10.84

 

100/4/1

600

120

908

14.82

100/4/1

 

 

 
  

1500

120

1200

120

908

120

908

908

26.55

16.74

21.08

100/4/1

100/4/ 1

 

 

800

 

 

 

363

9.4 Gas Temperature Measurements

Matlab program used for ﬁtting the R25 to R45 emission lines to the experimental

data in determining the rotational temperature.

%Calculation of the intensity of the optical emission spectrum

%Rotational quantum number J

%Bvu and ng molecular constant (cm'l)

%

a=zeros(90);

lam=zeros(90);

for 1:] :90

Jp=J;

Jpp=Jp-l;

Bvu=1.7456;

ng=1.62405;

nu=Bvu*Jp*(Jp+l)-ng*Jpp*(Jpp+1); %(cm’l)

h=6.63*10"(-34);

%(Joule*sec)

center=5165.2*10"(-10); %wavelength (m)

cencm=5165.2*10"(-8);

%wavelength (cm)

nucen=1/cencm;

%wavenumber (cm'l)

em=nucen+nu;

%(cm'1)+(cm'l)

lambda: 1/em;

%(cm)

an g=lambda* 1 .0* 104(8);

364

S=(J+1)*(J-l)/J;

c=3.0*10"(8);

%(m/sec)

nu1=Bvu*Jp*(Jp+1);

En=100.0*h*c*nul;

k=1.38*10"(-23);

%[Joule*(K)"(-1)]

T=2500.0;

arg=¢Xp(-En/(k*T));

amp=S*arg;

%Intensity

a(J)=amp;

lam(J)=ang;

end

plot(lam,a,’ k.’ , lam,a,’ g-’)

365

 

9.5

X-ray Diffraction Crystallography

The most outstanding fact which x-ray diffraction technique has revealed about

crystals is that they are composed of atoms arranged in a pattern periodic in three dimen-

sions. Unlike amorphous materials, a crystal has constant interplanar spacing dhkl which is

a function both of the plane indices (hkl) and the lattice constants of the crystal. From the

Bragg’s law (The diffraction directions are solely determined by the size and shape of the

unit cell of the crystal.):

n)» = 2dsin0

(9.1)

11: order of reﬂection, n=l,2,3,... (n=0, the diffracted beam is in the same direction

as the transmitted beam. It can not be observed.)

A. : wavelength of the light source

(1: interplanar spacing

0: incident and reﬂected angles

With x-ray of known wavelength A , varying incident angle 0, and measuring

peaks, then we can determine the unknown spacing of crystal planes. The intensities of the

diffracted beams which are corresponding to the structure factor of the crystal are the

resultant wave scattered by all the atoms of the unit cell of that crystal. Comparing the

peak-position information of our samples in the intensity versus 20 plot (20: the angle

between diffracted beam and transmitted) which were obtained by the diffraction analysis

with a x-ray diffractometer to ASTM (American Society for Testing Materials) Inorganic

366

 

Powder Diffraction File, we can identify the diamond peaks to prove the presence of the

diamond in the ﬁlms. The different intensity ratio of the peaks to the powder diffraction

ﬁle can be an indicate of crystal orientation/texture of the ﬁlms.

The x-ray diffractometer at Chemistry department of MSU is a Rigaku 200 rotat-

ing anode powder diffractometer with maximum power: 60 kV and maximum current: 200

mA.

9.6 Atomic Force Microscopy

TappingMode AFM operates by scanning a tip attached to the end of an oscillating

cantilever across the sample surface. The cantilever is oscillated at or near its resonance

frequency and it can be on either side of the resonant frequency. During scanning, the tip

lightly taps on the sample surface, contacting the surface at the bottom of its swing. The

tip undergoes deﬂections as it is scanning along the sample surface. The deﬂection is

detected by a laser light from a solid state diode which is reﬂected off the back of the can-

tilever and collected by a position sensitive detector consisting of two closely spaced

photodiodes whose output signal is collected by a differential ampliﬁer. Angular displace-

ment of cantilever results in one photodiode collecting more light than the other

photodiode, producing an output signal (the difference between the photodiode signals

normalized by their sum) which is proportional to the deﬂection of the cantilever. By

maintaining a constant RMS of the oscillation signal acquired by the split photodiode

detector, the feedback loop maintains a constant oscillation amplitude. The topographic

image of the sample is formed by the vertical position of the scanner at each (x,y) data

367

 

J

point in order to maintain a constant setpoint amplitude. A constant tip-sample interaction

is maintained during imaging by maintaining a constant oscillation amplitude.

The probes used in this research were TESP (TappingMode Etched Silicon Probe)

which had nominal tip radius of curvature of 5 -10 nm. For this thesis research, 5 pm x 5

11m images were acquired with the number of samples parameter set to 512, then the pixel

size is 9.8 nm (5 uni/512 = 0.0098 1.1m =9.8 nm). Thus, the features smaller than 9.8 nm

will not be resolved. The vertical resolution is affected by the conversion of 16 bits over

the vertical range of the scanner. For example, a scanner may have a full vertical move-

ment of 6 11m over 440 V. By reducing the Z-limit to 110V, the full vertical movement of

the scanner reduced to 1.5 1.1111 over the same 16 bits. There, the Z resolution is increased

by the smaller vertical sample size.

The AFM system at KMF (W. M. Keck Microfabrication Facility at Michigan

State University, a NSF MRSEC facility) at PHY department of MSU is a SPM (Scanning

Probe Microscopy) model DI D3100.

 

368

10

References:

[Angu]

J .C.Angus, H.A.Will, and W.S.Stanko, J. Appl. Phys. 39, 2915 (1968).

[Angu}

J.C.Angus, F.A.Buck, M.Sunkara, T.F. Groth, C.C. Hayman, and R. Gat, “Dia-

mond Growth at Low Pressures,” MRS Bulletin, pp38-47, Oct.1989.

[Angu}

J.C.Angus, C.C.Hayman, and R.W.Hoffman, in Diamond Optics, SPIE Pro-

ceedings, 969, Intl. Soc. For Optical Engineering, Bellingham, Washington, 2,

1988.

[Asma]

M.Asmann, J.Heberlein, and E.Pfender, Diamond Relat. Mater. 5,], 1999.

[Asma]

M.Asmann, R.F.Cook, J.V.Heberlein, and E.Pfender, J. Mater. Res. 16, 469,

2001.

[Asmu]

J.Asmussen, J.Mossbrucker, S.Khatami, W.S.Huang, B.Wright, and V.Ayres,

“The effect of nitrogen on the growth, morphology, and crystalline quality of

MPACVD diamond ﬁlms,” diamond and Related Materials 8: 220-225, 1999.

[Asmu]

J.Asmussen and W.S.Huang,”Process For Synthesizing Uniform Nanocrystal-

line Diamond,” Practitioner’s Docket Nd. MSU 4.1-572 (ID01-060).

[Asso]

M.B.Assouar, F.Benedic, O.Elmazria, M.Belmahi, R.J.Rioboo, and P.Alnot,

“MPACVD Diamond Films for Surface Acoustic Wave ﬁlters,” Diamond and

Related Materials 10(3-7): 681-685, 2001.

[Bach]

P.K.Bachmann, D.Leers, and H.Lydtin, “Towards a general concept of diamond

chemical vapor deposition”, Diamond and Related Materials, Vol. 1, pp.1-12,

(1991).

[Batt]

C.C.Battaile, D.J.Srolovitz, and J.E.Butler, “Morphologies of Diamond Films

From Atomic-Scale Simulations of Chemical Vapor Deposition,” Diamond and

Related Materials, Vol. 6(9): 1198-1206,]997.

[Begh]

M.G.Beghi, A.C.Ferrari, C.E.Bottani, A.Libassi, B.K.Tanner, K.B.K.Teo,

J .Robertson, “Elastic constants and structural properties of nanometer-thick

diamond-like carbon

ﬁlms,” Diamond and Related

Materials

11(3-6):

1062-1067, 2002

[Begh]

M.G.Beghi, C.S.Casari, A.L.Bassi, C.E.Bottani, A.C.Ferrari, J.Robertson, and

P.Milani, “Acoustic phonon propagation and elastic properties of nano-sized

carbon ﬁlms investigated by Brillouin light scattering,” Thin Solid Films,

420-421, pp.300-305, 2002.

369

 
 

[Bhus]

D.M.Bhusari, J.R.Yang, T.Y.Wang, K.H.Chen, S.T.Lin, and L.C.Chen, “Effects

of substrate pretreatment and methane fraction on the optical transparency of

nanocrystalline diamond thin ﬁlms,” Journal of Materials Research, 13(7):

1769-1773, 1998.

[Bi]

[Bil

B.K.Bi, B.Golding, W.S.Huang, and J.Asmussen, “Surface Acoustic Wave

Devices Based on Unpolished Nanocrystalline Diamond,” Practitioner’s Docket

Nd. MSU 4.1-571 (ID01-061).

B.K.Bi, W.S.huang, J.Asmussen, and B.Golding, “Surface Acoustic Wave on

Nanocrystalline Diamond,” Diamond and Related Materials 11(3-6), 677-680,

2002.

[Bogl]

U.Bogli, A.Blatter, S.M.Pimenov, E.D.Obraztsova, A.A.Smolin, M.Maillat,

A.Leijala, J .Burger, H.E.Hintermann, E.N.Loubnin, “Tribological Properties of

Smooth Polycrystalline Diamond Films,” Diamond and related Materials 4(7):

1009-1019,1995.

[Butl]

J.E.Butler, R.L.Woodin, Philos. Trans. R. Soc. London Ser. A342, (1993),

pp.209-224.

[Case]

M.Casey and J.Wilks, “Friction of Diamond Sliding on Polishing Cube Faces of

Diamond,” J. Phys. D: Appl. Phys. 6(15): 1772-1781, 1973.

[Chen]

L.C.Chen, P.D.Kichambare, K.H.Chen, J .J .Wu, J .R.Yang, and S.T.Lin, “Growth

of highly transparent nanocrystalline diamond ﬁlms and a spectroscopic study

of the growth,” Journal of Applied Physics, 89(1): 753-759, 2001.

[Chen]

Y.Y.Chen, T.T.Wu, and T.T.Chou, “Analysis of the frequency response of a dis—

persive IDT/ZnO/sapphire SAW ﬁlter using effective permittivity and the cou-

pling of modes model,” J. Phys. D: Appl. Phys. 37, pp.120-127, 2004.

[Clay]

K.J.Clay, S.P.Speakman, G.A.J.Amaratunga, and S.R.P.Silva, “Characterization

of a-C:H:N Depsoition from CH4/N2 RF Plasmas Using Optical Emission

Spectroscopy,” J. Appl. Phys. 79(9), 7227, 1996.

[Coll]

C.B.Collins, F.Davanloo, T.J.Lee, J.H.You, and H.Park, in Laser Ablation in

Materials Processing: Fundamentals and Applications, edited by B.Braren,

J .J .Dubowski, and D.Norton (Mater. Res. Soc. Symp. Proc. 285, pp. 547, Pitts-

burgh, PA, 1993.

[Coll]

C.B.Collins, F.Davanloo, J.H.You, and H.Park, in Laser Applications, edited by

A.A.Mak, SPIE Proc. 2097. PP.129, 1994.

370

 

[Csen]

R.Csencsits, D.M.Gruen, A.R.Krauss, and C.Zuiker, “TEM Study of Diamond

Films Grown From Fullerene Precursors,” Mat. Res. Soc. Symp. Proc. Vol. 403,

pp.29l-296, 1996.

[Dava]

F.Davanloo, H.Park, and C.B.Collins, “Protective coatings of nanophase dia-

mond deposited directly on stainless steel substrates,” Journal of Materials

Research, 11(8): 2042-2050, 1996.

[Deli]

B.V.Derjaguin, D.V.Fedosseev, V.M.Lukyanovich, B.V.Spitsyn, V.A.Ryabov,

and A.V.Lavrentyev, J. Cryst. Growth 2, 380 (1968).

[Diam]

Abstracts, 14th European Conference on Diamond, Diamond-Like Materials,

Carbon Nanotubes, Nitrides & Silicon Carbide, 2003.

[Endl]

I.Endler, K.Bartsch, A.Leonhardt, H.J.Scheibe, H.Ziegele, I.Fuchs, C.Raatz,

“Preparation and wear behavior of woodworking tools coated with superth

layers,” Diamond and Related Materials, 8(2-5): 834-839, 1999

[Erde]

A.Erdemir,

C.Bindal,

G.R.Fenske,

C.Zuiker,

R.Csencsits,

A.R.Krauss,

D.M.Gruen, “Tribological characterization of smooth diamond ﬁlms grown in

Ar-C-60 and Ar-CH4 plasmas,” Diamond ﬁlms and Technology, 6(1): 31-47

1996

[Erde]

A.Erdemir, G.R.Fenske, A.R.Krauss, D.M.Gruen, T.McCauley, and R.T.Csenc-

sits, “Tribological properties of nanocrystalline diamond ﬁlms,” Surf. Coat.

Technol. 121: 565-572, 1999.

[EI‘Z]

R.Erz, W.Dotter, K.Jung, and H.Ehrhardt, “Preparation of smooth and nanoc-

rystalline diamond ﬁlms,” Diamond and Related Materials, 2, pp.449-453,

1993.

[Ever]

W.G.Eversole, US. Patent No. 3030188, 23.07.1958.

[Ferr]

A.C.Ferrari, J.Robertson, “Elastic constants of tetrahedral amorphous carbon

ﬁlms by surface Brillouin scattering,” Applied Physics Letters 75(13):

1893-1895, 1999.

[Gabl]

J .Gabler, L.Schafer, and H.Westermann, Diamond Relat. Mater. 9, 921, 2000.

[Gayd]

A.G.Gaydon, “Dissociation Energies and Spectra of Diatomic Molecules,”

London, Chapman and Hall LTD, 1968.

[Gicq]

A.Gicquel, K.Hassouni, Y.Breton, M.Chenevier, and J.C.Cubertafon, “Gas

Temperature Measurements by Laser Spectroscopic Techniques and by Optical

Emission Spectroscopy,’ Diamond and Related Materials 5, 366, 1996.

371

 

[Good]

D.G.Goodwin, J.E.Butler, in:M.A. Prelas, G.Popovia, L.K.Bigelow (Eds),

Handbook of Industrial Diamonds and Diamond Films, Marcel Dekker. New

York, 1996, pp.1-70.

[Goy l l

A.N.Goyette, J .E.Lawler, L.W.Anderson, D.M.Gruen, T.G.McCauley, D.Zhou,

and A.R.Krauss, “C2 Swan Band Emission Intensity as A Function of C2 Den-

sity”, Plasma Sources Sci. Technol. 7, pp.149-153, 1998.

[Goy2]

A.N.Goyette, J .E.Lawler, L.W.Anderson, D.M.Gruen, T.G.McCauley, D.Zhou,

and A.R.Krauss, “Spectroscopic Determination of Carbon Dimer Density in

Ar-HZ-CH4 and Ar-HZ-Cw plasmas”, J. Phys. D: Appl. Phys. vol. 31,

pp.1975-1986, 1998.

[Grim]

M.H.Grimsditch, Anastassakis, E. and Cardona, M.; Phys. Rev. B 5 p 580

(1978)

[Grae]

J.E.Graebner, S.Jin, G.W.Kammlott, B.Bacon, L.Seibles, and W. Banholzer,

Anisotropic, “Thermal Conductivity in Chemical Vapor Deposition Diamond,”

J. Appl. Phys, 71(11): 5353-5356, 1992.

[Grot]

T. Grotjohn, “A Review of CVD Diamond Research,” 1994.

[Grot]

T.A.Grotjohn, W.S.Huang, J .Sivagnaname, J .Asmussen, D.Story, ‘Plasma

Diagnostic Measurements and Numerical Modeling of Argon-Hydrogen-Meth-

ane Discharges used for Nanocrystalline Diamond Deposition in a Microwave

CVD System,” 10th European Conference on Diamond, Diamond-Like Materi-

als, Carbon Nanotubes, Nitrides & Silicon Carbide, 1999.

[Grue]

Deter M.Gruen, Xianzheng Pan, Alan R.Krauss, Shengzhong Liu, Jianshu Luo,

and Christopher M.Foster, J. Vac. Sci. Technol. A 12(4), Jul/Aug 1994.

[Grue]

D.M.Gruen, C.D.Zuiker, A.R.Krauss, and X.Pan, “Carbon dimer, C2, as a

growth species for diamond ﬁlms from methane/hydrogen/argon microwave

plasmas,” J. Vac. Sci. Technol. A 13(3), pp.1628-1632, 1995.

[Grue]

D.M.Gruen, P.C.Redfem, D.A.Horner, P.Zapol, and L.A.Curtiss, “Theoretical

studies on nanocrystalline diamond: Nucleation by dicarbon and electronic

structure of planar defects,”. J. Phys. Chem., B, 103(26): 5459-5467, 1999.

[Hami]

D.R.Hamilton and R.G.Seidensticker, J. Appl. Phys. 34, 1450 (1960).

[Hayw}

1.P.Hayward, I.L.Singer, L.E.Seitzman, “Effect of Roughness on The Friction

of Diamond on CVD Diamond Coatings,” Wear, 157(2): 215-227, 1992

[Herz]

G.Herzberg, “Spectra of Diatomic Molecules,” New York: Van Nostrand Rein-

hold, 1950.

372

 

[Higa]

K.Higaki,

H.Nakahata,

H.Kitabayashi,

S.Fujii,

K.Tanabe,

Y.Seki,

and

S.Shikata, “High Power Durability of Diamond Surface Acoustic Wave Filter,”

IEEE TRansactions on Ultrasonics Ferroelectrics and Frequency Control 44(6):

1395-1400, 1997.

[Holl]

P.Hollman, A.Alahelisten, T.Bjork, S.Hogmark, “CVD-Diamond Coatings in

Sliding Contact With AL, AL-17SI and Steel,” Wear 179(1-2): 11-16, 1994.

[Holl]

P.Hollman, O.Wanstrand, S.Hogmark, “Friction Properties of smooth nanocrys-

talline diamond coatings,” Diamond and Related Materials, 7: 1471-1477,

1998.

[Hongl

T.M.Hong, S.H.Chen, Y.S.Chiou, and C.F.Chen, “Optical Emission Spectros-

copy Studies of the Effect of Nitrogen Addition on Diamond Synthesis in a

CH4-C02 Gas Mixture,” Appl. Phys. Lett. 67(15), 2149, 1995.

[Hongl

S.RHong, H.Yoshikawa, K.Wazumi, Y.Koga, “Synthesis and tribological char-

acteristics of nanocrystalline diamond ﬁlm using CH4/I-I-2 microwave plas-

mas,” Diamond and Related Materials, 11(3-6): 877-881, 2002.

[Huan]

W.S.Huang,

J.Asmussen,

B.Wright,

D.M.Gruen,

S.L.Jiao,

A.R.Krauss,

A.Sumant, M.A.Crimp, and L.Zeng, “Microwave-Plasma Assisted Chemical

Vapor Deposition of Ultra-Nanocrystalline Diamond Films”, CFMR 2000 Con-

ference.

[Huan]

W.S.Huang, J.Asmussen, B.Wright, A.R.Krauss, D.M.Gruen, and A.Sumant,

“The Deposition of Ultra-Nanocrystalline Diamond Films Using a Ar/112/CH4

Microwave Discharge”, ICOP 2000 Conference.

[Huan]

W.S.Huang, J .Asmussen, T.A. Grotjohn, D.M.Gruen, S.L.Jiao, A.R.Krauss, and

A.Sumant, “Microwave-Plasma Assisted Chemical Vapor Deposition of Nanoc-

rystalline Diamond Films”, NSF-MRSEC 2001.

[J ian]

X.Jiang, “Rayleigh mode in amorphous hydrogenated carbon ﬁlms,” Physical

Review B 43(3), 2372-2377, 1991.

[J iao]

S.L.Jiao, “JiaoWorkSummary-DEC99”, Two cross-section TEM dark ﬁeld

images of the MSU-nanocrystalline diamond ﬁlm with SAED (Selected Area

Electron Diffraction) patterns were taken. The SAED pattern in diamond was

obtained from a selected area including the cone-shaped feature. The arcs on

the rings strongly indicate the material is textured. By forming a dark ﬁeld

image with the arc (by selecting the marked spot on the (111) ring), these ori-

ented grains

are

located and shown uniformly distributed

inside

the

cone-shaped object. It’s intrigue to observe that nanocrystalline diamond crys-

tals have the tendency to agglomerate to form the oriented texture. Unpub-

lished.

373

 

[Kani]

D.R.Kania, M.I.Landstrass and M.A.Plano, “Diamond Radiation Detectors,”

Diamond and Related Materials, 2 (1993), pp1012-1019.

[Khat]

Saeid Khatami: Controlled synthesis of diamond ﬁlms using a microwave dis-

charge (Non-equilibrium plasma), Ph. D. thesis at the Michigan State Univer-

sity, Department of Electrical Engineering (1997).

[Klei]

I. R.Kleindienst: Master Thesis at the Michigan State University, Department of

Electrical Computer Engineering 1999.

[Knigl

D.S.Knight, W.B.White, “Characterization of diamond ﬁlms by Raman spec-

troscopy,” J. Materials Research, Vol. 4, No. 2, 1989, pp.385-393.

[Kohz]

M.Kohzaki, S.Noda, Diamond Films Technol. 3(3) 135, 1994.

[Kond]

K.Kondo, S.Sawai, “Fabricating Nanocrystalline Diamond Ceramics by A

Shock Compaction Method,” Journal of The American Ceramic Society, 73 (7):

1983-1991, 1990.

[Kono]

V.I.Konov,

A.A.Smolin,

V.G.Ralchenko,

S.M.Pimenov,

E.D.Obraztsova,

E.N.Loubnin, S.M.Metev, and G.Sepold, “D.C. arc plasma deposition of

smooth nanocrystalline diamond ﬁlms,” Diamond and Related Materials, 4,

pp.1073-1078, 1995.

[Kuo]

K.P.Kuo and J .Asmussen, Diamond Relat. Mater. 6, 1907 (1992).

[Kuo]

Kuo-Ping Kuo: High pressure synthesis of diamond ﬁlms using microwave cav-

ity plasma reactor,Ph. D thesis at the Michigan State University, Department of

Electrical Engineering (1997).

[Lee]

[Lil

S.T.Lee, Z.Lin, and X.Jiang, Mater. Sci. Eng. R 25, 123, 1999.

J .Q.Li, D.Y.He, W.T.Guo, J .H.Zhang, Y.I.Sun, Q.S.Lei, and X.Gao, “Nanocrys-

talline Diamond Thin ﬁlms As Infrared Optical Protective Coatings,” Intema-

tional Journal of Modern Physics B, 16(6—7): 1013-1017, 2002.

[Lin]

T.Lin, G.Y.Yu, A.T.S.Wee, Z.X.Shen, and K.P.Loh, “Compositional mapping of

the argon-methane-hydrogen system for polycrystalline to nanocrystalline dia-

mond ﬁlm growth in a hot-ﬁlament chemical vapor deposition system,” Applied

Physics Letters, 77(17): 2692-2694, 2000.

[Liu]

Huimin Liu and David S. Dandy, Diamond Chemical Vapor Deposition, Noyes

Publications, 1995.

[Luo]

J.S.Luo, D.m.Gruen, and A.R.Krauss, X.Z.Pan, S.Z.Liu, Electrochemical

Scoiecty Proceedings Volume 95-10.

374

[Mats]

S.Matsumoto, Y.Sato, M.Kamo, N.Setaka, “Vapor deposition of diamond parti-

cles from methane“, Jpn. J. Appl. Phys. 21, part2, 1982, pp. L183-185.

[May]

Paul W. May “CVD Diamond - a new Technology for the future”, Endeavour

Magazine 19(3), 1995 pp. 101-106.

[McCa]

T.G.McCauley, D.M.Gruen, and A.R.Krauss, “Temperature dependence of the

growth rate for nanocrystalline diamond ﬁlms deposited from Ar/CH4 micro-

wave plasma,” Appl. Phys. Lett. Vol. 73, No. 12, 1998, pp.1646-1648.

[Mirt]

M.Mirtich, J. C.Angus, “Dual ion beam deposition of carbon ﬁlms with dia-

mondlike properties,” National Aeronautics and Space Administration, Wash-

ington D. C. (1984).

[Miyo}

K.Miyoshi, R.L.C.WU, A.Garscadden, P.N.Barnes, H.E.Jackson, “Friction and

Wear of Plasma-Deposited Diamond Films,” Journal of Applied Physics, 74

(7): 4446-4454, 1993

[Mori]

Y.Mori, H.Yagi, M.Deguchi, M.Kitabatake, K.Nishimura, A.Hatta, T.Ito, T.

Hirao, T.Sasaki, and A.Hiraki, “Crystallinities and electrical properties of

homoepitaxial diamond ﬁlms grown from carbon monoxide”, 2nd Int. Conf. on

the appli. of Diamond Films and Related Materials. Yoshikawa, Vol. 2, Tokyo,

Japan, 1993, pp393-398.

[Mori]

T.Morita, K.Banshoya, T.Tsutsumoto, and Y.Murase, “Effects of Work Materi-

als on cutting performance of diamondOcoated cemented carbide tools,” Forest

Products Journal, 48(5): 43-50, 1998.

[Moss]

J .Mossbrucker: Polarized Raman Spectra of diamond ﬁlms, Dissertation at the

Kaiserslautem University (1997).

[Moss]

EMISSION.BAS, 03-04-1998.

[Moss]

J.Mossbrucker, S.Khatami, W.S.Huang, B.Wright, V.Ayres, and J.Asmus-

sen,”The Effect of Nitrogen on the Growth, Morphology, and Crystalline Qual-

ity of MPACVD Diamond Films,” Diamond and Related Materials 8 (2—5):

220-225, 1999.

[Naka]

H.Nakahata, K.Higaki, A.Hachigo, S.Shikata, N.Fujimori, Y.Takahashi, T.Kaji-

hara, and Y.Yamamoto, “High-Frequency Surface-Acoustic-Wave Filter Using

ZnO/Diamond/Si Structure,” Japanese Journal of Applied Physics Partl-Regu-

lar Papers Short notes & Review Papers, 33(1A): 322-328, 1994.

[N aka]

H.Nakahata, A.Hachigo, K.Higaki, S.Fujii, S.I.Shikata, and N.Fujimori, “Theo-

retical Study on SAW Characteristics of Layered Structures Including a Dia-

375

mond Layer,” IEEE Transactions on Ultrasonics Ferroelectrics and Frequency

Control 42(2): 362-375, 1995.

[Naka]

H.Nakahata, H.Kitabayashi, T.Uemura, A.Higaki, S.Fujii, Y.Seki, K.Yoshida,

and S.Shikata, “Study on Surface Acoustic Wave Characterization of

Si02/interdigital-transducer/ZnO/Diamond Structure and Fabrication of 2.5

GHz Narrow Band Filter,” Japanese Journal of Applied Physics Partl-Regular

Papers Short notes & Review Papers, 37(5B): 2918-2922 Sp. Iss. SI 1998.

[Naka]

H.Nakahata, S.Fujii, K.Higaki, A.Hachigo, H.Kitabayashi, S.Shikata, and

N.Fujimori, “Diamond-based surface acoustic wave devices,” Semiconductor

Science and Technology, 18, 896-8104, 2003.

[Neer]

D.Neerinck, P.Persoone, M.Sercu, and A.Goel, Thin Solid Film 317, 402, 1998.

[Nied]

P.Niedzielski, S.Miklaszewski, P.Beer, and A.Sokolowska, “Tribological prop-

erties of NCD coated cemented carbides in contact with wood,” Diamond and

Related Materials, 10(1): 1-6, 2001.

[Nist]

L.C.Nistor, J .Van Landuyt, V.G.Ralchenko, E.D.Obraztsova, and A.A.Smolin,

“N anocrystalline diamond ﬁlms: transmission electron microscopy and Raman

spectroscopy characterization, “Diamond and related Materials, 6, pp.159-168,

1997.

[Nuss]

A.Nussbaum and R.A.Phillips, “Contemporary Optics for Scientists and Engi-

neers”, Prentice Hall Inc, Englewood Cliff, NJ 1976, pp.191.

[Ongl

T.P.Ong and R.P.H.Chang, “Low-temperature deposition of diamond ﬁlms for

optical coating,” Appl. Phys. Lett., 55(20): 2063-2065, 1989.

[Orr]

R.Orr, “The impact of thin ﬁlm diamond on advanced engineering systems,”

presented at the global business and technical outlook for high performance

inorganic thin ﬁlm coatings, Gorham advanced materials institute, Monterey,

California, Oct.30-Nov. l ,1988.

[Pan]

L.S.Pan, S.Han, D.R.Kania, M.A.Plano, and M.I.Landstrass, “Electrical Prop-

erties of High Quality Diamond Films”, Diamond and Related Materials, Vol. 2,

1993, PP820-824.

[Pier]

H.O.Pierson, Handbook of Carbon, Graphite, Diamond and Fullerenes, Noyes

Publications, Park Ridge, New Jersey 1993.

[Pras]

C.V.V.Prasad and P.F.Bemathm,”Fourier Transform Spectroscopy of The Swan

((1311 g.1311 11) System of The Jet-Cooled C2 Molecule”, Astrophy. 1., Vol. 426,

pp.812-821, 1994.

376

 

[Qadr]

S.B.Qadri, C.Kim, E.F.Skelton, T.Hahn, and J.E.Bulter, “THermal Expansion

of Chemical Vapor Deposition Grown Diamond Film, Thin Solid Films,

236(1-2): 103-105, 1993.

[Qin]

L.C.Qin, D.Zhou, A.R.Krauss, and D.M.Gruen, NanoStructure Materials,

Vol.10, No.4, pp.649-660, 1998.

[Rang]

S.Ranganathan, Acta. Crystallogr. 21, 197, 1966.

[Redf]

P.C.Redfem, D.A.Homer, L.A.Curtiss, and D.M.Gruen, “Theoretical Studies of

Growth of Diamond (110) from Dicarbon,” J. Phys. Chem. 100(28):

pp.11654-11663, 1996.

[Rein]

D.K.Reinhard and I.R.Kleindienst, “Diamond ﬁlms as optical coatings for sili-

con,” New Diamond and Frontier Carbon Technology 10(1): 13-23, 2000.

[Shar]

T.Sharda, M.M.Rahaman, Y.Nukaya, T.Soga, T.Jimbo, and M.Umeno, “Struc-

tural and optical properties of diamond and nano-diamond ﬁlms grown by

microwave plasma chemical vapor deposition,” Diamond and Related Materi-

als, 10, pp.561-567, 2001.

[Shar]

T.Sharda, T.Soga, and T.Jimbo, “Optical properties of nanocrystalline diamond

ﬁlms by prism coupling technique,” J. Appl. Phys., 93(1): 101-105, 2003.

[Shio]

Hiromu Shiomi, Keiichirou Tanabe, Yoshiki Nishibayashi and Naoji Fujimori,

“Epitaxial Growth of High Quality Diamond Film by the Microwave

Plasma-Assisted Chemical-Vapor—Deposition Method”, Japanese Journal of

Applied Physics, Vol. 29, No. 1, January, 1990, pp34-40.

[Silv]

M.Silveira, M.Becucci, E.Castelucci, F.Polla Mattiot, V.Barbarossa, R.Toma-

ciello, EGalluzi, “Non-diamond carbon phases in plasma-assisted deposition of

crystalline diamond ﬁlms: a Raman study,” Diamond and Related Materials 2

(1993), p. 1257-1262.

[Silv]

F.J.G.Silva, A.J.S.Femandes, F.M.Costa, V.Teixeira, A.P.M.Baptista, E.Pereira,

“Tribological behavior of CVD diamond ﬁlms on steel substrates,” Wear, 255:

846-853 Part 2, 2003

lSpitl

B.V.Spitsyn, L.L.Bouilovand and B.V.Derjaguin,”Vapor Growth of Diamond on

Diamond and Other Surfaces,” J. Cryst. Growth 52, 2 (1981), pp219-226.

[Suma]

A.Sumant: This ﬁlm looks smoother than the previous one. I am thinking of

sending that seal to Tony at ﬂowserve to have look at its waviness then I can test

that seal here at ANL. Conversations on August 29th 2000.

377

[Sun]

Y.N.Sun, H.Z.Yun, W.T.Guo, and J .Q.Li, “Diamond IR optical coatings by hot

ﬁlament chemical vapor deposition,” Thin Solid ﬁlms, 212, pp.137-139, 1992.

[Tamh]

Amit Arvind Tamhane: Growth ledges and morphological development in dia-

mond ﬁlms, (1992) Ph.D. thesis, Ohio State University.

[Tang]

C.J.Tang, A.J.Neves, A.J.S.Femandes, J .Gracio, and N.A1i, “A new elegant

technique for polishing CVD diamond ﬁlms,” Diamond and Related Materials,

12. pp.1411-1416, 2003.

[Tenn]

S.Tennant, “On the Natural of the Diamond,” Phil. Trans. Roy. Soc. London,

87, 123 (1797).

[Thom]

M.Thomas and W.J.Tropt, Johns Hopkins APL Technical Digest, 14, pp.16-23,

1993.

[Thor]

A.P.Thome, “Spectrophysics,” London, Chapman and Hall, 1988.

[Ulcz]

M.J.Ulczynski, “Low-Temperature Deposition of Transparent Diamond Films

with a Microwave Cavity Plasma Reactor,” Ph. D thesis at the Michigan State

University, Department of Electrical Engineering (1998).

[Ulcz]

M.J.Ulczynski, B.Wright, and D.K.Reinhard,”Diamond-coated glass sub-

strates,” Diamond and Related Materials, 7(11-12):1639-1646, 1998.

[Vavi]

V.S. Vavilov, “Diamond as a Material in Solid State Electronics,” Mat. Res.

Soc. Symp. Proc. Vol.242, 1992, pp87-95.

[Wang]

T.Wang, H.W.Xin, Z.M.Zhang, Y.B.Dai, and 1-1.S.Shen, “The fabrication of

nanocrystalline diamond ﬁlms using hot ﬁlament CVD,” Diamond and Related

Materials, 13, pp.6-13, 2004.

[Wang]

X.H.Wang, L.Pi1ione, W.Zhu, W.Yarbrough, W.Drawl, and RMessier, “Infra-

red Optical Properties of CVD Diamond Films,” J. Mater. Res., 5, 2345, 1990.

[Wild]

C.Wild, W.Muler-Sebert, T.Bckermann and P.Koidl,“Applications of Diamond

Films and Related Materials,” Editor: Y.Tzeng, M.Yoshikawa, M.Murakawa,

and A.Feldman, Elservier Science Publishers. B.V. 1991.

[Xu}

T.Xu, S.R.Yang, M.Chen, J.Tian, Q.Xue, J.Li, and W.T.Guo, “Inﬂuence of

Nitrogen Ion Implantation on Tribological Properties of Nanocrystalline Dia-

mond Films,” J. Phys. D: Appl. Phys. 35(8), 788-793, 2002.

[Yama]

K.Yamanouchi, N.Sakumi, T.Satoh, Proceedings IEEE Ultrasonics Sympo-

sium, p351, 1989.

378

 

 

-
_
f
m
i
‘
u
+

 

[Yang]

W.B.Yang, F.X.Lu, and Z.X.Cao, “Growth of nanocrystalline diamond protec-

tive coatings on quartz glass,” J. Appl. Phys., 91(12): 10068-10073, 2002.

[Yosh]

H.Yoshikawa, C.Morel, and Y.Koga, “Synthesis of nanocrystalline diamond

ﬁlms using microwave plasma CVD,” Diamond and Related Materials, 10,

pp.1588-1591, 2001.

[Yosh]

M.Yoshikawa, G.Katagiri, H.Ishitani, A.Ishitani, M.Ono, K.Matsumura, “Char-

acterization of crystalline quality of diamond ﬁlms by Raman spectroscopy,”

Applied Physics Letters 55 (1989). 1313.2608-2610.

[Yosh]

M.Yoshikawa, G.Katagiri, H.Ishida, A.Ishitani, T.Akamatsu, “Raman spectra of

diamondlike amorphous carbon ﬁlms,” Journal of Applied Physics 64 (1988), p.

6464-6468.

[Zait]

A.M.Zaitsev, “Optical Properties”, Chapter 7, M.A.Prelas, G.Popovich, and

L.K.Bigelow, Hand book of Industrial Diamonds and Diamond Films, Marcell

Dekker, Inc., New York, 232, 1998.

[Zarr]

M.Zarrabian, C.N.Fourches, G.Turban, C.Marhic, and M.Lancin, “Observation

of nanocrystalline diamond in diamondlike carbon ﬁlms deposited at room tem-

perature in electron cyclotron resonance plasma,” Applied Physics Letters,

70(19):2535-2537, 1997.

[Zhan]

Jie Zhang: Experimental development of microwave cavity plasma reactors for

large area and high rate diamond ﬁlm deposition, Ph. D. thesis at the Michigan

State University, Department of Electrical Engineering (1993).

[Zhan]

J .Asmussen and J .Zhang,”Apparatus for the coating of Material on a Substrate

Using 3 Microwave or UHF Plasma,”U.S.Patent N.5,311,103,May 10,1994.

[Zhou]

D.Zhou, D.M.Gruen, L.C.Qin, T.G.McCauley and A.R.Krauss, J. Appl. Phys.

84, 1981 (1998).

thu}

W.Zhu, A.R.Badzian, and R.Messier, J. Mater. Sci. Lett. 2, 785, 1983.

[Zhu]

W.Zhu, A.Inspektor, A.R.Badzian, T.Mckenna and R.Messier, ‘Effect of

Noble-Gases on Diamond Deposition from Methane-Hydrogen Microwave

Plasmas,’ J. Appl. Phys. 68, 1489, 1990.

[Zhu]

W.Zhu, B.R.Stoner, B.E.Williams, and J.T.Glass, “Growth and Characterization

of Diamond Films on Nondiamond Substrates for Electronic Applications,”

Proceedings of IEEE 79, 5, p621-646, 1991.

379

 

[Zuik]

C.Zuiker, A.R.Krauss, D.M.Gruen, X.Pan, J .C.Li, R.Csencsits, A.Erdemir,

C.Bindal, and G.Fenske, “Physical and tribological properties of diamond ﬁlms

grown in argon-carbon plasmas,” Thin Solid Films, 270, pp.154-159, 1995.

 

380