'-'. a, “539's Milli\lllllllllllll lilalllzlll‘llllllllllll ‘, 3 12930 ‘ LIBRARY Michigan State University This is to certify that the dissertation entitled MICROWAVE ASSISTED PLASMA CVD OF DIAMOND FILMS USING A THERMALzLIKE PLASMA DISCHARGE presented by Kuo-Ping Kuo has been accepted towards fulfillment of the requirements for Ph.D. degree in Electrical Eng M d V Major professor \ \ Date a! ”I 2] MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 (0 ml 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 DATE DUE W JCT a 52003 b92305 DEC 1 6 2005 £42810 p iMAR 1) 3 2613‘- -" 1M WWW" MICROWAVE ASSISTED PLASMA CVD OF DIAMOND FILMS USING THERMAL-LIKE PLASMA DISCHARGE By Kuo-Ping Kuo A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Electrical Engineering 1997 ABSTRACT Microwave Assisted Plasma CVD of Diamond Films Using A Thermal-like Plasma Discharge By Kuo-Ping Kuo Two new microwave plasma reactor configurations have been devel- oped to improve the deposition efficiency and controllability of microwave plasma reactors. The first concept extends the conventional microwave plasma reactor from a 20-80 Torr, nonequilibrium discharge regime to the higher pressure 80-150 Torr, thermal-like discharge operating regime. This new reactor configuration includes a water cooling stage and an improved input feed gas flow configurations. A thermal-like discharge is created and uniform diamond films are deposited over 2" diameter sub- strates with average growth rates of 5-6 um/h(45-50 mg/h). Growth emery efficiencies of 70 kW-h/ g (~ 20 kW-h/ ct) and carbon conversion efficiencies of 7 -10% are comparable to the best performance reported in the literature. This reactor concept opens up the possibility of applying microwave plasma assisted CVD of diamond to thick film deposition applications. The second of these prototype reactors, i.e., microwave plasma jet reactor, is operated at pressure between 10 to 45 Torr. It utilizes an unique force-flow concept to create a microwave plasma discharge in a region where the electromagnetic field strength is low. The force flow fea- ture reduces gas bypassing and forces the excited species and radicals onto irregular and complex shaped conducting substrates. Mixtures of CO, CH4, and H2 source gases together with this force-flow concept are utilized to enhance the diamond film growth on thirty six 1/ 8" diameter tungsten carbide round tools and thousands of 5-7 um diameter carbon fibers. COPYright by Kuo-Ping Kuo To my parents Fonro-Yisi and Miko ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation to Professor Dr. Jes Asmussen, Jr. for his guidance, encouragement, and support for this thesis research. Thanks are also due the other members of the author’s quidance committee: Professor Dr. Donnie Reinhard, Professor Dr. Timothy Grotjohn and Professor Dr. Brad Golding. In addition, the author owes a debt of gratitude to Dr. Jie Zhang, Dr. Saeid Khatami, Dr. Jorge Mossbrucker, Dr. PengUn Mak, Uwe Kahler, and Wen-shin Huang for providing me with their knowledge. The author have to thank Roxanne Peacock and Brian Wright for their technical support. Last, the author would like to thank Professor Dr. Stanley L. Flegler for training me SEM and Professor Dr. John McGrath for providing thermal conductivity infor- mation on the CVD diamond films. Sincere appreciation is also extended to Dr. Chow Ling Chang and Brian Cline at Norton Diamond Film Corp. and the Michigan Research Excellence Fund for the financial assistance which made this thesis research possible. vi TABLE OF CONTENT LIST OF TABLES ....................................... xviii LIST OF FIGURES ....................................... xiv CHAPTER ONE Introduction 1 . 1 Introduction ........................................ l 1 .2 Research Objectives .................................. 4 1.3 Dissertation Outline .................................. 4 CHAPTER TWO Literature Review of Diamond Film Deposition Reactors 2. 1 Introduction ........................................ 6 2.2 A General Description of Chemical Vapor Deposition (CVD) of Diamond Films ...................................... 6 2.3 The Multivariable Diamond Film Deposition Reactor .......... 9 2.3.1 Experimental Variables of a Diamond Film Deposition Reactor ....................................... 9 2.3.2 “Figures of Merit” of Reactor Performance ............ 1 l 2.4 Diamond Thin Film CVD Reactor Technologies ............. 12 2.4.1 Notes on Literature Review ........................ 12 2.4.2 Hot Filament CVD (HFCVD) Reactors ................ 14 2.4.2. 1 Conventional HFCVD Reactor ................. 14 2.4.2.2 Electron-assisted HFCVD Reactor .............. 17 2.4.3 Direct Current(DC) Plasma CVD Reactors ............ 20 2.4.3. 1 Conventional DC plasma CVD Reactor .......... 20 2.4.3.2 Enclosed DC Arc Jet CVD Reactor ............. 23 2.4.4 Combustion Flames ............................. 29 2.4.4.1 Atmospheric Pressure Combustion Flame ........ 29 2.4.4.2 Enclosed Flat Combustion Flame .............. 32 2.4.5 RF Plasma CVD Reactors ......................... ' 35 vii 2.4.5. 1 Conventional RF Plasma CVD Reactor .......... 35 2.4.5.2 RF Thermal Plasma CVD Torch ............... 37 2.5 Microwave Plasma CVD Reactors ........................ 40 2.5. l Magneto-microwave Plasma CVD Reactors ............ 40 2.5.2 Tubular Microwave Plasma Reactors ................ 44 2.5.3 Microwave Plasma Jet Reactors .................... 47 2.5.4 Bell Jar Microwave Plasma CVD Reactors ............ 50 2.5.4. 1 Astex Bell Jar Microwave Plasma CVD Reactor. . . . 50 2.5.4.2 UC-Berkeley Bell Jar Microwave Plasma CVD Reactor .................................. 53 2.5.4.3 MSU Bell Jar Microwave Plasma CVD Reactor . . . . 56 2.5.5 Surface-wave Microwave Plasma CVD Reactors ........ 59 2.6 Summary ....................................... 62 CHAPTER THREE High-Pressure Microwave Cavity Plasma Reactor: Experimental Systems, Experimental Procedures, Experimental Parameter Space, Measurement Methodologies, and Reactor Configuration 3. I Introduction ....................................... 67 3.2 Experimental Systems ................................ 68 3.2. 1 Introduction ................................... 68 3.2.2 Microwave Power Supply and Waveguide / Transmission system ............................ 68 3.2.3 Flow Control and Vacuum Pump System ............. 71 3.2.4 Computer Monitoring System ...................... 74 3.3 Common Experimental Procedures ...................... 76 3.3. l Seeding Procedure .............................. 76 3.3.2 Start-up and Shut-down Procedures ................ 77 3.4 Experimental Parameter Space ......................... 78 3.4. 1 Introduction ................................... 78 3.4.2 Independent Experimental Input Variables ........... 80 3.4.3 Dependent Experimental Internal Variables ........... 82 3.4.4 External Experimental Output Variables ............. 82 3.5 Measurement of Experimental Output Variables ............ 83 3.5. 1 Measurement of Film Characteristics ................ 83 3.5.1. 1 Measurement of Film Uniformity .............. 83 3.5.1.2 Measurement of Film Structural Quality ......... 83 3.5.1.3 Measurement of FIlm Texture ................. 84 3.5. 1.4 Measurement of Film Morphology .............. 85 3.5.2 Calculation of Reactor Performance ................. 87 3.5.2. 1 Measurement of Film Growth Rate ............. 87 3.5.2.2 Measurement of Specific Yield ................ 87 3.5.2.3 Measurement of Gas Efficiency ................ 88 3.5.2.4 Measurement of Carbon Conversion Efficiency . . . . 88 viii 3.6 Design of High-Pressure MCPR ......................... 89 3.6. 1 Introduction ................................... 89 3.6.2 Operation of Moderate-Pressure MCPR ............... 91 3.6.2.1 Operation of Microwave Cavity Excitation and Tuning .................................. 91 3.6.2.2 Operation of Air Cooling System ............... 91 3.6.2.3 Operation of Thermally Floating Substrate Holder Setup ................................... 93 3.6.3 Substrate Cooling Stage - The First Design ........... 94 3.6.3. 1 Introduction .............................. 94 3.6.3.2 The Change in The Air Cooling System .......... 94 3.6.3.3 Substrate Cooling Stage - The First Design ....... 96 3.6.3.4 Drawbacks of The First Design ................ 98 3.6.3.5 Final Substrate Cooling Stage Design . . . . . ..... 100 3.7 Summary ........................................ 104 CHAPTER FOUR Reactor Experimental Output Performance - Reactor Performance (Y 1) 4.1 Introduction ...................................... 105 4.2 Reactor Performance (Y 1)- Linear Growth Rate ............ l 10 4.2.1 Linear Growth Rate=f (c, ft,TS,t) ................... 1 10 4.2.2 Linear Growth Rate=f ( Pt-p) ...................... l 19 4.3 Reactor Performance (Y 1)- Specific Yield ................. 124 4.3. 1 Specific Yield=f (c, ft, Ts, t) ....................... 124 4.3.2 Specific Yield=f ( Pt-p) ........................... 127 4.4 Reactor Performance (Y 1)- Carbon Conversion Efficiency ..... 128 4.4.1 Carbon Conversion Efficiency=f ( c, ft, TS, t) .......... 128 4.4.2 Carbon Conversion Efficiency=f ( Pt-p) .............. 132 4.5 The Growth of Thick Free Standing Diamond Films ......... 133 4.6 Summary ........................................ 133 ix 6.1 CHAPTER FIVE Reactor Experimental Output Performance- Film Characteristics (Y2) Introduction ...................................... 136 Effect of Methane Concentration (c) ..................... 139 5 .2. 1 Introduction .................................. 1 39 5.2.2 Experimental Variables ......................... 139 5.2.3 Growth Rate=flc) .............................. 140 5.2.4 Film Morphologmflc) ........................... 143 5.2.5 Film Texture=flc) .............................. 146 5.2.6 Film Structural Quali =flc) ...................... 149 5.3 Effect of Total Flow Rate (ft) ........................... 152 5.3. 1 Introduction .................................. 1 52 5.3.2 Experimental Variables ......................... 152 5.3.3 Growth Rate=flft) .............................. 153 5.3.4 Film Morphology=flft) ........................... 154 5.3.5 Film Texture=f(ft) .............................. 158 5.3.6 Film Structural Quality=f(ft) ...................... 161 5.4 Effect of Substrate Temperature (Ts) .................... 164 5.4. 1 Introduction .................................. 1 64 5.4.2 Experimental Variables ......................... 164 5.4.3 Growth Rate=f(Ts) ............................. 165 5.4.4 film Morphology=flTs) .......................... 167 5.4.5 Film Texture=fl’1‘s) ............................. 170 5.4.6 Film Structural Quality=flTs) ..................... 173 5.5 Effect of Deposition Time(t) ........................... 175 5.5.1 Introduction .................................. 17 5 5.5.2 Experimental Variables ......................... 17 5 5.5.3 Growth Rate=flt) .............................. 176 5.5.4 Film Morphologmflt) ........................... 17 7 5.5.5 film Texture=f(t) .............................. 180 5.5.6 Film Structural Quality=flt) ...................... 183 5.6 Summary ........................................ 185 CHAPTER SIX Development and Optimization of the Microwave Plasma Jet Reactor Introduction ...................................... 187 6.2 Development of the MPJR ............................ 188 6.2. 1 Experimental Systems .......................... 188 6.2.2 Early Version of MPJR .......................... 189 6.2.3 Basic Configuration of the MPJR for Diamond Film Deposi- tion on Round Tools ............................ 192 6.3 Optimization of the MPJR for Diamond Film Deposition on Round Tools ............................................ 194 6.3. 1 Introduction .................................. 194 6.3.2 Optimization ................................. 194 6.3.3 Operation of the Final MPJR ..................... 201 6.4 Summary ........................................ 203 CHAPTER SEVEN The Experimental Performance of the Microwave Plasma Jet Reactor 7. 1 Introduction .................................... 205 7.2 Experimental Input/ Output Variable Space ............ 205 7.3 Experimental Procedures .......................... 209 7 .3. 1 Substrate Pretreatment Procedure ................. 209 7.3.2 Start-up and Shut-down Procedures ............... 210 7.3.3 Measurement of Substrate Temperature ............ 210 7.3.4 Measurement of Reactor Output Variables (Y) ........ 210 7 .3.4. 1 Measurement of Reactor Performance Variable W1) ............................. 21 1 7.3.4.2 Measurement of Film Characteristic Variables (Y2) ............................ 21 1 7.4 Diamond Film Deposition on WC-6%Co Round Tools ..... 21 1 7.4.1 Reactor Configuration .......................... 212 7.4.2 Experimental Controllable Input Variables (U) and Internal Input Variables (X) ............................. 212 7.4.3 Reactor Performance (Y 1) - Linear Growth Rate ....... 214 7.4.4 Film Characteristic (Y2) ......................... 215 7.4.4.1 Film Uniformity .......................... 215 7.4.4.2 Film Morphology .......................... 217 7.4.4.3 Film Structural Quality .................... 222 7.5 Diamond Film Deposition on Carbon Fibers ............ 224 7.5.1 Reactor Configuration .......................... 224 7.5.2 Experimental Controllable Input Variables (U) and Internal Input Variables (X) ............................. 226 7.5.3 Film Characteristic (Y2) ......................... 227 7.6 Summary ...................................... 231 CHAPTER EIGHT Conclusions 8. 1 Introduction ...................................... 232 8.2 Operational Characteristics of the High-Pressure MPCR ..... 234 8.2. 1 Introduction .................................. 234 xi 8.2.2 Diamond Film Deposition on 2” Silicon Wafers ........ 235 8.3 Operational Characteristics of the MPJR ................. 239 8.3. 1 Diamond Film Deposition on 1/8” diameter WC—Co Round Tools ....................................... 239 8.3.2 Diamond film Deposition on 5-7 pm diameter Carbon Fibers ...................................... 239 8.4 Recommendations for Future Research .................. 240 8.4.1 High-Pressure MPCR ........................... 240 8.4.2 MPJRs ...................................... 241 APPENDIX A ........................................... 243 LIST OF REFERENCES ................................... 248 xii LIST OF TABLES Table 2. 1 Summary of “figures of merit” ........................ 64 Table 3. 1 Definition of the experimental variables for the high-pressure MCPR. ................................................ 81 Table 3.2 Location of XRD peaks and their relative intensity in a powder pattern. ............................................... 85 Table 4. 1 The Ranges of reactor variables investigated in this chapter. .................................................... 1 10 Table 7. l Reactor input/ internal/ output variables investigated in this and next chapters. ...................................... 209 Table A. 1 Experimental conditions of the diamond films presented described in Chapter 3 and 4 ............................... 243 xiii LIST OF FIGURES Figure 2.1 A general process for diamond film growth by chemical vapor deposition[3]. ............................................ 3 Figure 2.2 Multivariables of a diamond film deposition reactor ....... 10 Figure 2.3 Schematic diagram of a conventional HFCVD reactor [18],[19]. ..................................................... 15 Figure 2.4 Schematic diagram of an Electron-assisted HFCVD reactor[22]. ..................................................... 18 Figure 2.5 Schematic diagram of the DC plasma CVD reactor[23]. . . . 21 Figure 2.6 Schematic diagram of a DC plasma jet CVD reactor[26].. . . 24 Figure 2.7 Schematic illustration of an atmospheric pressure DC are jet CVD reactor and its substrate biasing system[27]. ............... 27 Figure 2.8 Schematic diagram of an atmospheric-pressure oxygen-acety- lene combustion flame[29] .................................. 30 Figure 2.9 Schematic of an enclosed flat flame[32] ................ 33 Figure 2. 10 A schematic diagram of a conventional RF plasma CVD reac- tor[35] ................................................. 36 Figure 2. 1 1 A schematic diagram of a RF thermal plasma torch[36][37]. ..................................................... 38 Figure 2. 12 The schematic drawing of a magneto-microwave plasma CVD reactor[38],[39]. ......................................... 41 Figure 2. 13 Schematic diagram of a tubular microwave plasma-assisted CVD reactor[l] ........................................... 45 xiv Figure 2. l4 Schematic drawing of a microwave plasma jet torch[41]. . ..................................................... 48 Figure 2.15 Schematic drawing of the Astex bell jar microwave PACVD reactor[43l-[46]. ......................................... 51 Figure 2.16 Schematic of the 4" UC-Berkeley bell jar PACVD system[47]- [50]. .................................................. 54 Figure 2. 17 Schematic diagram of the moderate-pressure microwave cav- ity plasma reactor (MCPR)[9],[51] ............................. 57 Figure 2. 18 Schematic illustration of a surface-wave microwave PACVD apparatus[52]-[54] ........................................ 60 Figure 2. 19 Summary of linear growth rate vs. area power density for dia- mond film deposition reactors. .............................. 65 Figure 2.20 Summary of linear growth rate vs. pressure for diamond film deposition reactors. ...................................... 66 Figure 3. 1 Microwave power supply and microwave waveguide / transmis- sion system for high—pressure MCPR and MJPR .................. 70 Figure 3.2 Flow control and vacuum system for high-pressure MCPR and MPJR. ................................................ 73 Figure 3.3 Flow chart of Computer monitor program[9]. ........... 75 Figure 3.4 Multivariables for the high-presure MCPR a diamond film dep- osition reactor ........................................... 79 Figure 3.5 A typical SEM photo from JEOL 6400 at MSU ........... 86 Figure 3.6 Reactor operating field map under thermally floating configu- ration for 5" quartz dome/ 3” substrate reactor configuration[5 1] . . . . 90 Figure 3.7 The cross sectional view of the thermally floating MCPR sys- tem. .................................................. 92 Figure 3.8 The cross sectional view of the high-pressure MCPR system. ..................................................... 95 Figure 3.9 Typical substrate holder setup used in high-pressure MCPR for diamond film deposition: (a) 2" silicon wafer, (b) 2" graphite substrate holder, (c) boron nitride ring, ((1) insulation discs, (e) graphite flow regula- XV tor, (f) water coolling stage .................................. 97 Figure 3.10 Can Body of the 3" Water Cooling Stage[107] .......... 99 Figure 3.1 1 Bottom Body of the 3" Water Cooling Stage[107]. ....... 99 Figure 3.12 Typical substrate holder setup used in high-pressure MCPR for diamond film deposition: (a) 2” silicon wafer, (b) 2” molybdnrum sub- strate holder, (c) boron nitride ring, ((1) insulation discs, (e) graphite flow regulator, (i) water coolling stage. ........................... 101 Figure 3.13 Can Body of the 3" Water Cooling Stage[107] .......... 102 Figure 3.14 Bottom Body of the 3" Water Cooling Stage[107]. ...... 102 Figure 4. 1 High-pressure MCPR block diagram for the experiments described in this chapter .................................. 109 Figure 4.2 Summary of growth rate vs. methane / hydrogen concentration in high-pressure MCPR. .................................. 1 14 Figure 4.3 Summary of growth rate vs. total flow rate in HPMCPR. . . 114 Figure 4.4 Summary of linear growth rate vs. substrate temperature. 1 15 Figure 4.5 Summary of linear growth rate vs. deposition time. ..... I 15 Figure 4.6 Summary of growth rate vs. methane / hydrogen concentraion and substrate temperature in the high-pressure MCPR ........... 1 16 Figure 4.7 Locations of different methane concentrations for the experi- ments that were investigated in this thesis. The solid curve represents the diamond growth region for the moderate pressure MCPR[9),[51] and the dash curve represents the growth region for the high pressure MCPR dis- cribed in this thesis ...................................... 1 17 Figure 4.8 Location of diamond zones vs. c, ft, and T8 for big pressure and moderater pressurel51] MCPRS .......................... 1 18 Figure 4.9 Summary of growth rate vs. pressure for the high pressure MCPR. ............................................... l 19 Figure 4.10 Summary of film growth rate vs. absorbed power ....... 120 xvi Figure 4.1 1 The effect of %CH4/H2 on the specific yield in high-pressure MCPR ................................................ 125 Figure 4.12 Summary of specific yield vs. total flow rate in the high-pres- sure MCPR. ........................................... 125 Figure 4. 13 Summary of specific yield vs. Ts in high pressure MCPR. .................................................... 126 Figure 4.14 Summary of specific yield vs. deposition time in high-pres- sure MCPR. ........................................... 126 Figure 4.15 Summary of specific yield vs. absorbed micorwave power in HPMCPR .............................................. 127 Figure 4. 16 Summary of carbon conversion efficiency vs. %CH4/H2 .................................................... 128 Figure 4.17 Summary of carbon conversion efficiency vs. total flow rate. .................................................... 129 Figure 4.18 Summary of carbon conversion efficiency vs. Ts ....... 130 Figure 4.19 Summary of carbon conversion efficiency vs. deposition time. .................................................... 131 Figure 4.20 Summary of carbon conversion efficiency vs. absorbed micro- wave power in the high pressure MCPR ....................... 132 Figure 5. 1 Linear growth rates (weight gain) versus methane concentra- tions. The experimental conditions are shown in Section 5.2.2 ..... 142 Figure 5.2(a)-(h) Surface morphology vs. %CH4/H2 varied from 1% to 8%. .................................................... 145 Figure 5.3 The X-ray diffraction spectra as a function of methane concen- trations of (a) 2%, (b) 3%, (c) 4%, (d) 5%, (e) 6%, (i) 7% and (g) 8%. Maxi- mum ratio of 1(220) / HI 1 1) was observed at c=4% and 5%. They are summarized in Figure 5.4. ................................ 147 Figure 5.4 The ratio of the XRD peak height of (220) to (1 1 1) versus the methane concentration. The experimental conditions are shown in Section 5.2.2 ........................................... 148 Figure 5.5 Raman spectra of various methane concentrations.The xvii Experimental conditions are shown in Section 5.2.2. ............ 150 Figure 5.6 FWHM of Raman spectra as a function of CH4/H2 concentration. The experimental conditions are shown in Section 5.2.2.1 .................................................... 151 Figure 5.7 Linear growth rate as a function of total flow rate at fixed methane/ hydrogen concenctration(3%). ...................... 154 Figure 5.8(a)-(f) Film morphologies vs. total flow rates. The experimental conditions are shown in Section 5.3.2. ....................... 157 Figure 5.9 XRD spectra of diamond films synthesized by CH4/H2 flow rates of (a)12/400 sccm, (b)18/ 600 sccm, (c) 24/800 sccm, (d) 30/ 1000 sccm, (e) 36/ 1200 sccm and (f) 42/ 1400 sccm, using the experimental conditions described in Section 5.3.2 ......................... 159 Figure 5.10 The XRD height of (220) relative to that of (1 1 1) as function of total flow rate in microwave plasma cavity system. The experimental con- ditions for these experiments are described in Section 5.3.2. . . . . . . 160 Figure 5.1 1 The effect of total flow rate on Raman spectrum. The experimental conditions are listed in Section 5.3.2 ............... 162 Figure 5.12 The effect of total flow rate on FWHM of Rarnan spectrum. The experimental conditions are shown in Section 5.3.2. . . . . . . . . . 163 Figure 5.13 The effect of Ts on the film growth rate .............. 166 Figure 5.14(a)-(f) Film morphologies vs. substrate temperatures. The experimental conditions are described in Section 5.4.2. .......... 169 Fi e 5.15 X-ray diffraction s ectra of diamond films s thesized for (a) 9 OC, (b) 10250, (c) 10500, (d 1075C, (e) 1100C and ( 11250. Maxi- mum ratio of I(220)/I(1 1 1) was observed at (e) 1 1000. They are summa- rized in Figure 5.16 ...................................... 171 Figure 5.16 The XRD peak height of (220) relative to that of (1 1 1) as a function of substrate temperature. The experimental conditions for these experiments are shown in Section 5.4.2 ....................... 172 Figure 5. 17 The effect of substrate temperatures on FWHM of diamond peak lines from Raman Spectra. Their Raman spectra are Shown in Figure 5.18 .................................................. 173 Figure 5.18 The effect of substrate temperature on Raman spectra. The experimental conditions are listed in Section 5.4.2 ............... 174 xviii Figure 5.19 Linear growth rate as a function of deposition time. The experimental conditions are described in Section 5.5.2. .......... 177 Figure 5.20 (a)-(f) Film morphology vs. deposition time. The experimental conditions are shown in Section 5.5.2. ....................... 179 F1 e 5. 21 X—ray diffraction s ectra of diamond films synthesized for (a) ,(b) 10 h, (c) 20 h, (d) 30 h. e) 55 h and (f) 100 h, using the experimen- tal conditions described in Section 5. 5. 2. ..................... 181 Figure 5.22 The XRD height of (220) relative to that of (1 1 1) as a function of deposition time in high pressure MCPR. .................... 182 Figure 5.23 FWHM of the diamond peak lines vs. deposition times. Their Raman spectra are shown in Figure 5.24 ...................... 183 Figure 5.24 The effect of deposition time on Raman spectrum. The experi- mental conditions are listed in Section 5.5.2. .................. 184 Figure 6.1 Cross sectional view of the early version of MPJR for high rate diamond film deposition on flat substrates[9] ................... 191 Figure 6.2 The cross sectional view of the basic version of MPJR for dia- mond film coating on two WC-6%Co round tools[1 10]. ........... 193 Figure 6.3 The cross sectional view of the first variation of MPJR for dia- mond film coating on six WC-6%Co round tools ................. 196 Figure 6.4 The cross sectional view of the first variation of MPJR for dia- mond film coating on eighteen and thirty six WC-6%Co round tools.. .................................................... 198 Figure 6.5 The cross sectional view of the thirdth variation of MPJR for diamond film coating on many carbon fibers. .................. 200 Figure 6.6 The cross sectional view of the final optimized configuration of MPJR which was used for diamond film coating on eighteen and thirty six WC-6%Co round tools .................................... 202 Figure 7.1 MPJR block diagram for diamond thin film deposition on the benchmark substrates .................................... 208 Figure 7 .2 Thickness measurements of 36 WC-6%Co round tools. . .215 Figure 7 .3 Film Uniformity of diamond films for thirty six WC—6%CO round tools. ........................................... 216 xix Figure 7 .4 SEM photo of the pre-deposited WC-6%Co round tool surface. .................................................... 218 Figure 7 .5 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 0.5 h ................................... 218 Figure 7 .6 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 1 h. ................................... 219 Figure 7.7 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 2 h. ................................... 219 Figure 7.8 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 3 h. ................................... 220 Figure 7.9 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 4 h. ................................... 220 Figure 7.10 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 5 h. ................................ 221 Figure 7.11 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 6 h. ................................ 221 Figure 7.12 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 8 h. ................................ 222 Figure 7. 13 Raman spectrum of the diamond film that was deposited on a WC-6%Co round tool surface ............................... 223 Figure 7. 14 The cross sectional view of the microwave plasma jet reactor for diamond thin film coating on carbon fibersll4) ............... 225 Figure 7. 15 The SEM picture of a diamond film coated carbon fiber. The experimental conditions are shown in Section 7 .3.2 .............. 228 ]Figure 7. 16 The SEM picture of the diamond film on carbon fibers. This figure is an enlarge of Figure 7.15 that contains more carbon fibers.. .................................................... 229 Figure 7. 17 Raman spectrum of the diamond film on the carbon fibers. .................................................... 230 XX CHAPTER ONE Introduction 1. 1 Introduction Microwave discharges were utilized in many of the first experimen- tal demonstrations of plasma assisted CVD deposition of diamond thin films[1]-[8]. These early experiments made use of a 3-4 cm diameter tubu- lar microwave reactor to create a 2.45 GHz excited discharge. Using CH4/ H2 feed gases polycrystalline diamond films were deposited on substrates that were placed adjacent to or into the microwave discharge. Since these early experiments many other types of plasma sources have also demon- strated plasma assisted diamond thin film deposition. Some of these plasma sources are (1) RF thermal discharges, (2) DC are discharges, (3)DC glow discharges, etc. These plasma sources, especially the higher pressure thermal-like discharges, have demonstrated linear deposition rates exceeding 100 um/h. In addition to a number of plasma assisted deposition processes diamond films can also be synthesized by hot filament and combustion reactors. However, microwave plasma assisted deposition still remains one of the most common laboratory methods used in scientific deposition investigations. Microwave discharges have a number of intrinsic advan- tages. They are a relatively simple technology, are easy to Operate and can deposit films on small substrate areas over a 10-70 Torr operating pres- sure range. While deposition rates are usually a modest 0.5-3 um/h, the film quality is high and repeatable. The modest deposition rates together with the electrodeless nature of the discharge produce films of good qual- ity and experiments that are repeatable from run to run. Thus microwave discharges have developed an excellent reputation as a plasma source in which to investigate the fundamental science of diamond deposition. The research described in this dissertation is concerned with improving microwave plasma assisted CVD diamond film deposition tech- nology. Of particular importance are the objectives of increasing the depo- sition rate and the deposition area of microwave plasma reactors. In addition, methods of depositing films on irregular shaped substrates and especially the deposition of films on objects with sharp edges and points are investigated. This thesis is part of a long-term research efiort at Mich- igan State University where microwave plasma technology was invented, developed and evaluated in specific applications. Thus this thesis builds upon earlier research by J. Zhang (see Ph.D Thesisl9]) and US. Patent Number 5,31 1.103(1071) where an improved 12.5 cm diameter plasma reactor, identified as microwave cavity plasma reactor(MCPR), was devel- oped and applied to diamond thin film deposition in the 20-70 Torr pres- sure regime. It also makes use of a new microwave reactor concept, identified as the microwave jet reactor (see J. Zhang’s Ph.D Thesis [9] and U.S Patent Number 5,571,577[108]) to deposit films on irregular shaped objects. This dissertation summarizes the experimental development and investigation that lead to a successful demonstration of a high pressure microwave cavity plasma reactor (HPMCPR) and a microwave plasma jet reactor (MPJR). Some of this work has been published in part in scientific journals[10],[1 1] and has been presented at international confer- ences[12],[13) and a patent disclosure has been submitted to MSU[14]. The research in this thesis utilizes the moderate-pressure MCPR reactor design[9] and modifies its design to extend its operation to higher pressures. The research activities involved the addition of a substrate cooling assembly to the MCPR and then exploring how to operate the reactor under the higher pressure (80-150 Torr) and higher power (2-4.5 kW) regime. A part of the experimental work in this thesis is also con- cerned with applying the microwave jet reactor concept[108] to diamond film deposition on irregular shaped objects. A series of experiments are described that demonstrate the usefulness of this technology to deposit films on round tools, and even very small, fragile objects such as carbon fibers. The research described in this thesis was carried by the author over a period of four years from 1993 to 1997 at Michigan State University under the guidance of Dr. Jes Asmussen. Research funding in this period was provided by Norton Diamond Film Corp., Michigan State Research Excellent Fund, and NSF MRSEC / Center for Sensor Materials 1.2 Research Objectives The research described in this thesis has three objectives. The first objective was to develop and test a high pressure MCPR for the synthesis of large area and uniform diamond films with high deposition rates (3-7 rim/h). The second objective was to experimentally characterize the high pressure microwave cavity plasma reactor for CVD diamond processes by evaluating the reactor performance over a wide range of reactor experi- mental variables. The third objective was to apply and optimize the micro- wave plasma jet reactor concept for the important application of depositing uniform and adherent diamond thin films on cemented carbide cutting tools and carbon fibers. 1.3 Dissertation Outline This dissertation is divided into three parts: (1) literature review (Chapter 2), (2) description, development, and characterization of the high pressure microwave plasma reactors (Chapter 3, 4, 5), and (3) description and application of the microwave plasma jet reactor (Chapter 6,7). Chapter 2 is a two-part review of the literature. The first section describes and compares the nonmicrowave plasma-assisted CVD reac- tors. A brief review of the “figure of merits” of the microwave plasma- assisted CVD reactors is presented. Chapter 3 has three sections. First section describes the experimental systems and the redesigned high pres- sure MCPR and second section describes the experimental procedures that were used in the experiments reported in this thesis research. The last section describes the methods of the measurement of the output vari- ables. Chapter 4 presents an overview of the reactor performance (Y 1) of the high-pressure MCPR vs. five independent input variables, i.e., pres- sure, absorbed power, flow rate, substrate temperature, gas chemistry, and deposition time. Chapter 5 describes the deposited film properties (Y3) versus the multidimensional input variable space. The experimental details and results of CVD diamond deposition films vs. each single input variable are highlighted. The first part of Chapter 6 describes the micro- wave plasma jet reactor. The state of the art performance of this micro- wave plasma jet reactor is presented and the method of measurement of the output variables (Y) for diamond film deposition on cemented carbide cutting tools and carbon fibers is reported in third section of Chapter 6. Chapter 7 presents the growth and characterization of microwave plasma CVD diamond films on WC-Co cutting tools and carbon fibers. Experi- mental details, results and analysis of the experimental evaluation of CVD diamond films in terms of tool packing capacity, film uniformity, film growth rate, film quality, film morphology are presented. Chapter 8 sum- marizes the research that was investigated in this thesis. CHAPTER 2 Literature Review of Diamond Film Deposition Reactors 2.1 Introduction This chapter reviews a number of diamond film chemical vapor dep- osition (CVD) technologies. These technologies include hot filament CVD, direct current (DC) CVD, combustion flame, radio frequency (RF) and microwave plasma assisted CVD. The experimental performance of these difi'erent reactors is compared by computing a number of reactor perfor- mance measures. These reactor output measures are (1) film growth rate, (2) film weight gain, (3) specific yield, (4) gas efficiency, and (5) carbon conversion efiiciency. Since the major focus of this dissertation is con- cerned with microwave plasma assisted diamond film CVD, microwave plasma reactors are reviewed and discussed in detail in Section 2.5. 2.2 A General Description of the Chemical Vapor Deposition of Dia- mond Films A general process for diamond film growth by chemical vapor depo- sition is demonstrated by the diagram shown in Figure 2.1[3]. The gas input is a mixture of hydrogen (H2) and hydrocarbon (CH4, C2H2, etc.) gases. The energy input is supplied by electrical (DC, Radio Frequency, or Microwave) energy and chemical energy. With sufficient energr input, the input gas mixtures are dissociated and chemically react with each other. A mixture of activated hydrocarbon and atomic hydrogen species are gen- erated and diffuse onto substrate surface. During the surface reconstruc- tion, clusters of diamond, amorphous carbon (a-C), amorphous hydrocarbon (a-C:H), and graphite are formed. As shown in Figure 2.1, the atomic hydrogen facilitates the diamond formation with only Sp3 bonds and suppresses the graphitic carbon clusters. The role of atomic hydrogen was first noted by Angus et aL[15], who reported on the prefer- ential etching of graphite compared to diamond by atomic hydrogen. Hydrocarbon Gas Hydrogen Gas newsman Dissociation Activated Hydrocarbon Species Atomic Hydrogen Diffusion Diffusion Reaction on Substrate Surface Etching Graphite, a-C, a—C:H Diamond Figure 2.1 A general process for diamond film growth by chemical vapor deposition[3]. 2.3 The Multivariable Diamond Film Deposition Reactor 2.3. 1 Experimental Variables of a Diamond Film Deposition Reac- tor The many experimental variables of a typical diamond film deposi- tion reactor are identified in Figure 2.2. As shown, controllable input vari- ables (U 1) are defined as the variables which can be independently controlled during an experiment such as input power, gas chemistry, gas flow rate, deposition pressure, etc. Reactor design variables (U2) are iden- tified as the variables which can be prechosen by the reactor designer such as reactor configuration and reactor wall materials, etc. Deposition process variables (U3) include the variables concerned with substrate pro- cess procedures and substrate preparation such as substrate biasing/ cooling/ heating, substrate seeding procedure, start-up and shut-down procedures, deposition time, etc. Internal variables are concerned with the variables related to internal discharge states such as discharge vol- ume, species densities and temperatures, electron density and tempera- ture, substrate temperature, etc. Output variables are identified as output performance or “figures of merit” of a diamond film deposition reactor such as linear growth rate, total growth rate, specific yield, gas flow efficiency, carbon conversion efficiency, etc., or film characteristics such as film uniformity, film Raman Spectrum, film morphology, etc. 10 neuomoa 5560ro 55 98:86 a .«o mofimguqsz Nd mama 85885 $8.5;w use 9335 35 Ba 3222 SEES o 2.683 58mm 39% am a 9522.5 a m a .08 . 238.45 85880 2mm 26: am 0 Hr 92.5 mm _ . egofioafi ”a = 33$...» 23.228 5.33%; «some .3...» sex: at ~> Engages 2.: 2m . swomwfiamocgm m Mi SBESEoH 2825 3.5.8.5 SE a H_I we: 8828c . . . segues. 8.85 > eczemacmfl 80 85.980 598 o 6339.; 85:5 8 .285 ahwewmgfiwmowmmwfi 59.5 #8325, See... m c _ h . . Sum 5390 88: m > 85:55.. .638: .808: 528.3 2.: Basso Eeazassoasfim 22.33% :89: :80 .208 a use a; Seesaw". 859m 526m 3 on .r a 3.83%.; Sea 358.. 2.3.2 11 “Figures of Merit” of Reactor Performance In order to evaluate the performance of individual diamond film deposition reactors and compare the performance between difierent reac- tors, it is necessary to define the performance measures for a diamond film deposition reactor. These “figures of merit” may describe the effi- ciency of the deposition process or other important outputs of the deposi- tion process such as film deposition rate. These figures of performance or merit can then be used to compare the performance of different deposi- tion reactor concepts. The “figures of merit” that are employed in this dis- sertation are defined below: (1) (2) (3) (4) W (um/ h) is defined as the thickness increase of the diamond film (um) divided by deposition time (h). W (mg/h) is defined as the weight increase of the substrate (mg) divided by deposition time (h). Most papers in the open literature only report linear growth rate. Therefore if weight increase is not given in the reference then the total growth rate is calculated from the thickness increase by: Weight Increase=(Thickness Increase)x(Deposition Area)x(Dia- mond Density of 3.51 g/cm3). Wm (kW-h/g) is defined as the power input (kW) per diamond film total growth rate (g/h)[9]. W (mg/liter) is defined as the total growth rate (mg/ min) divided by total flow rate (liter/ min). The total flow 12 rate is defined as the sum of all input gas flow rates[9]. (5) WSW is defined as the percentage of carbon atoms in the input gases which are converted into dia- mond[9],[16],[l7]. 2.4 Diamond Film CVD Reactor Technologies Diamond film CVD reactors can be arranged into five maj or catego- ries: (1) Hot filament CVD (HFCVD) reactors: (2) Direct current(DC) plasma-assisted CVD reactors; (3) Combustion flame; (4) RF plasma-assisted CVD reactors: (5) Microwave plasma-assisted CVD reactors. The state-of-art of (1)-(4) is briefly described in Section 2.4.1-2.4.5 below while in Section 2.5 microwave plasma assisted CVD reactors are reviewed in greater detail. 2.4.1 Notes on Literature Review The review presented in the following sections is based on the experimental data that were published in the open literature. The input variables of a diamond film deposition reactor such as input power, gas chemistry, deposition pressure, reactor configuration, substrate material, etc., are directly cited from the references. However, except for the sub- 13 strate temperature, the internal variables of a diamond film deposition reactor are usually not described in the literature. The “figures of merit” of a diamond film deposition reactor examined in the following‘sections are estimated or calculated fiom the best data available in the references. Some factors that can limit the accuracy of the “figures of merit” presented in this review are discussed below: (1) In this review, deposition area is assumed to be the area of the entire substrate surface. It is assumed that the diamond film is deposited over uniformly the entire substrate surface. This assumption may result in overestimating some of the reactor “figures of merit”. (2) Linear growth rate may be measured differently by different research groups, resulting in the differences in the accuracy of “figures of merit” from on report to another. Some groups measure the film thick- ness from the SEM photos of the cross-sectional area of a diamond film while other groups use weight increase of the diamond film along with the deposition area and diamond density to calculate the film thickness. For both methods, the linear growth rate was obtained by dividing the thick- ness increase with the deposition time. If maximum linear growth rates are used, then the total growth rate expressed in mg/ h is an overestima- tion of the growth rate unless the diamond film is uniformly coated over the entire deposition area. (3) Since the method of determining film growth rate is different from one research group to another, the error bars for the “figures of l4 merit” may difier considerably from one reactor to another. 2.4.2 Hot Filament CVD (HFCVD) Reactors 2.4.2.1 Conventional HFCVD Reactor Figure 2.3 schematically shows the conventional HFCVD reactor developed by Matsumoto et aL[18],[19] in 1982. After the reactor was evacuated to a low pressure by a vacuum pump, a gas mixture of meth- ane and hydrogen was introduced from the top of the silica glass tube at a total flow rate of 4-200 sccm. A tungsten filament of 30 mm long, 0.15 mm in diameter[19] was placed several millimeters above the substrate. Once the reactor reached the operating pressure of 10-100 Torr, the fila- ment was then heated by an electric current to a temperature of 2000°C. The gas mixture was thermally dissociated as it passed over the hot fila- ment. A substrate (Si, Mo, natural diamond, etc.) was put on a silica holder and its temperature was monitored by a Pt-PtRh thermocouple set just below the silica holder. Typical experimental variables that were reported or calculated in recently developed HFCVD reactors[20] are: (A) Input Variables(U): (1) Controllable input variables (U1): a) Input electrical power was not reported, 15 Feed Gas Furance ! Silica Tube // Tungsten Filament / / Substrate W// £ Silica Holder Thermocouple I I To Pump Figure 2.3 Schematic diagram of a conventional HFCVD reactor [181,[19]. 16 b) Gas mixture: 0.067%-0.8%CH4/H2(CH4 =0.1-1.2 sccm, H2 =150 sccm), c) Total flow rate~l50 sccm. d) Deposition pressure=25 Torr (3.33 kPa), (2) Reactor design variables(Uz): a) Type of Power source: DC power, b) Reactor configuration is shown in Figure 2.3, c) Reactor size: The filament was 10 mm long, 0.13 mm in diameter, (3) Deposition process variables(Uz): a) Deposition time=1 h, b) Substrate seeding was not reported, c) Substrate material and size: l.5x1.5 mm2 silicon, d) Substrate was not biased, cooled or electrically heated, (B) Internal variables(X): (1) Substrate temperature=700-1060°C(973-1333°K), (C) Figures of merit calculated for this review: (1) Linear growth rate=1-4 um/h, (2) Deposition area=0.0225 cmz, (3) Total growth rate=0.008-0.03 mg/h, (4) Specific yield is not available, (5) Gas flow efficiency=0.0008-0.003 mg/ liter, (6) Carbon conversion efiiciency~0.08%. 17 2.4.2.2 Electron-assisted HFCVD Reactor An electron-assisted HFCVD reactor modified from the conventional HFCVD reactor has been used for diamond film growth. As an example, Figure 2.4 schematically displays the configuration of an electron- assisted HFCVD reactor[21],[22]. An AC current of 20A and 60V was applied and heated the filament to a temperature of 2000°C. In addition, a DC bias of approximately 90V was applied between the filament (nega- tive) and the substrate (grounded, not Shown in Figure 2.4). After the reactor was evacuated, a total flow rate of 50-100 sccm of a mixture of methane and hydrogen was supplied from the gas inlet and was forced to flow over the filament and substrate by the quartz tube shown in Figure 2.5. The reactor pressure was maintained at 30 Torr with a throttle valve. A tungsten filament of 20 mm long, 1.2 mm in diameter was supported approximately 1cm above the substrate. A substrate (Si, Mo, SiC, WC, etc.) was placed on a graphite holder, which in turn was placed on a water cooling stage. The substrate temperature was measured by a thermocou- ple attached to the back of the substrate. Typical experimental variables that were reported or calculated from Ref. [22] are: (A) Input Variables(U): (1) Controllable input variables(U1): a) Input power level: 1.2 kW, b) Gas mixture: 1-8% CH4/H2 concentration, 18 A} _ .— 0ptica| Pyrometer 20A, 60V AC Power A 1 :34: 90v DC Bias / / e—‘J; Gas Inlet ; i J ‘ Quartz Tube Filament > l—mm fl ::(:l Pressure Sensor Substrate Graphite Water Cooling Holder Stage Chamber '11 Water ln/0ut 1rL Thermocouple Vacuum Pump Figure 2.4 Schematic diagram of an Electron-assisted HFCVD reac- tor[22]. 19 c) Total flow rate=100 sccm. d) Deposition pressure=30 Torr (4 kPa), (2) Reactor design variables(Uz): a) Type of power source: AC power, b) Reactor configuration is shown in Figure 2.4, c) Reactor size: The filament was 20 mm long and 1.2 mm in diameter, (3) Deposition process variables(Uz): a) Deposition time=4~l4 h, b) Substrate seeding procedure was not reported, c) Substrate material and size: 2xlcm2 silicon, d) Substrate was positively biased by 90V (with respect to fil- ament) and was water cooled, (B) Internal variables(X): (1) Substrate temperature=750°C, (C) Figures of merit calculated for this review: (1) Linear growth rate=0.5-3 um/ h, (2) Deposition area=2 cm2, (3) Total growth rate=0.35-2.1 mg/h, (4) Specific yield=571-3428 kW-h / g, (5) Gas flow efficiency=0.058-0.35 mg/ liter, (6) Carbon conversion efficiency=0.54-l.31%. 20 2.4.3 Direct Current (DC) Plasma CVD Reactors 2.4.3.1 Conventional DC Plasma CVD Reactor Figure 2.5 schematically displays a conventional DC plasma CVD reactor[23],[24]. After the reactor was evacuated to 10'4 Pa(~7.5x10’4 mTorr), a gas mixture of CH4 and H2 was introduced at a total flow rate of 20-100 sccm. The input gas was distributed by a the water-cooled, grilled cathode. The reactor pressure was held at 200 Torr by a throttle valve. DC power was supplied to create a DC discharge between water-cooled anode and cathode when the chamber reached the operating pressure of 200 Torr. A substrate (Si, Mo, W, A1203, SiC, etc.) was mounted on the anode. This upward flow configuration allowed the impingement of dissociated species upon the substrate, resulting in the formation of diamond films. The temperature of the substrate during deposition was measured with an optical pyrometerl24]. Typical experimental variables that were obtained or calculated in this DC plasma CVD reactor are[24]: (A) Input Variables(U): (1) Controllable input variables(U1): a) Input power=3. 14 kW, b) Gas mixture: 2%CH4/H2, c) Total flow rate=100 sccm. d Deposition pressure=200 Torr (26.6 kPa), 21 Water Anode Substrate DC Power Supply Pressure water \ l / 7— Controller l l IJV Gas Inlet To Pump Figure 2.5 Schematic diagram of the DC plasma CVD reactor[23]. 22 (2) Reactor design variables(Uz): a) Type of power source: DC power, b) Reactor configuration is shown in Figure 2.6, c) Reactor size: Reactor consisted of a 1 cm diameter anode and a 2 cm diameter cathode, (3) Deposition process variables(Ug): a) Deposition time=30 min, b) Substrate seeding procedure was not reported, c) Substrate material and size: 5x5 mm2 Si, Mo, W, etc. (i) Substrate was positively biased with 1kV (with cathode grounded) and was water cooled, (B) Internal variables(X): (1) Substrate temperature=850°C, (C) Figures of merit calculated for this review: (1) Linear growth rate=20 um/h, (2) Deposition area=0.25 cm2, (3) Total growth rate=1.76 mg/ h, (4) Specific yield=1784 kW-h / g, (5) Gas flow efiiciency=0.29 mg/ liter, (6) Carbon conversion efiiciency=2.73%. 23 2.4.3.2 Enclosed DC Arc Jet CVD Reactor A DC arc jet CVD reactor similar to the conventional DC plasma CVD reactors[24],[25] is illustrated in Figure 2.6[26]. During the diamond deposition, only Ar was passed through the plasma gun electrode gap, creating a high velocity plasma. The CH4 and H2 were introduced into the vacuum chamber through separate feed lines at a total flow rate of approximately 5.5 liter/ min. The H2 gas was injected 1 cm downstream of the electrode gap through opposing side ports integral to the plasma gun, while CH4 gas injected 3.5 cm downstream from the H2 ports through a stainless steel tube mounted onto the substrate holder. The reactor pres- sure was kept between 50-400 Torr by a throttle valve. The water-cooled molybdenum substrate with the size of 25mmx38mm was positioned 2.5 cm from the electrode gap. This downstream configuration allowed the impingement of dissociated species upon the substrate, resulting in the formation of diamond films. Typical experimental variables that were reported or calculated in this DC plasma jet CVD reactor are[26]: (A) Input Variables(U): (1) Controllable input variables(U1): a) Input power=2.8-3.0 kW(20V, 140-150A DC Power), b) Gas mixture: 2.7%-15%CH4/H2(Ar=13.7 liter/min, CH4: 150-700 sccm, H2=5.4 liter/min), c) Total flow rate=20 liter/ min, 24 Ar (+Electrodes DC Arc H2 I"l2 Plasma Jet Mo Substrate Protective Plate I Water-Cooled Mount [ l—l__ ] H _ _::l——> Cooling Water Out To Vacuums Pump l Cooling Water In Figure 2.6 Schematic diagram of a DC plasma jet CVD reactor[26]. 25 d) Deposition pressure=55-60 Torr (~18 kPa), (2) Reactor design variables(Uz): a) Type of power source: DC power, b) Reactor configuration is Shown in Figure 2.6, c) Reactor size: The size of anode and cathode were not reported, (3) Deposition process variables(Uz): a) Deposition time was not reported. b) Substrate seeding procedure was not reported, c) Substrate material and size: 2.5x3.8 cm2 silicon, d) Substrate was electrically insulated and water cooled. (B) Internal variables(X): (1) Substrate temperature was not reported, (C) Figures of merit calculated for this review: (1) Linear growth rate=0.6-3.6 urn/h, (2) Deposition area=9.5 cm2, (3) Total growth rate=2-12 mg/h, (4) Specific yield=250- 1500 kW-h / g, (5) Gas flow efficiency=0.01-0.0017 mg/ liter, (6) Carbon conversion efficiency=0.06-0. 12%. Another DC plasma jet CVD reactor was operated at atmospheric Pressure with two DC power supplies. The schematic of this reactor con- 26 figuration can be seen in Figure 2.7[27]. A secondary discharge was used to enhance the film growth rate using a DC substrate bias. As Shown, a mixture of CH4 and H2 gases with flow rates of approximately 20 liter/ min were fed from six injection ports located at the copper anode. In the meantime, Ar gas at a flow rate of 450 liter/ min was fed into the anode- to-cathode spacing. The arcjet power supply provided the DC power to the tungsten cathode and copper anode, resulting in the formation of a DC are around the converging nozzle. When input mixed gases passed down- stream through the DC arc and the converging nozzle, the gas mixtures were dissociated and created a downstream plasma jet. These dissociated species impinged upon the substrate, resulting in the formation of dia- mond film on its surface. A substrate consisting of 1/2" (1.27 cm) diame- ter molybdenum rod was threaded in a water-cooled copper holder as shown in Figure 2.7. This copper holder was DC biased by a biasing power supply and electrically insulated with a boron nitride (BN) insula— tor. Arciet Power Supply Bias'ng Power Supply 27 .L —_ d- — —- - Converging Nozzle Plasma Jet Thoriated Tungsten Cathode r Copper Anode H2, 80d CH4 (6 injection ports) Molybdenum Rod Copper Holder BN Insulator Water Cooling Figure 2.7 Schematic illustration of an atmospheric pressure DC are jet CVD reactor and its substrate biasing system[27]. 28 Typical experimental variables that were reported or calculated in this DC plasma jet CVD reactor are[27): (A) Input Variables(U): (1) Controllable input variables(U1): a) Input power: 100 kW, b) Gas mixture=2.5%CH4/H2(CH4=0.467 liter/ min, H2 = 19.125 liter/ min, Ar=450 liter/ min), c) Total flow rate~470 liter/ min, (1) Deposition pressure: atmospheric pressure, (2) Reactor design variables(ng a) Type of power source: TWO DC power supplies b) Reactor configuration is shown in Figure 2.7, c) Reactor size: The size of anode and cathode were not reported, (3) Deposition process variables(U-j: a) Deposition time=1 h, b) Substrate seeding procedure was not reported, c) Substrate material and size=1/2” diameter Mo rod, (1) Substrate was positively bias up to 200V and was also water cooled, (B) Internal variables(X): (1) Substrate temperature: 1 1 15°C(1388°K), (C) Figures of merit calculated for this review: 29 (1) Linear growth rate=10-30 um/h, (2) Deposition area=1.27 cmz, (3) Weight gain=2.67-13.37 mg/ h, (4) Specific yield=7479 kW—h / g, (5) Gas flow efiiciency=9.47x10'5-4.74x10'4 mg/ liter, (6) Carbon conversion efficiency=0.018-0.089%. 2.4.4 Combustion Flames 2.4.4.1 Atmospheric Pressure Combustion Flame The schematic drawing of a conventional combustion flame dia- mond film deposition reactor, which was operated at atmospheric pres- sure, is shown in Figure 2.8[28],[29],[30]. As shown, this combustion flame consists of torch, a torch nozzle, shroud, and the substrate assem- bly. A 0.95-1.27 cm diameter molybdenum rod was used as the substrate and was placed on a water-cooled block. A total flow rate of 2-6 liter/ min of a mixture of C2H2 (acetylene) and 02 was introduced from the top of the torch and flowed downstream to the torch nozzle and the shroud gas (N2 or air) was fed downstream between torch and shroud. A flame of approximately 3000°C[27][29] was ignited by an electrical spark. The gas mixture of C2H2 and 02 was dissociated by passing through the flame. Diamond films were then deposited on the substrate. The substrate tem- perature in this reactor was measured with an optical pyrometer. 30 C2H2'02 MIXIUTB Shroud Gas (N2 or air) Shroud Gas (N2 or air) 1 / iii Torch Nozzle a — Premixed Flame Torch Atmosphere Flame Feather Water Cooling Block Substrate Figure 2.8 Schematic diagram of an atmospheric-pressure oxygen- acetylene combustion flamel29]. 31 Typical experimental variables that were used in this combustion flame are[29]: (A)Input Variables(U): (l) Controllable input variables(U1): a) The flame power was not reported, b)Gas mixture: C2H2 (0.643 liter/min), 02 (0.607 liter/min) and N2=1.5 liter/min, c) Total flow rate=2.7 5 liter/min, d) Deposition pressure: atmospheric pressure, (2) Reactor design variables(Uz): a) Type of power source: combustion flame, b) Reactor configuration is shown in Figure 2.8, c) Reactor size: the torch size was not reported (3) Deposition process variables(Uz): a) Deposition time=25 min, b) Substrate was seeded by being polished with No.600 SiC sandpaper and 1 um diamond paste, c) Substrate material and size=3/8"(0.95cm) diameter Mo, d) Substrate was water cooled, no bias was applied. (B) Internal variables(X): (1) Substrate temperature=1050°C(1323°K), (C) Figures of merit calculated for this review: (1) Linear growth rate=10-15 um/h, 32 (2) Deposition area=0.7 1 cm2, (3) Weight gain=2.49-3.74 mg/ h. (4) Specific yield was not available, (5) Gas flow efficiency=0.033-0.050 mg/ liter, (6) Carbon conversion efficiency=0.006-0.009%. 2.4.4.2 Enclosed Flat Combustion Flame A low-pressure flat flame was developed by Goodwinl31][32] and Cappelli[33][34] to deposit diamond film at the pressure ranging from 40 to 200 Torr. Figure 2.9 illustrates the reactor described in Ref.[33]. As shown, a multi-nozzle burner was supported by a T-stage. After the reac- tor was evacuated, a gas mixing chamber injected a mixture of 02 and fuel (C2H2[19,20], CH4[21], etc.) at flow rates between 9-21 liter/min through the multiple nozzles located on the top of burner and onto the substrate. The flame was ignited with an electric spark by an ignitor at proper pressure and the upward flow gas mixture then was dissociated in the flame. The dissociated species impinged on the substrate where the diamond films were deposited. The substrate used in this reactor was a threaded molybdenum rod of diameter ranging from 1.6 cm[33] to 5 cm[32], depending on the size of the burner. The substrate temperature was controlled by adjusting the amount of molybdenum that was threaded into the water-cooled block. The substrate temperature was measured by a two-color infrared pyrometer. 33 Water In Water Out I nitor F l L— Power supply ——gi:l— i_M°__JFIamt-3‘T 2:181:36; Water In I Multinozzle Water Out 24 3 Array C Burner , 02 ——J Mixing '7 ——|_] Chamber Reactor Fuel Figure 2.9 Schematic of an enclosed flat flamel32]. 34 Typical experimental variables were given or calculated from the data supplied in Ref. [32] are: (A)Input Variables(U): (1) Controllable input variables(U 1): a) Type of power source: combustion flame b) Flame power was not reported, 0) Gas mixture: - 0.6 liter/min-cm2 of C2H2,i.e., 1.884 liter/min in a 2 cm diameter burner, and 7.536 liter/min in a 4 cm diameter burner, - 1.3 liter/min--cm2 of 02,i.e., 4.082 liter/min in a 2 cm diameter burner, and 16.328 liter/ min in a 4 cm diameter burner, ii) Total flow rate=9.42-20.41 liter/ min, c) Deposition pressure=180 Torr (~24 kPa), (2) Reactor design vanables(U2): a) Reactor configuration is shown in Figure 2.9, b) Reactor size: The diameter of the burner shown in Figure 2.10 was from 2 cm to 4 cm[31],[32], (3) Deposition process variables(Uz): a) Deposition time=25 h, b) Substrate seeding information was not reported, c) Substrate was a 5 cm diameter molybdenum rod, 35 d) Substrate was water cooled, no bias was applied. (B) Internal variables(X): (1) Substrate temperature=830°C(l 103°K), (C) Figures of merit calculated for this review: (1) Linear growth rate=1 um/h, (2) Deposition area=19.6 cm2, (3) Weight gain=6.88 mg/h, (4) Specific yield is not available, (5) Gas flow efficiency=0.012-0.056 mg/ liter, (6) Carbon conversion efficiency=0.0014-0.0057%. 2.4.5 RF Plasma CVD Reactors 2.4.5. 1 Conventional RF Plasma CVD Reactor A schematic view of a conventional RF plasma CVD reactor oper- ated at low pressure (3.75-22.5 Torr) is displayed in Figure 2.10[35]. As shown in Figure 2.10, the center section of quartz tube was coaxially sur- rounded by the working coil. The substrate was placed on a silica boat in the middle of quartz tube and experiments were performed without any external biasing or heating. After the reactor was evacuated, a mixture of CH4/H2 was fed into the quartz tube. One half to one kW of RF power was supplied by a 13.56 MHz generator. A RF plasma discharge was ignited at the center of quartz tube when the pressure was between 3.75 and 22.5 36 Torr. This plasma discharge dissociated the input gas mixtures and the radical species reacted on the substrate. Silicon, molybdenum, and silica glass plates were used as the substrate materials and the temperatures were monitored by an optical pyrometer during each experimental run. Two example deposition results are[35]: (l) a 5 um film was depos- ited under 7 .5 Torr (1 kPa) with 700 W RF power at a substrate tempera- ture of 800°C, and (2) a 20 um film was deposited under 22.5 (3 kPa) with 1 kW RF power at substrate temperature of 940°C. Since the deposition time and the substrate size were not reported in Ref.[35], the “figures of merit” are not available. Substrate Holder Working Coil 0000000 F eed Gas Substrate To Pressure Controller To Pump / o o o o o Quartz Tube I Q’W | Water 0ut ater n I! ll '1 13.56 MHz RF Generator Figure 2.10 A schematic diagram of a conventional RF plasma CVD reactor[35). 37 2.4.5.2 RF Thermal Plasma CVD Torch The schematic drawing of a RF thermal plasma diamond filrn depo- sition reactor is shown in Figure 2.1 1[36],[37). As shown, the RF thermal plasma torch consists of a torch head, coaxial double quartz sleeves, and an RF working coil. The torch was mounted on the flange of a reactor chamber, where the water cooled substrate holder was positioned. After the chamber was evacuated, sheath gases (mixture of Ar, H2 and CH4) and plasma gas (Ar only) were introduced from the torch head. A 3.4-4 MHz RF power supply (not shown in Figure 2.1 l) was used to sustain the thermal plasma at deposition pressure of an atmosphere[36] or subatmo- spherel37]. A molybdenum substrate of 10 cm diameter was placed within or at the end of the thermal plasma on a water-cooled substrate holder. Diamond film was deposited on the substrate as the dissociated gas species reacted on its surface. 38 Sheath Gas Plasma Gas Cooling Water Outlet :i i: T h H d Chamber Flange Water out Substrate Nozzle T0 Pump <—: ‘ —__—>_ T0 EXhQUSI I —_.Water Out Water In Figure 2.14 Schematic drawing of a microwave plasma jet torchl4l]. 49 (2) Reactor design variables(U,): a) Type of power source=a 2.45 GHz microwave power, b) Reactor configuration is shown in Figure 2.14, c) Reactor size: A 57 .2 mm diameter outer conductor and a 20 mm diameter center conductor with a 22 mm jet nozzle (3) Deposition process variables(Uz): a) Deposition time=35 min, b) Substrates were mechanical blasted for 2 h with SiC pow- der, c) Substrate size=25x10 mm2, d) substrate was water cooled only, (B) Internal variables(X): (1) Substrate temperature=887-927°C(l 160- 1200°K), (C) Figures of merit calculated for this review: (1) Linear growth rate=6-12 um/h, (2) Deposition area=2.5 cm2, (3) Weight gain=5.27-10.53 mg/h. (4) Specific yield=360-800 kW-h / g, (5) Gas flow efficiency=0.002-0.004 mg/Iiter, (6) Carbon conversion efiiciency=0.018-0. 109%. 50 2.5.4 Bell-jar Microwave Plasma-assisted CVD Reactors 2.5.4. 1 Astex Bell-jar Microwave Plasma-assisted Reactor The bell-jar microwave PACVD system, manufactured by Astex Inc., is schematically shown in Figure 2. 15(431-[46]. After the reactor was evac- uated by the vacuum pump, a mixture of methane, hydrogen and oxygen was introduced from the gas inlet and guided into the bell jar by a cylin- drical fused silica tube. The gas mixtures were dissociated by the micro- wave energy which was coupled into the cylindrical microwave cavity through the rectangular waveguide and coaxial antenna. The substrate, which can be silicon, fused silica, MgO, etc., was placed on a graphite holder and this holder was placed on the substrate cooling stage. A ball- shapedl44) plasma was created inside the quartz bell jar and the dissoci- ated species reacted on the substrate surface, producing the diamond film over its surface. Detailed experimental variables that were reported or calculated from Ref. [45] are: (1) Controllable input variables(U1): a) Input power=1.5 kW, b) Gas mixture: 0.5% CH4/H2, c) Total flow rate=200-600 sccm, d) Deposition pressure=40-70 Torr, (2) Reactor design variables(Uz): 51 Waveguide Magnetron : i bmema 6:52;?3? I l—iCylidrical Microwave Cavity _Quartz Bell Jar Substrate Ball Shape Plasma Graohite Substrate Holder Cylindrical Fused Silica Tube Gas Inlet J- ". Gas Inlet Water Cooling Stage - =--—.To Vacuum Pump Water Out Water In If Thermocouple Figure 2. 15 Schematic drawing of the Astex bell jar microwave PACVD reactor[43l-[46]. (3) 52 a) Type of power source: a 2.45 GHz microwave power, b) Reactor configuration is shown in Figure 2. 15, c) Reactor size: A 4" diameter(10.16 cm) quartz bell jar, Deposition process vanables(U2): a) Deposition time was not reported, b) Substrates were polished by 0.5 um diamond powder, 0) Substrate size=2 cm diameter, d) substrate was water cooled only, (B) Internal variables(X): (1) Substrate temperature=850- 1 030°C( 1 123- l 303°K), (C) Figures of merit calculated for this review: (1) (2) (3) (4) (5) (6) Linear growth rate=3.5 um/h, Deposition area=12.5 cm2, Weight gain=15.4 mg/h, Specific yield=97 kW-h / g, Gas flow efliciency=0.42-l.28 mg/ liter, Carbon conversion efficiency=16-48%. 53 2.5.4.2 UC-Berkeley Bell-jar Plasma-assisted CVD Reactor A schematic diagram of a UC-Berkeley bell-jar MPACVD system is shown in Figure 2.16[471-[50]. After the reactor chamber was evacuated to 30 mTorr[49], a mixture of CH4 and H2 gases with flow rates up to 100- 300 sccm was fed into a 10.2 cm diameter quartz bell jar[48] via an annu- lar gas feed system(not shown in Figure 2.16) in the circular metal base- plate. The mixture was then guided to the substrate by a short quartz tube, and was pumped out at the baseplate near the substrate holder. Microwave power entered the reaction chamber from the top and dissoci- ated the gas mixtures. The substrate, which was a 25 mm diameter sili- con wafer[49], was supported by a quartz sample holder at the position near the plasma ball. The dissociated species reacted on the substrate and formed the diamond film on its surface. The substrate temperature was measured from the back of the substrate by an infrared pyrometer. Detailed experimental variables that were reported from Ref.[50] (1) Controllable input variables(U1): a) Input power=560 W, b) Gas mixture: 0.1% Vol. CH4, c) Total flow rate: approximately 300 sccm, d) Deposition pressure=80 Torr, (2) Reactor design variables(Uz): a) Type of power source: a 2.45 GHz microwave power sup- ply. b) Reactor configuration is shown in Figure 2.16, Quartz Bell Jar Graphite Retaining Ring 54 2.45 GHz Microwave Power I I\ Pyramidal Waveguide Secondary Waveguide Plasma Ball Quartz Sample Holder Quartz Tube Substrate Gas In&0ut i.— Metal Baseplate \ >T To Vacuum Pump L 55. To Optical Pyrometer Figure 2.16 Schematic of the 4” UC-Berkeley bell jar PACVD system(47l-[50]. (3) 55 c) Reactor size: A 10.2 cm diameter(4") quartz bell jar, Deposition process variables(Uz): a) Deposition time =24 h, b) Substrates were scratched with 1 pm diamond paste, 0) Substrate size=25 mm diameter, (B) Internal variables(X): (l) Substrate temperature=800-900°C( 107 3- 1 1 73°K), (C) Figures of merit: (1) (2) (3) (4) (5) (6) Linear growth rate=0.02 m/ h, Deposition area=4.91 cm2, Weight gain=0.034 mg/ h, Specific yield=16470 kW-h / g, Gas flow efficiency=0. 1 1 mg/ liter, Carbon conversion efficiency=0.35%. 56 2.5.4.3 MSU Microwave Cavity Plasma Reactor (MCPR) The schematic drawing of a microwave cavity plasma reactor (MCPR) developed at MSU in 1986-1991 is shown in Figure 2.17[9]. The microwave discharge was produced inside a quartz dome located at one end of a microwave cavity. After the chamber was evacuated to 1 mTorr, the reactive gases were introduced radially from the bottom of the cavity and flowed upward into and through the plasma. The gases then flowed out through a flow pattern regulator. The microwave emery was transmit- ted from a 2.45 GHz microwave generator through a rectangular waveguide into a 7" cylindrical cavity via an end-feed probe. A 12 cm diameter hemisphere shaped plasma was created and stabilized by adjusting the length of the conduction probe and sliding short, i.e, I.p and Ls. A silicon substrate was placed on the flow pattern regulator. Thus the discharge was in direct contact with the substrate and also directly heated the substrate. Detailed experimental variables that were reported or calculated from Ref.[51] are: (1) Controllable input variables(U1): a) Input power=2.4 kW, b) Gas mixture: 1.5%CH4/H2 (CH4/H2=6/ 400 sccm), c) Total flow rate=406 sccm, d) Deposition pressure=37 Torr (5 kPa), 57 Rectangular Waveguide . 1 Magnetron l_—_ Microwave Generator Finger Stock > _. . Cylidricel Microwave Cavity Conducting Probe :1: Sliding Short Tuning ' ‘ Finger Stock Lo LP Substrate Quartz Bell Jar Hemisphere shape Plasma Gas Inlet IL Gas (lilo; Pattern Regulator Resonant Breaker - ~- Quartz Tube To Vacuum Pump Figure 2.17 Schematic diagram of the moderate-pressure micro- wave cavity plasma reactor (MCPR)[9],[51]. 58 (2) Reactor design variables(Uz): a) Type of power source: a 2.45 GHz microwave power sup- ply. b) Reactor configuration is shown in Figure 2.17, c) Reactor size: A 5" diameter(12.7 cm) quartz bell jar, (3) Deposition process variables(Ug): a) Deposition time=8 h, b) Substrates were seeded by 0.1 pm diamond powder/pho- toresist solution(detailed see Ref.[51]) c) Substrate size=3” diameter (7.62 cm) diameter, d) substrate was thermally isolated and no substrate bias was applied, (B) Internal variables(X): (1) Substrate temperature=1000°C(1273°K), (C) Figures of merit calculated for this review: (1) Linear growth rate=0.66 um/ h, (2) Deposition area=45.60 cm2, (3) Weight gain: 10.56 mg/ h, (4) Specific yield=227 kW-h/g, (5) Gas flow efficiency=0.43 mg/Iiter, (6) Carbon conversion efficiency=2. 19%. 59 2.5.5 Surface-wave Microwave PACVD reactor A schematic diagram of a surface-wave microwave PACVD reactor is displayed in Figure 2.18[52]-[54]. After the chamber was evacuated to 106 Torr, mixtures of CH4 and H2 gases were fed into a fused silica vessel from the top of the reactor. The microwave energy was transmitted from a 2.45 GHz microwave generator through a waveguide surfatron into the fused silica vessel. A plasma was created and stabilized by adjusting the coaxial tuning stub and waveguide tuning stubs in the waveguide surfa- tron. A silicon substrate was placed on a molybdenum holder, which con- tained a tungsten filament heater. Detailed experimental variables that were reported from Ref.[54] (1) Controllable input variables(U1): a) Input power=1.15 kW, b) Gas mixture:0.75%CH4, c) Total flow rate=100 sccm, d) Deposition pressure=l5 Torr (2 kPa), (2) Reactor design variables(Uz): a) Type of power source: a 2.45 GHz microwave power, b) Reactor configuration is shown in Figure 2.18, c) Reactor size=25 mm i.d quartz tube, (3) Deposition process variables(ng a) Deposition time=2h[53], 60 Fused Silica Vessel CH4 + H2 Coaxial Tuning Stub Coaxial Tuning Stub Waveguide 2.45 GHz Microwave : Surfatron Plasma ‘ — ~— - _ “ J Substrate Waveguide Tuning Stub Molybdenum Holder Tungsten Filament Figure 2.18 Schematic illustration of a surface-wave microwave PACVD apparatus[52]-[54]. 61 b) Substrates were pretreated by 20-40 um diamond powder in an ultrasonic methanol bath for 60 min, followed by cleaning with methanol, c) Substrate size=2 cm in diameter, d) substrate was heated by an external tungsten heater, (B) Internal variables(X): (l) Substrate temperature=950°C(l223°K), (C) Figures of merit calculated for this review: (1) (2) (3) (4) (5) (6) Linear growth rate=0.4-0.6 um/ h, Deposition area: 3.14 cmz, Weight gain=0.4-0.6 mg/ h. Specific yield=2445 kW-h / g, Gas flow efficiency=0.08 mg/ Liter, Carbon conversion efficiency=1.95. 62 2.6 Summary The performance “figures of merit” of diamond film deposition reac- tors described in Section 2.4 and 2.5 are summarized in Table 2.1, Figure 2.19, and Figure 2.20. Table 2.1 lists the “figures of merit” of various dia- mond film deposition reactors. Figure 2.19 shows the relationships between linear growth rate (um/ h), area power density (defined as input power divided by deposition area, kW/cmz), and specific yield (kW-h/ Ct) of the diamond film deposition reactors. Note that performance data from all reactors except the microwave reactors are denoted by empty circles. Data from microwave diamond film deposition reactors are denoted by solid circles. All of the diamond film synthesis techniques that have been reviewed in this chapter have the following aspects in common: ( 1) Deposition rates increase as area power density increases. As shown in Figure 2.19, the conventional DC arcjet reactor( (2)), the atrno- sphere DC arcjet reactor(@), the RF thermal plasma torch(@), and the microwave plasma jet (®) employ high area power densities to produce high film growth rates. (2) Figure 2.20 shows the relationship between the linear growth rate and deposition pressure. Deposition rates increase as deposition pressure increases. As shown, in the low-pressure nonequilibrium dis- charge regime (<80 Torr), linear growth rates are less than 3 m/ h. When the reactors operated in the high-pressure thermal discharge regime (>200 Torr) such as (7), @, Q), @, @, growth rates are above 10 um/ h. (3) Bell jar microwave plasma reactors described in Section 2.5.4 63 have better gas utilization efiiciency of depositing diamond films than other plasma reactors. (4) The carbon source in the reactive gases is less than 10% by the volume diluted in hydrogen, (5) The substrate temperatures for synthesis of high quality dia- mond films are between 300-1 100°C. (6) The best energr efiiciency expressed in terms of energy/weight gain approaches 10 kW-h/ Ct. Assuming an energy cost of ten cents per kW-h, the electrical energy costs approach one dollar/ Ct. Using the data displayed in Figure 2.19 and Figure 2.20, this cost can be achieved by both low-pressure and high-pressure diamond film deposition reactors. 64 Table 2.1 Summary of “figures of merit” Figum of Merits Deposition Carbon Area Conversion (m2) Efficinecy (%) Conventional COOS-0.03 3X10'3 ~0.08% HFCVD[20] urn/h mg/h mglliter Electron-amisted 05-3 .0 0.35-2.1 2 cm2 571- 0058-035 0.54-1.31% HFCVD[22] urn/h org/h 3428 mg/liter Conventional DC 20 1.76 0.25 cm2 1784 0.29 2.73% PlasmalM] um/b mg/h kW-h/g mg/liter Enclosed DC are 0.6-3.6 2-12 9 5 m2 250- 0.01- 0.06-0.12% jetl26] urn/b mg/h 1500 0.0017 mglliter Atmospheric DC Arc 10-30 2.67-13.37 1.27 cm2 7479 9.47x10‘5 0.018-0.089% “”71 “mm mm 4.74x10“ mglliter Atmospheric Combus- 10-15 2.49-3.74 0.71 cm2 NIA 0.033- 0.006-0.009% tion Flamel29] urn/h mg/h 0.050 rug/liter Enclosed Flat 1 rim/h 6.88 19.6 cm2 N/A 0.012- 0.0014- Flarne[32] mglh 0.056 0.0057% mglliter RF Thermal PACVD 30 826.6 735 cm2 72.5 0.134 3.21% 'lbrchl37] )1th mg/h kW-hlg mg/liter Magneto-microwave 0.4-0.8 0.69-1.38 4.91 cm2 217-434 0.097- 0.21-0.43% PACVD[39] [.1th mg/h kW—h/g 0.193 mg/liter Thhular Microwave 0.8 1.12 4 cm2 536 0.059- 0.97-1.24% PACVDl8] rim/h mg/h kW-h/g 0.092 mg/liter Microwave Plasma 6-12 527-1053 2.5 cm2 360-800 0.002- 0.018-0.109% Jdl42] urn/h mglh kW-hlg 0.004 Dig/liter Astex Bell-Jar Mleow- 3.5 15.4 125 cm2 97 0.42-1.28 16-48% ave PACVDI45] urn/b rug/b kW-hlg mg/liter UC-Berlrery Micro- 0.02 0.034 4.91 “1,2 16470 0.11 0.35 wave PACVD[50] rim/h rug/h kW-h/g mglliter MSU Moderate- 055 14.86 31 cm2 163 0.61 2%-32% Pressure MCPPISI] urn/h mg/h kW-hlg mg/liter MSU High-Pressure 6.27 44.68 2027 cm: 69 1.2 7.7% MCPR[II] jun/h mg/h kW-h/g mg/liter Surface-wave 0.4-0.6 0.4-0.6 3.14 cm2 2445 0.08 1.95 PACVDI54] pm/h mg/h kW-hlg rug/liter 65 10 h V I TY‘V'Y' V I VtT‘Y'l j fi—r7""' 7 I V 'I'I'I p d O to I r Linear Growth Rate (um/h) 8 .5 o 1000 kW-h/CI _: -1 10 : 1 10.2 1 .a.... r A naaanal a . AIAAAJL 4 a .o....l . a AAAAAA 10‘2 to" 10° 10‘ lo2 103 Area Power Density (kW/cmz) LEQEND 1.HFCVD[22] 8.Magleto-microwave CVD[39] 2.Conventional DC CVD[24] 9.Tubular Microwave CVD[8] . - VD 10.Microwave Plasma Jet CVD[42] 3 E"°'°s°d DC Arcletp [26] 11.Astex Bell Jar Microwave CVD[45] 4.Atrnosphere DC Arclet CVD[27] . 12.MSU Bell Jar Microwave CVD[51] 73F Thelma' P'asma T°'°hl371 13.MSU Bell Jar Microwave CVD(High Pressure)[11] Figure 2. 19 Summary of linear growth rate vs. area power density for diamond film deposition reactors. 66 1o_ . . mm, . . ....., . . .W E i : ® @ 1 ® 8 ‘ 101 - 3 E ‘9 = a CD ; r: 10° .- 1 E 2 6 % ® 3 ‘2 : 3 8 i J C :r l . 10'1 :- . 10'2 i . . ...i.1 L . . ..Lr.l i A a .iLL. 10° 101 102 103 Pressure (Torr) LEGEND 1.HFCVD[22]] 7.RF Thermal Plasma Torch[37] 2.Conventional oc cvo[24] g-Muiorllaetnmicrowavec $33139] .Enclosed DC Ar 'et cvo 26 - U ’ “0W9 : ”nowhere mam 030327] 10.Microwave Plasma Jet CVD[42] ' , 11.Astex Bell Jar Microwave CVD[45] 5-Amsph9'9 COmDUSt'O" Flam°l29l 12.MSU Bell Jar Microwave CVD[51] 6.Enc|osed Combustion Flame[32] 13.MSU Bell Jar Microwave CVD(High Pressure)[11j Figure 2.20 Summary of linear growth rate vs. pressure for diamond film deposition reactors. CHAPTER 3 High-Pressure Microwave Cavity Plasma Reactor: Experimental Systems, Experimental Procedures, Experimental Parame- ter Space, Measurement Methodologies, and Reactor Configuration 3. 1 Introduction The objectives of this chapter were concerned with improving the moderate-pressure microwave cavity plasma reactor(MCPR) by extending its operation to higher pressures. Of particular importance was the objec- tive of increasing the deposition rate and the energy efficiency of the microwave plasma reactors. In order to achieve these research objectives, the thermally floating MCPR reactor needed to be redesigned to operate at higher pressures (80-140 Torr) and higher input powers (2.5-4 kW). The redesigned reactor, identified here as high-pressure MCPR, includes a water-cooled substrate holder setup and a better air cooling system that is applied to cool the microwave cavity, bell jar, etc. This chapter first presents the experimental systems that were employed in the experiments that are described in this chapter. The experimental systems consist of the microwave power supply and 67 68 waveguide / transmission systems, flow control and vacuum systems, and a computer monitor system. Then the common experimental procedures including the seeding procedure and start-up and shut-down procedures are described. The input and output parameter space that was used in the experiments presented in this thesis is also described. The design of water -cooled substrate holder setup which was needed in order to operate the reactor in high pressures is highlighted in the end of this chapter. 3.2 Experimental Systems 3.2. 1 Introduction The two microwave plasma diamond depositing reactors that are used in this thesis research, i.e., high-pressure MCPR (Chapter 3,4,5) and microwave plasma jet reactor (MPJR) (Chapter 6,7,8) use the same micro- wave power supply and waveguide/ transmission system, vacuum pump and the gas flow control system, and computer monitor system. These systems are described below in Sections 3.2.2, 3.2.3, and 3.2.4. 3.2.2 Microwave power supply and waveguide] transmission sys- tem Figure 3.1 schematically displays the microwave power supply and waveguide/ transmission systems. The microwave power supply (1) con- sists of a magnetron (2), a circulator (3), and a dummy load (4). The waveguide/ transmission systems consist of the rigid waveguides (5), a 69 dual-directional coupler (6), incident (7) and reflected (8) power meters, a flexible rectangular waveguide (9), and a rectangular waveguide to coaxial transition unit (10). The microwave energr supplied by the magnetron (2), identified here as the incident power, Pmc, is propagated through a set of rectangu- lar waveguides(5), (9), transmitted to a coaxial waveguide (l 1) by the tran- sition unit (10), and coupled into the cavity applicator (12) through a mechanically tunable coaxial excitation probe (13). The excitation probe is located in the center of the sliding short (14) and the sliding short can be 70 2de 98 ~50: mammoarnwfi .8.“ 889mm coammgmamb \ weavers? 963803" can 339% .8309 963803 H .m mama bmmsm 830m 385822 25 0 £20 med HonoUE \ll J raise ego cohocwaz E you—82mm? 193—«U 35 All. . .......... ... P Ls _ . L _C , Hoaaaoeufi . ..................N..l. e _ team mafia a: ll Boa aousoxm ac ll $3353 388:: ll 53:00 826m #6583235 . , @ cosmos; 2&on @ : / _, _ E Co aqua wocawgwg Ema B. u. D u. H 83 .832 830m “any: 5 .832 826m vmuuoamm 8v 833m aoeeoz eoaamaou $5 71 moved up and down along the reactor axis to control the applicator height, L8. The depth of coaxial excitation probe, Lp, is also independently adjustable. In the case of any tuning mismatch, some of the incident power will be reflected back from the cavity applicator. This reflected power, Pref, propagates through the dual-directional coupler (6) and is directed by the circulator (3) into the matched dummy load (4), where it is absorbed and dissipated as thermal energy. The power absorbed by dis- charge loaded applicator, Pt, is equal to the difference between the inci- dent power and reflected power, i.e., Pt: Pine-Pref . The microwave power source is a Cober (Model No. SGF/ 4503), 2.45 GHz, 6 kW power source. 3.2.3 Flow Control and Vacuum Pump System Figure 3.2 shows the schematic drawing of the flow control and vacuum systems. The source gases, which consist of C02 (1), H2 (2), CO (3), and CH4(4) of a respective purity of 99.9%, 99.999%, 99.9%, and 99.99%, are monitored by four MKS type 1 1 159A mass flow controllers (5) along with a 4-channel MKS flow controller (6). The flow rates of source gases are set and displayed on this 4-channel MKS flow controller (6) by the system operator. The source gases with desired flow rates are mixed before they enter the baseplate (7). An ALCATEL 2033 type mechanical pump (12) is used to pump down the chamber pressure to ~5 mTorr, which is measured by a low pressure gauge (13). Operating pressure is set on the pressure controller (14) by the system operator and its value is 72 displayed on the high pressure gauge (10). A manual valve (1 1), which is normally open, is installed and can be used to manually control the 73 (6) 4-Channel MKS flow controller [0000' (5) Mass Flow Controllers i I (7) Base Plate 1 . (8)Chamber l l (13) (11) Low Pressure Manual Valve Gauge l J (9) Automatic Throttle (16) Valve (14) Pressure (10) { S stem Vent‘ High Pressure Gauge alve Controller (12) Mechanical RoughingPpump (17) Exhaust Figure 3.2 Flow control and vacuum system for high-pressure MCPR and MPJR. 74 pressure of the chamber if the automatic throttle valve fails. Nitrogen (15) is released at gas pressure of 10 psi through the system vent valve (16) to bring the chamber into the atmospheric pressure. It is also used to dilute the exhaust gases (1?) through an exhaust valve (18). The mixing of nitro- gen into the exhaust gas stream insures that the exhaust gases are below the flammable condition. 3.2.4 Computer Monitoring System A computer monitoring system was used to observe and regulate the system operating conditions and control the shut down procedure in the experiments described in this thesis research. Figure 3.109 displays the flow chart of the computer monitoring program[9]. As shown, the experimental running time, reflected power upper limit, and the operating pressure threhold are first set in the computer. The experimental system is then enabled to allow the feed gases to flow and the microwave power to turn on. After the plasma is excited, a checking loop compares the pres- sure, reflected microwave power, and time with the preset values to deter- mine the state of the experiment. An emergency shut down of the microwave power and feed gas is performed if the experimental running time, reflected power, and operating pressure are over the preset values at any time during the experiment. In the meantime, the automatic throttle valve is closed in order to isolate the vacuum pump from the process chamber. On the other hand, when the experiment is completed. 75 (see Sectin 3.3.2) . Start timer ()1 #CJONl—I . Set the threhold pressure . Set experiment running time . Enable the experimental system . Experimental start-up procedure H 2. Check timer . Check and record the operating pressure and reflected power Timer Expires Operating pressure or reflected power over the threhold values Normal shut-down sequence (sec Section 3.3.2) Emergency shut-down squence 1. Turn off microwave power 2. Turn off all gas flows 3. Close the automatic throttle valve Figure 3.3 Flow chart of Computer monitor program[9]. 76 the computer program directs the system into a normal shut-down sequence (see section 3.3.2). 3.3 Common Experimental Procedures 3.3.1 Seeding Procedure A seeding procedure was perfOrmed on the substrate surface in preparation for the deposition experiments described in this dissertation. The diamond powder and photoresist were first mixed and then agitated in an ultrasonic bath. The resulting solution was then dropped over the entire substrate surface. The substrate was rapidly rotated in a spinner, resulting in the uniform seeding over the substrate surface. Using this photoresist seeding method, the diamond seeds were embedded more uniformly and the nucleation density was estimated to be 108cm"2 [101.[51]. The recipe for the photoresist substrate seeding, which was rou- tinely used in this research, is summarized as follows: (1) bake 200 mg Amplex, 0.1um diamond powder at 150°C for 30 minutes, (2) pour 20 ml Shipley type A photo resist thinner into a 53 ml Shapely 1830 photo resist, (3) mix 0.1 pm diamond powder in (1) with the solution in (2), (4) ultrasonically agitate the mixture for 1 hour, 77 (5) drop the mixture onto the substrate surface with full coverage, (6) spin the substrate at 4000 rpm for 40 seconds, (7) bake the substrate at 150C for 15 minutes. 3.3.2 Start-up and Shut-down Procedures After manually placing the substrate on the substrate holder and setting the cavity length (LS~21 cm) and probe depth (I1,~3 cm), the reactor was pumped to 5 mTorr by a mechanical pump ((12) in Figure 3.2). Source gases with appropriate gas mixtures and flow rates were introduced into the reactor and then the chamber pressure was gradually increased. At chamber pressure of ~10 Torr, microwave power of 800 W was incident and ignited the discharge. The desired chamber pressures between 80-140 Torr were achieved in approximately five minutes by the value that was preset on the pressure controller ((14) in Figure 3.2.). Absorbed microwave power was adjusted between 3-4.5 kW so that the 2" (5.08 cm) diameter substrate was covered with the discharge plasma. Substrate temperature was independently controlled between 700- 1 150°C by varying the number of substrate discs (see Section 3.6.3.2 and 3.6.3.3). Methane and hydrogen were used as source gases and their flow rates were controlled by a 4-channel flow controller. During an actual deposition experiment, all the experimental input variables such as flow rate, deposition time, microwave power, etc. are monitored and recorded vs. running time by a computer (see Section 3.2.4). After the experiment 78 was completed, the experimental system was shut down with the procedures described below: (1) automatically turn off CH4, C02, and CO gas flow by shutting off the channels in gas flow controller, (2) 3-minute self-cleaning process with H2 plasma, (3) automatically turn off microwave power, (4) automatically turn off H2 gas flow by shutting off the channel in gas flow controller, (5) automatically evacuate the chamber by activating the throttle valve to the open position. 3.4 Experimental Parameter Space 3.4.1 Introduction As mentioned in the literature review shown in Chapter 2, the out- put performance of a plasma diamond deposition reactor is a complex function of many experimental variables. Thus in a well understood and repeatable experimental system each of these experimental variables are identified and controlled. The reactor block diagram shown in Fig.2.2 arranges the reactor variables into three basic groups: (1) input variables, U, (2) the internal variables, X, and (3) the output variables, Y. Using the definitions of each variable group, as described in Chapter 2, a block dia- gram for the high-pressure MCPR diamond deposition reactor is dis- £3088 Confluence 83 95:86 8 NED—2 chamoarawfi «5 com mofimtgaaz Em mama ease 5.: 32292,. 8:... 0 £80 5.2% 5E a .555 SE a 59980 :83on seesaw 25: so 2c; 0588? 2am 5,35 .983 sec 5320 52:? H5 3.236220 as. % am as _ Es: 22.38 2:885 58mm 28% am oeccsgmo mo 5.8_£_s> 882.. 8... o a a «2.3 as: 8528". w 222858 5.85 > oaafimaea. mac Wl 83%.; oE=_o> $9285 595 2.832.; .23: m? 35535.. .288 3.08: 8.358 5.: Basso .08 . 2.2 25283 J 2385 8288961 2am 25: 8% EEEBO $63 I 826m 59% l x asaczaseoosfiam $.38? 5.2% =m>> 5.8% 55.6580 388». mesomaaom .o 8 .P o a a «2.8.3; 3.8: .288 5.833.; 5%. 23:95.8 80 played in Figure 3.4. The detailed descriptions of each group are dis- cussed in the sections below. 3.4.2 Independent Input Experimental Variables Figure 3.4 shows a general block diagram of the experimental variables for a high-pressure MPCR CVD diamond deposition reactor. Table 3.1 identifies the range of experimental deposition variables chosen for this experimental investigation. The independent controllable input variables, U1, are substrate temperature Ts, pressure p, incident microwave power Pine, gas chemistry c (expressed in %CH4/H2)and total gas flow rate ft. Their respective ranges varied from 700-1 150C, 80-140 Torr, 3-5 kW, 1-8%CH4/H2 and 100-1400 sccm. Reactor geometry variables, U2, were fixed except for the change in the substrate holders which allowed the substrate temperature to be independently varied. Deposition process variables, U3, include substrate size and material, seeding procedure and deposition time. Deposition time was varied from 4h to 100h. Undoped (100) 2" diameter silicon wafers were used as substrates for all the experiments presented in this thesis. The substrates are seeded with a 0.1 um diamond powder-photoresist solution as described in Section 3.3.1. 81 Table 3.1 Definition of the experimental variables for the high-pressure MCPR. a) Deposition Pressure, Variable: p=80-150Torr Controllable b) Absorbed Microwave Power, Variable: Pt=2.5-4.5kw Input Variables, c) Gas Chemistry, Variable: c=1-8% CH4/H2 01 d) Substrate Temperature, Variable: Ts=700-l 150C c) Total Flow Rate, Variable: {5100-1400 seem a) Reactor Configuration, Variable: Substrate Holder Input Reactor Setups Variable, Geometry b) End-feed Excitation, Fixed U Variables, U2 c) Electromagnetic Mode and Cavity Tuning, Fixed at mom (1) Quartz Dome Geometry, Fixed Deposition a) Substrate Material and Size, fixed: 2" dia. Silicon Process b) Substrate Seeding Procedure, Fixed: Seeded by a Variables U Spinner with 0.1lirn Diamond Powder-Photoresist . a Solution c) Deposition Time, Variable: t=4-100 I-Iours Internal a) Plasma Volume, Vd, Approximately Constant at Variable, 144cm3 X b) Absorbed Power Density, Variable: =Pt/Vd a) Linear Growth Rate: 4-7 um/h Output Reactor b) Total Growth Rate: 27-50 mg/ h Variable, Performance. Y1: c) Specific Yield: 69-300 kw-h/g Y (1) Gas Flow Efficiency: 0.6-1.2 rug/liter e) Carbon Conversion Efficiency: 1-12% Film Characteristics, Y2: a) Film Uniformity: 10% or better over 2" dia. area b) Film Structural Quality: FWHM=3-8 cm'l c) Film Morpholoy: mostly (I 1 1) morphology (1) Film Texture: mostly < 1 10> texture 82 3.4.3 Dependent Internal Experimental Variables The experimentally measured internal variables, X, in the high- pressure MCPR were (1) the power absorbed by the reactor, Pt, (2) the plasma volume, Vd, and (3) the absorbed power density,

=Pl/Vd. The power absorbed by the reactor Pt, is defined as the difference between the incident microwave power Pmc and the reflected power Pref, i.e. Pt=Plnc- Pref. The plasma volume Vd was always adjusted so that the hemispherical discharge covered the substrate and produced good deposition uniformity. As the pressure varied from run to run the absorbed power was adjusted to produce a discharge that covered the substrate. Thus the discharge volume was approximately constant at 144 cm3 for all experiments and Pt and

increased with pressure. 3.4.4 External Output Experimental Variables As shown in Figure 3.4, the reactor output variables (Y) in the high- pressure MCPR are defined as reactor performance variables (Y 1) and film characteristic variables (Y2). The reactor performance variables contain film linear growth rate, film total growth rate, specific yield, gas flow efficiency, carbon conversion efficiency. The film characteristic variables include film uniformity, film structural quality, film morphology, and film texture. Each of the reactor performance variables has been defined in Chapter 2 (see page 16-18). The techniques of measuring the film 83 characteristic variables are described in the next section. 3.5 Measurement of Experimental Output Variables 3.5. 1 ' Measurement of Film Characteristics 3.5.1.1 Measurement of Film Uniformity Film uniformity was determined by (l) breaking the wafer and mea- suring the thickness vs. position by a SEM, or (2) for thick films by directly measuring the thickness of free standing films with micrometer. Percent uniformity is defined by the difference between the maximum and minimum thickness measurements divided by the average thickness measurement over the 5 cm diameter substrate. 3.5.1.2 Measurement of Film Structural Quality The Raman Spectroscopy equipment was developed by Moss- brucker etaL[67] at MSU. In the system, the Raman spectra were excited with the 514 nm line of a green Ar+ laser with an output power of 600 mWatt. A 60X optical objective lens that illuminated a spot with a diame- ter of 30 um was used for all the Raman spectra reported here. In the later chapters, the full width at half of its height maximum (FWHM) of Raman spectra for the diamond peak at 1332 cm’1 will be used as an indication of the structural quality of diamond films. Since the laser 84 has a non-zero, small but finite band width, the corrected FWHM will be broadened by the following equationl67],[68): 2 2 2 M = A/FSample +FMono +FLaser ’ where M = Corrected FWHM measurement, FSample = Measured FWHM of the sample, FMono = FWHM of the monochromator, Fuse, = FWHM of the laser. 3.5. 1.3 Measurement of Film Texture X-ray diffraction (XRD) measurement of films with film thickness greater than 5 um[69] provide an indication of growth orientation in the crystalline films[70]. The location of XRD peaks and their relative intensi- ties in a powder pattern is shown in Table 3.2. Generally, the difi'ractome- ter for XRD spectra reported in this dissertation use a stop size of 0.10 and a 3 second integration. The growth orientation ratio utilized in the later chapters is defined by the ratio of [220] peak height to [111] peak height, i.e., I[220]/I[ l l 1]. Since the powder pattern intensity of [1 l 1] peak is dominant over other XRD peaks (note that the powder is spread ran- domly on the detecting stage), a growth orientation ratio << 1 indicates lit- tle orientation in that crystallographic direction while a ratio >> 1 shows significant orientation of these crystal grains in the filml22]. . ...__ _.-._._.-_ 85 Table 3.2 Location of XRD peaks and their relative intensity in a powder pattern. XRD peak 29 Position Powder Pattern Intensity [l l 1] 43.90 100% [220] 75.30 25% [3 l 1] 9 1 .50 l 6% [4001 1 19.50 8% [331] 140.50 16% 3.5. 1.4 Measurement of Film Morphology Since Scanning Electron Microscopy (SEM) is a well known tech- nique that has higher resolution, as well as a wider and higher magnifica- tion range than light microscopy[71], it has been utilized for identifying the surface morpholog' and thickness of the diamond films reported in this dissertation. In order to get a good SEM photo, the diamond film sample was required to be (1) devoid of solvents, water, or grease that may vaporize in the vacuum, (2) firmly mounted on the specimen stage, and (3) electrically conductive. Using these techniques, a typical SEM image that resulted from the J EOL 6400 SEM at MSU is shown in Figure 3.5. The terminology of surface morphologies used in this dissertation is also defined in Figure 3.5[72],[73]. 86 (l 10) Morphologr (100) Morphology (11 1) Morphology Figure 3.5 A typical SEM photo from JEOL 6400 at MSU. 87 3.5.2 Calculation of Reactor Performance 3.5.2. 1 Measurement of Film Growth Rate Film total growth rate, W in mg/ h was determined by measuring the total weight gain of the substrate, Wt, during an experimental run and dividing this gain by the deposition time t. The average linear growth rate(um/ h), d, was determined from d = AXE , where W is the weight gain per hour, A is the deposition area, D is the diamond density of 3.51 g/ cm3ll4]. As an example, the linear growth rate for the high-pressure MCPR is calculated here. As shown in Table 2.1, the total growth rate for the high-pressure MCPR is measured to be 44.86 mg/h over 2" diameter sili- con substrate (with experimental conditions of CH4/ H2=18/ 600 sccm and absorbed microwave power of 3.1 kW). Then the linear growth rate is: 4 44.68 as x m h cm linear growth rate (um/ h) = 3 20.27cm2x3.51 g3x10 mg cm =6.27 lim/h. 3.5.2.2 Measurement of Specific Yield Specific Yield (kW-h/ g) is defined as the power input (kW) per dia- mond film total growth rate in the units of g/h[9]. Using the same exam- 88 ple that is described in the section above, the total growth rate in the high-pressure MCPR is 44.86 mg/h over 2" diameter silicon substrate, then the specific yield is: 3.1kW as x _g h 103mg specific yield (kW-h / g) 44.68 =69 kW-h / g. 3.5.2.3 Measurement of Gas Efficiency Gas Efficiency (mg/liter) is defined as the total growth rate (mg/ min) divided by total flow rate (liter/ min). The total flow rate is defined as the sum of all input gas flow rates[9]. Using the same example that is described in the section above, i.e. the total growth rate in the high-pres- sure MCPR is 44.86 mg/ h over 2" diameter silicon substrate, then the gas efficiency is: 44.68 1:15 gas efficiency (mg/ liter) = cc 1 60min x x 618 . min lO—3liter h =1 .2 mg/ liter. 3.5.2.4 Measurement of Carbon Conversion Efficiency Carbon Conversion Efficiency is defined as the percentage of car- bon atoms in the input gases which are converted into dia- 89 mond[9],[16],[17]. If the total growth rate in the high-pressure MCPR is 44.86 mg/ h over 2” diameter silicon substrate, then the carbon conver- sion efficiency is: carbon conversion efficiency (%) 23 44.68 "T‘IEX g 1h mole 6.02x10 atom 103mg x 60min X 12g X mole x100% cc liter mole 6.02 x 1023atom 18—. X X . X min 103cc 22.411ter mole =7.7%. 3.6 Design of High-Pressure MCPR 3.6. 1 Introduction As mentioned in Section 3.1, the objectives of this chapter research were to increase the deposition rate and energy efficiency of the thermally floating MCPR that was developed by Zhangl9). In order to achieve these objectives, the reactor has to be operated in a higher pressure, higher power density regime (see Figure 2.19 and Figure 2.20). However, when the reactor was operated in this regime, the temperature of the substrate increases above 1 150°C where diamond deposition does not take place and the reactor components such as quartz bell jar, microwave cavity, etc., became overheated. A reactor operating field map for the thermally floating reactor, which was developed by Zhangl9) and Khatami [51], 90 shows that as the operating pressure and absorbed microwave power increase, the substrate temperature also increases. As shown in Figure 3.6, the substrate temperature reaches 1000°C at pressure of 56 Torr. Thus the addition of a water cooling stage and a better air cooling system became necessary when the operation of ther- mally floating MCPR is extended to higher pressures. This section begins with the description of the original MCPR that was developed by J. Zhang[9]. Then the design of the water cooling stage setups and the air cooling system that were developed and employed during the course of this thesis research is described. A I I / / 1000 // 56 Torr 4// 6 / ' '7 / 2.. / / i—«m / / . 950 . / 50 Torr 4— / as; (Vd) mm /__f / / +5 \ / // a 900 / 45 Torr 4.. / E I ” ,’ f-' / / g // 38 Torr x__ / \ a 850 / 7 // (Vd) max .o / , ”5’ / 34 T ’ 800 L 0” ‘7 1.4 1.6 1.8 2.0 2.2 2.? Absorbed microwave power, Pt (KW) Figure 3.6 Reactor operating field map under thermally floating config- uration For 5” quartz dome/3” substrate reactor configurationlSl]. 91 3.6.2 Operation of The Thermally Floating MCPR 3.6.2.1 Operation of Microwave Cavity Excitation and Mg A cross sectional view of the thermally floating MCPR is shown in Figure 3.7. As shown, the applicator side wall(1) consists of a 17.8 cm inside diameter cylindrical brass tube. This brass tube is electrically connected to a water-cooled baseplate assembly(2-4) and a water cooled sliding short(l 1) via finger stock(12). Thus the cylindrical volume bounded by the sliding short, the side walls and the baseplate forms the cylindrical cavity applicator electromagnetic excitation region. As shown, 2.45 GHz (CW) microwave power is coupled into the cylindrical cavity applicator through a mechanically tunable coaxial excitation probe(l3) located in the center of the sliding short. The sliding short can be moved up and down along the reactor axis to control the applicator height, L , and the depth of coaxial excitation probe, Lp, is also independently adjustable. Thus by changing the probe depth and the cavity height, the applicator can be excited and matched to the desirable electromagnetic resonance. 3.6.2.2 Operation of Air Cooling System The only air cooling system that was used in the thermal floating MCPR consists of a air blower with 60 CFM (cubic foot per minute). As 92 I"S CoolingAir 20 2 Cooling Air : -53.; 32 i?“ 4 . ..17 _ 19 Legend (1) Cavity Side Wall (2) Baseplate (3) Annular Plate (4) Distribution Plate (5) Quartz Dome (10) Substrate (11) Sliding Short (12) Finger Stock (13) Excitation Probe (14) Plasma Discharge (15) View Window (16) Flow Paitem (17) Quartz Tube (18) Metal Tube (19) Metal Plate (20) Air Blower Inlet (21) Teflon Pieces (32) Laser Ports Figure 3.7 The cross sectional view of the thermally floating MCPR sys- tem. 93 shown in Figure 3.7, this air blower forms an air cooling stream by blow- ing the cooling air from the air blower inlet (20), onto the bell jar (5) and cavity side walls (1), and finally flows out of the cavity through the view window (15) and two pairs of laser diagnostic ports (32). 3.6.2.3 Operation of Thermally Floating Substrate Holder Setup The baseplate assembly consists of a water-cooled and air-cooled baseplate(2), an annular input gas feed plate(3), a gas distribution plate(4), a 12.5cm i.d. quartz dome(5), the thermallyfloating substrate holder setup assembly (16-19), and the substrate itself (10). The baseplate. the annular gas plate, and the gas distribution plate introduce an uniform ring of input gas into the quartz dome volume. A hemisphere shaped discharge(14) is positioned over and is adjusted to be in direct contact with the top of the substrate surface by varying LF and Ls to excite and match the TMO 13 plasma-loaded resonant mode. The discharge can be viewed through the screened side window(15) and the substrate temperature is measured by focusing an optical pyrometer through the screened side -window(15). The substrate (10) is placed on top of a flow pattern regulator (16) 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. A metal tube (1 8) which serves as a conductive tube is placed inside the quartz tube (17) and prevents the plasma discharge from forming underneath the substrate by reducing the 94 the substrate by reducing the electric field underneath the substrate. The metal tube (18) and quartz tube (17) are placed on a metal plate (19) which has 30 mm diameter hole in its center to pass the hot gases from with the quartz dome (5) to the pump. 3.6.3 Operation of The High-Pressure MCPR 3.6.3.1 Introduction This section first presents the change in the air cooling system that was used to cool the microwave cavity and bell jar. Then the design of the water-cooled substrate holder setup that replaced the thermally floating substrate holder setup in the MCPR (described in section above) is described. Since the high-pressure MCPR, which is shown in Figure 3.8, used the same microwave cavity and microwave power deliver systems, the operation of microwave cavity excitation and tuning is referred to Sec- tion 3.6.2.1. 3.6.3.2 The Change In The Air Cooling System In order to utilize the air blower existing inside the Cober micro- wave power supply and add another air cooling stream into the high-pres- sure MCPR, two Teflon pieces (as shown in (21) Figure 3.8) were drilled four of 1/8” diameter through holes. This change allows the cooling air (from the air blower in the microwave power supply, not shown in 95 Legend (1) Cavity Side Wall (2) Baseplate (3) Annular Plate (4) Distribution Plate (5) Quartz Dome (6) Cooling Stage Wall (10) Substrate (11) Sliding Short (12) Finger Stock (13) Excitation Probe (14) Plasma Discharge (15) View Window (16) Flow Pattern Regulator (17) Quartz Tube (19) Insulation Disc Set (20) Air Blower Inlet (21) Teflon Pieces (32) Laser Ports Figure 3.8 The cross sectional view of the high-pressure MCPR system. 96 Figure 3.8) to flow through the coaxial waveguide, onto the bell jar (5) and cavity side walls (1), and to flow out of the view window(15) and the three laser ports (32). This additional air cooling not only allowed the additional cooling of the quartz bell jar (5) but also provided cooling of the input coaxial waveguide system. 3.6.3.3 Substrate Cooling Stage - The First Design Figure 3.9 shows the cross sectional view of the first prototype water-cooled substrate holder setup that was developed for high-pressure diamond film deposition in Michigan State University. As shown, a 2" in diameter silicon substrate(Figure 3.9(a)) was placed on the 2” in diameter graphite holder(Figure 3.9(b)). A boron nitride ring(Figure 3.9(e)) was used to keep the substrate at the center of the reactor. A set of insulating discs(Figure 3.9(d)), the gas flow pattern regulator(Figure 3.9(e)), and the 2” tall, 3" diameter water cooling stage(Figure 3.9(fl) were employed to control the substrate temperature. The gas flow pattern regulator(Figure 3.9(e)), which is placed on the top of water cooling stage, is an electrically conducting disc (made from either graphite or molybdenum) with a ring of holes circumferentially surrounding the substrate holding recess. Substrate temperature could be independently controlled by placing different layers of either graphite or boron nitride insulation disks(Figure 3.9(d) between the gas flow pattern regulator(Figure 3.9(e)) and the 2” diameter graphite holder(Figure 3.9(b)). 97 1 (a) P H (b) :1 .__—_1 (c) ; :(d) as: a (e) r 4 (D W flmfiTn—I—l—l—l Bottom Plate 1 l Water Outlet Water Outlet Figure 3.9 Typical substrate holder setup used in high-pressure MCPR for diamond film deposition: (a) 2” silicon wafer, (b) 2” graphite substrate holder, (c) boron nitride ring, ((1) insulation discs, (e) graphite flow regu- lator, (1) water coolling stage. 98 Figure 3.10 and Figure 3.1 1 schematically illuminate the construc- tion of the prototype water cooling stage. The can body, which is shown in Figure 3.10, consists of a 3.125” inside diameter stainless plate and a 2" tall stainless steel tube. These two parts were jointed by a vacuum weld. The bottom body, which is shown in Figure 3.1 1, contains the water inlet tube, the water outlet tube, and bottom plate. The cooling water was guided to the center of the body can by the 0.25" in diameter inlet tube and then released through the 0.25” in diameter outlet tube. 3.6.3.4 Drawbacks of The First Design The first substrate holder setup design and water cooling stage had the following drawbacks: Substrate holder setup: (1) The edge of 2" graphite holder was etched by the plasma after each run. This resulted in an unknown and an additional amount of car- bon contribution to the diamond film deposition process. The mainte- nance of the reactor was also increased by the need to replace the 2” graphite holder. (2) The resulting temperature at the center of the substrate was higher than the temperature at the edge of the substrate area. This is due to the observation that the plasma shape becomes more arc-like than hemisphere-like at high pressure regime. Water cooling stage: (1) The 0.25” in diameter water inlet and outlet were easily plugged 99 +— 3.250" = l——— 3.125” ———1 Top of Can body >C; '3 Tube of Can body I _,. 2.000” 1 .1 Figure 3.10 Can Body of the 3" Water Cooling Stage[107] 2.500” —‘-‘ ] 4— 0.250” Water Outlet Bottom Plate 1 W l I | l | | l | l I —~ <— T 0.250” Water Inlet 6.000” Figure 3.1 1 Bottom Body of the 3” Water Cooling Stage[107]. 100 by the debris from the university water. The installation of water filters became necessary. (2) A water leak from the welding section between the top of can body top and the tube of can body (shown in Figure 3.10) was observed. A redesigned substrate holder setup and water cooling stage elimi- nated these drawbacks. This redesigned system is described in the sec- tions below. 3.6.3.5 Final Substrate Cooling Stage Design Figure 3.1 2 schematically shows the redesigned high-pressure MCPR configuration for diamond film deposition. The major difierences between the revised configuration and prototype version are: For substrate holder setup: (1) 2" graphite holder was replaced by 2" conductive plate. This replacement can prevent the substrate holder from being etched by the plasma discharge. (2) The 2' conductive plate, as shown in Figure 3. 12, was curved. This configuration solved the problem of temperature nonuniformity over the substrate that was described in Section 3.2.3. For water cooling stage: (1) As shown in Figure 3.13, the can body was machined from one solid piece of stainless steel. This prevented the problem of water leakage. (2) As shown in Figure 3.13, the inside surface of top body can was 101 W \n (a) =3 =2 (b) .. .2 (c) m ‘F—‘CIFJ (d) (e) i 1 Water Outlet Water Outlet Figure 3. 12 Typical substrate holder setup used in high-pressure MCPR for diamond film deposition: (a) 2" silicon wafer, (b) 2" conductive plate, (c) boron nitride ring, ((1) insulation discs, (e) graphite flow regulator, (f) water coolling stage. 102 3.250” ——> 3.125” I l “APT I 2.000” .. .1 Figure 3.13 Can Body of the 3” Water Cooling Stage[107]. |-—— 2500” —~' I [ | ::1 I _, l ‘— | 0.500” Water Outlet Bottom Plate I l —~l‘l_l_‘—’“1‘T I (“WU—Fri l I | l l 0.500” Water Inlet 6.000” ~ Figure 3.14 Bottom Body of the 3” Water Cooling Stage[107]. 103 machined from a flat surface to a curved surface. This revision enhanced the cooling at the center of body can, where higher temperature takes place under thermal plasma discharge condition. (3) As shown in Figure 3.14, stainless steel tubes for water inlet and outlet were extended from 0.25" to 0.5” in diameter. This revision allowed more cooling water to flow inside the can body and prevented the input and output cooing tubes from being plugged by debris from university cooling water. 3.7 Summary As described earlier in this chapter, a 12.5 cm diameter plasma reactor, identified as thermally floating MCPR, was developed by Zhangl9) and KhatamiISl] and applied to diamond thin film deposition over 3" or 4” in diameter substrates in the 20-70 Torr pressure regime[9],[51]. As shown in Table 2.1, the total growth rate and specific yield over 3" silicon substrates for this reactor are 14.86 mg/h(0.55 um/h) and 163 kW-h/ g, respectively. This chapter has described a redesigned thermally floating MCPR reactor which includes a water-cooled substrate holder setup and a better air cooling system. With these additions, the reactor, identified here as high-pressure MCPR, can be operated under the higher pressure (80-150 Torr) and higher power (2-4.5 kW) regime. When operating under these higher pressure higher power conditions the plasma changes from a non-equilibrium discharge to a more thermal-like discharge. The next two chapters describe the operational performance and the characteristics of 104 the diamond films synthesized by this new reactor configuration. CHAPTER FOUR Reactor Experimental Output Performance - Reactor Performance (Y 1) 4.1 Introduction In general, the output performance of the reactor (Y) is a function of many input variables, i.e., Y=f(U1, U2, U3). As shown in Figure 4.1, the input variables are divided into three groups, (1) independent input vari- ables U1 which include (a) methane concentration, (b) total flow rate, (c) substrate temperature, ((1) absorbed microwave power, (e) deposition pressure; (2) reactor geometry variables 02 which include (a) quartz dome size, (b) water cooling stage geometry, (c) substrate type and size, ((1) reac- tor size, Ls, (e) coupling probe position, LP; (3) process variables, U3, which include (a) seeding procedure, (b) start-up and shut-down proce- dures, and (c) deposition time (t). Important internal variables X are the discharge plasma volume Vd and the absorbed microwave power density . The plasma volume Vd is held constant at ~144 cm3 for all the experimental runs in this chapter and chapter 5. As shown in Figure 4.1, 105 106 absorbed microwave power (Pt), deposition pressure (p), and microwave power volume density () are not independent of each other, i.e.,,=Pt/Vd. Thus the absorbed power and the pressure are directly related in all the experiments and thus these variables are referred to as a doublet input variable, i.e., as the absorbed microwave power- deposition pressure (Pt-p) variable. As also shown in Figure 4.1, the outputs of the reactor (Y) can be divided into two groups. Outputs Y1 are concerned with the reactor per- formance such as growth rate, specific yield, and carbon conversion effi- ciency. A second group of outputs Y2 are concerned with the film properties such as morphology, structural quality, and texture. The understanding of the reactor behavior is very difficult not only because of the numerous variables but also because the input vs. output behavior of each variable is nonlinear and complex. Theories that describe and pre- dict reactor behavior do not exist. Therefore it is necessary to experimentally explore the operational behavior of the redesigned high pressure MCPR. The experimental data presented in this chapter are the result of over 100 separate experiments representing over 1,000 hours of experimental operation. The resulting multiple dimensional input/ output experimental variable space is diffi- cult to visualize and plot. Thus this chapter presents an overview of the high pressure MCPR reactor performance Y1 versus a selected number of input variables. Each experimental run represents a data point in the multi ments tor in l expert- and Tl cones ture, ' D. As fixed, elect] Gift: for] plot able Elei tack 107 multidimensional input/ output variable space. In each of these experi- ments the output reactor performance Y1 was measured for specific reac- tor inputs. The reactor input and output variables and their associated experimental ranges explored in this chapter are displayed in Figure 4.1 and Table 4.1. The independent input variables chosen were (a) methane concentration, c=%CH4/H2, (b) total flow rate, ft, (c) substrate tempera- ture, Ts, (d) absorbed microwave power (Pt), and (e) deposition pressure, p. As indicated in Figure 4.1 and Table 4.1, all other input variables were fixed, i.e., the substrates were limited to two inch (100) silicon wafers, the electromagnetic excitation was fixed at TM013 mode, etc. This chapter presents an overview of the output reactor perfor- mance Y1 versus the five dimensional input variable space, i.e., Y1=f(c, ft, Ts, t, Pt-p). The output performance variable Y1 that are of interest are (a) film linear growth rate, expressed in um/h, (b) total growth rate, expressed as mg/h, (c) specific yield, expressed as kW-h/g, ((1) gas flow efficiency, expressed in mg/liter, and (e) carbon conversion efficiency. Experimental data in this chapter is only presented for films with an uni- formity of 15% or better. The multiple dimensional experimental variable space is difficult to plot on a single graph, Therefore the various reactor performance vari- ables Y1 are plotted as a group of experimental data points versus a sin- gle input variable, i.e., Y1=f (c), Y1=f(ft), Yl=f(Ts), Y1=(t), and Y1=f(Pt-p). In each of these plots the other input variables are varied over their respec- l I _' t . — 108 tive ranges described in Figure 4.1 and Table 4.1. Thus for each perfor- mance curve there is considerable scatter of the experimental data as the other four input variables are changed. However, a shaded curve is included in each plot to represent the performance boundary of the reac- tor. It is this performance boundary (or surface in the multidimensional input/ output variable space) that describes the experimental perfor- mance of the MCPR reactor. The high pressure MCPR output performance, i.e., growth rate, specific yield, and carbon conversion efficiency are summarized in the sections below. These reactor outputs are then compared with the perfor- mance of other reactors that were described in Chapter 2. U2 109 .83an 25 S conuowoc 358:098 05 .38 gamma x003 NEOE 8:389:33 ~ .e 0.59m MD— ezsaa _ _ e 83 u were 8:330 Germansuaem macaw Apmxmmvmmacmuoum Esfilsnm can wauufim N> $3“me Emu“ 1|. 1‘ llllllll I— . $293.82 8mm \ @3383 m .musmmmum _ 323.0 3365.5 SE 1 3 2.8 V _ 538.535 Sum _ a . E . > _ . . . _ A>mv m 830m 32 uwnwwmneu r .I: xucmfiuwm _ 535mm Egon \ t .l @3383th ...|. 256.8380 259.80 r I I I l l \ . mgaummfimh mumbmnsm H II 556% 33m 80 $33.83 D I. Be» “seesaw . so is cam 22m 38. a» II 3am 5380138. mu—QQESV Am_nmflm>vam\fuo\ou I Sum 5380 .825 353,: u downbgucou ~355va a Eu mam.— .fimmD 38m Acmchtmuufimfifim Acmxmmvtmnumuofima Acmxfivtmnufimfima Eu Harem .598..— tonm wficmm Ema—33m Goummm mwfim wqaooU “one; mEoD 5.830 _ _ n _ L a: 4.1 110 .Table 4.1 The Ranges of reactor variables investigated in this chapter a) Deposition Pressure, Variable: p=80-14OTorr Input Controllable b) Absorbed Microwave Power, Variable: Pt=3-5kw Variable Input 0) Methane Concentration, Variable: c=1-8% CH4/H2 U Variables, U1 d) Substrate Temperature, Variable: Ts=800-113OC c) Total Flow Rate, Variable: ft=100-14OO sccm R a) Quartz Dome, Fixed Geactor b) Water Cooling Stage Configuration, Fixed eometry c) Sliding Short Length LS~210m and Probe Depth Lp ~3cm Variables, U2 . . . . Approxrmately F1xed(at TM013 Excrtatron Mode) (1) Substrate Material and Size, Fixed: 2” dia. (100) Si D 'ti P32068818 on 21) Starting and Shut-down Procedure, Fixed Variables U b) Substrate Seeding Procedure, Fixed ’ 3 c) Deposition Time, Variable: t=4- 100 Hours Internal a) Plasma Volume, V , Approximately Constant at 144cm3 Variable d X b) Absorbed Power Density, Variable: =PtIVd Reactor a) Linear Growth Rate: 4-7 um/h Output Performance b) Total Growth Rate: 27-50 mg/h Variable Y , ’ 0) Specific Yield: 69-300 kw-h/g Y 1' (1) Gas Flow Efficiency: 0.6-1.2 mg/liter 6) Carbon Conversion Efficiency: 1-l2% Film a) Film Uniformity: 10% or better over 2" dia. area Characteristics b) Film Structural Quality: FWHM=3-8 cm'l Y2: c) Film Morphology: mostly (111) morphology (1) Film Texture: mostly <1 10> texture 4.2 Reactor Performance (Y 1)- Linear Growth Rate 4.2.1 Linear Growth Rate=f(c,ft,’r,,t) Figure 4.2 to Figure 4.5 display the linear growth rate as a function 111 of a single input variable, i.e., the growth rate is plotted as a function of the CH4/H2 concentration, c, (Figure 4.2), of total flow rate, ft , (Figure 4.3), of substrate temperature, Ts, (Figure 4.4), and of deposition time, t, (Figure 4.5). The right hand vertical axis in Figure 4.2 displays the total growth rate in mg/ h. Since all the films were reasonably uniform over the two inch diameter, the relationship between the total growth rate, W, and the linear growth rate, d, is a constant determined from d~ W/(AD)~W/ 7.1, where A is the deposition area(~20.27 cm2) and D is the diamond density (3.51 g/cm3). The relationship between these two growth rates is the same for all figures in this chapter. As shown, the shaded curves in each performance map represent the performance boundary of the reactor’s growth rate versus the specific input variables and thus defines the experimental performance of the high pressure MCPR. Instead of plotting these curves versus a single vari- able the same data points can be displayed vs. two variables. For example in Figure 4.6 the same experimental growth rate data is displayed vs. substrate temperature and methane concentration. The shaded colored curves in this figure represent an upper boundary for the maximum growth rate at either a constant T$ (red curve Ts=1080°C) or a constant c (green curve c=3%). Indeed in Figure 4.6 there exists a three dimensional surface (not shown) that represents the upper boundary of the growth rate vs. Ts or c. 112 It is interesting to compare the CH4/H2 deposition regions for the moderate pressure MCPR[9],[51] and for the high pressure MCPR. Figure 4.7 displays a comer of the “well known” Bachmann Phase Diagram for CVD diamond deposition[45]. When both reactors use only CH4/H2 chemistry the “zone” of diamond deposition occurs along the H/H+C line. In the moderate pressure MCPR. good diamond films were grown between methane concentrations of 0.6-4.75% corresponding to the H / H+C values of O.997-0.979[9],[51]. In high pressure MCPR, diamond films were grown between methane concentrations of 1-8% corresponding to the H / H+C values of 0995-0967. Figure 4.8 shows the locations of diamond deposition zones vs. three independent input variables c, ft, and TS for both high pressure and moderate pressure MCPRS. As shown, the three dimensional zone where diamond deposition occurs was increased in the high pressure MCPR. This deposition zone was increased by extending c fi'om 0.6-5% to l-8%, ft from 60-200 sccm to 400-1400 sccm, and TS from 700-lOOOC to 600 to 1 1300. This significant increase in the volume of diamond deposition zone is believed to be due to the shift of the discharge plasma chemistry from a non-equilibrium cold plasma regime to a thermal-like plasma dis- charge regime. The experimental data displayed in Figure 4.2 to Figure 4.6 also indicates that the maximum uniform film growth rate using high pressure MCPR is 6.27 um/h(or~45 mg/ h) and it is achieved at c=3%, ft=618 sccm, ls~lt lunct phok) spaci trese descr allow carb sern 113 T8~1060°C, t=100h, and p=135 Torr. Clearly film growth rate is a strong function of c, ft, TS, t, and p. The details of how the growth rate, film mor— pholog, film texture, etc. change vs. the multiple dimensional variable space can be investigated by taking cross sectional slices of the data set presented here. This output performance vs. single input variables will be described in the next chapter. The experimental data set presented here allow the calculation of reactor performance such as specific yield and carbon conversion efficiency. The results of these calculations are pre- sented in Section 4.3. 114 .58. 0389 meeABmEv Hon: 0383 ”mafia—QB . A 2. 8 1 4. 1” 2 2 1... 7 _ a q q a q . q 0. 0. 9 7. 6. A 2 1 7 2 8 1. . 5 w 4. M” 2 2 M... 7 1 . a q q . m o 1w... o i‘fooogo or ... o o f o o o o . iv o o o . 2345678910 Methane/ Hydrogen Concentration, c, (%) 1 0 200 400 600 800100012001400 p...-.L.._...Lr 85756555453525150 7 6 5 4 3 2 1 O Agivofim 526.5 Boa: . p . . b p _ . . . p P P h p O 85756555453525150 7 6 5 4 3 2 1 0 Agvamm £380 383 Figure 4.2 Summary of growth rate vs. methane/ hydrogen concen- tration in high-pressure MCPR. Total Flow Rate, ft, (sccm) Figure 4.3 Summary of growth rate vs. total flow rate in HPMCPR. 115 8 7'; _ “570 - i g 6.3 _ _500 a 5.?) __ _42.9g 8 _ _ § 4.2 _ _35.70 g 35 _ use; 5 3 ' 71.4” a 2.5 " ‘ m or 2 ‘ “14.25. E 1.5 - r 0% 0.11-3 . :7" 5 500 600 700 800 900 100011001200 Substrate Temperature(°C) Figure 4.4 Summary of linear growth rate vs. substrate temperature. o A m w A m m V .a O 0000 :8 . ON M01 Linear Growth Rate(p.m/h) ohAhmmwuahmhmhvmm a a 0 >1 War are qr 0 10 20 30 40 50 60 70 80 901001020 Deposition Time(h) Figure 4.5 Summary of linear growth rate vs. deposition time. 116 '552 -48.3 -41.4 Q g (1080C,3%,6.27umlh,43.2mg/h) 0 0 :25; E1 I i g 0 -13:8 A ' 3"”? 239% 8 Z o 8,. (9OOC,5%,0.lum/h,0.7mg/h) 6 63% 600 700 45 UBS’RAMJBOO 1100 1 2&5“ PERA 1200 8’“ Figure 4.6 Summary of growth rate vs. methane / hydrogen concentra- tion and substrate temperature in the high-pressure MCPR. Solid curve represents an upper boundary for the maximum growth rate at con- stant c=3% and dash curve presents an upper bounday for the maxi- mum growth rate at constant TS=1080°C. 117 LEGEMQ High Pressure MCPR .- as u. n Moderate Pressure MCPR 0.95 0.96 (J x 8%CH../H2 {2‘ 0.97 3, Cart ~Z‘ i C i 9’ e 00 0.98 if 0951’ .9 or“ I 0.99 4.75%CH4/H2 0.6%CH4/H2 1.0 / 1%CPL/Hz H ’llTlWlfl}llll‘llll‘IITW 0 0.01 0.02 0.03 0.04 0.05 O/O+H Figure 4.7 Locations of different methane concentrations for the experiments that were investigated in this thesis. The solid curve rep- resents the diamond growth region for the moderate pressure MCPR[9],[51] and the dash curve represents the growth region for the high pressure MCPR discribed in this thesis. 118 0 zrrs, C) (8%,1130C) HPMCPR X(c, %) (89ft, 1442 scorn) Y(ft, sccm) Figure 4.8 Location of diamond zones vs. c, ft. and T8 for high pres- sure and moderate pressurel51] MCPRS. 119 4.2.2 Linear Growth RateaflPt-p) Figure 4.9 shows the relationships between the linear growth rate and doublet input variable Pt-p and Figure 4.10 shows the linear growth rate as a function of absorbed microwave power. The experimental growth rates for the moderate pressure MCPR are also presented for compari- son[9],[51]. The data points shown in Figure 4.9 were taken under the experimental conditions when the plasma volume (V d) was held approxi- mately constant as deposition pressure (p) was increased in 70-140 Torr pressure regime. In order to keep the plasma volume constant as deposi- tion pressure increased, the absorbed microwave power (Pt) was increased. The dashed curve in Figure 4.9 presents the experimentally measured increase in the power density (=Pt/Vd) as the pressure increased. The solid curve is again an upper boundary for the growth rate versus deposition pressure for the high pressure MCPR. The dashed curve presents the experimentally measured power density as a fuction of the pressure, i.e., Pt=k1p (k1 was approximately measured to be constant). Since Pt is defined as Vd, then Pt=k1p becomes Vd=k1p, i.e., =(k1/Vd)p= k2p (k2=k1/Vd~constant). Thus the volume power density increases as the deposition pressure increases. The other important experimental power density, area power density , which is also shown on the left hand axis in Figure 4.9. This power density is defined as =Pt/A where A is the deposition area. The rela- 120 tionsz between the and is a constant determined from = (Vd/A)~7.l, where A is the deposition area(~20.27 cm2) and Vd is the plasma volume (~144 cm3). As the pressure increases from 30 Torr to 80 Torr to over 100 Torr, the discharge plasma changes from a non-equilibrium discharge to a more thermal like discharge. When the reactor is operated in the thermal plasma regime, the film growth rate increases. Figure 4.9 shows that increasing the operating pressure from the 20-80 Torr pressure regime to the 80-140 Torr regime increases the maximum deposition rate by a fac- tor of five. Also as shown in Figure 4.9 for a constant pressure the growth rate data displays a considerable variation. For example at the thermal-er plasma regime (p=80-140 Torr), the growth rate varies between a maxima of 6.5 um/h(~43 mg/h) to a minima of 0.5 um/h(~3 mg/h). This variation indicates that at a constant pressure the growth rate is strong function of many input variables such as methane concentration (c), substrate tem- perature (Ts), total flow rate (ft), etc., and varies considerably as these input variables are changed. It is interesting to compare the growth rates and the associated area power densities of the low pressure and the high pressure MCPRS in this section. The low pressure MCPR operated with an area power density of 0.03- 0.04 kW/cm2 [51] which resulted in the maximum growth rate of ~O.55 urn/h. The area power densities for the high pressure MCPR. as 121 shown in Figure 4.9, are mostly in the range of 0.15 and 0.25 kW/cmz. The area power density of high pressure MCPR is almost 10 times that of the low pressure MCPR. Interestingly, the linear growth rates of the high pressure MCPR are also approximately ten times the linear growth rates of the low pressure MCPR. Thus the linear growth rate, i.e., the deposition per substrate area, increases directly as the discharge area power density increases. 122 908— 0.3an ”m8 ABE—d 454 0 +--- Em: 3825.. Hana—Eh?“ e m 23.3 $8.26 382ch 5 m w ”mmmlgmaml llllllllll .v I I I Holel I H I u r m a 7.8388 5635..“ 1 w... w... =o=-3cEancB + o . w. 338» wamgomoh: + 89 o o W Av + 0 + 1 On av 88 + l D... 0 + m x. IIIIIIIIIIIIIIIIIII H I M I 1 - mo ow has 3828 Emma» wmmsfi 33 I < l 5 O 5 O 5 O 5 O 5 4 4 3 3 2 2 1 1 AmEobE A>mv Samson 830m 053:5 n.o _ ow :w J _ J4 _ _ _ n 1 3. m 2 2 2 m 1. n m m 0 O. O O. O 0. 0 O O . O @8335 A orientation and (1 l 1) morphologt with approximate thickness of 7 mm were routinely produced using high-pressure MCPR system. This recipe was derived from the experimental input parameters that produced the best growth rate, i.e., the experimental input variables were pressure=l35 Torr(Figure 4.9), CH4/ H2=3%(18/ 600 sccm) from Figure 4.2, total flow rate~600 sccm(Figure 4.3), substrate temperature = 1060C (Figure 4.4), deposition time = 100h (Figure 4.5), absorbed microwave power~3.5 kW(Figure 4.9). After synthesis, a free standing diamond film was produced by back etch- ing the silicon using 50%HNO3-25%HF-25%CH3COOH solution. An example of the surface morphology of these films is displayed in Figure 5.21“). 4.6 Summary Using high pressure MCPR microwave plasma assisted diamond thin film synthesis was experimentally investigated with CH4/H2 gas 134 mixtures at 80-140Torr. Some of the important performance data is plotted in Figure 2. 19, and Figure 2.20 and in Table 2.1 where it can be directly compared with the published performance of other reactors. The performance of the high pressure MCPR can be summarized as follows: (1) Optimum diamond film growth conditions of the high-pressure MCPR are CH4/H2=3-4%, ft=600-700sccm, TS=1060-1100C, p=120- 135Torr, and =30W/cm3. The reactor performance (Y 1) under these conditions are: (a) linear growth rate: 6.27 m/h, (b) total growth rate=44.68 mg/h, (c) specific yield: 69 kW-h / g ((1) Gas flow efficiency: 0.08 mg/ liter (e) Carbon conversion efficiency=7.7%. (2) The low pressure MCPR reactor[9],[5l] operates with an area power density of 0.03-0.04 kW/cm2, resulting in maximum linear deposi- tion rates of approximately 0.55 um/ h. The area power densities for the high pressure MCPR is 0.15- 0.25 kW/cm2, i.e., almost ten times the area power density of the low pressure MCPR. Interestingly, the maximum lin- ear growth rates of the high pressure MCPR are also approximately ten times the linear growth rates of the low pressure MCPR. Thus the linear growth rate, i.e., the deposition per substrate area, increases directly as the area power density increases. 135 (3) A comparison of the specific yield of the moderate pressure MCPR and high pressure MCPR indicates that the energy efficiency improves considerably from 163 kW-h / g to 69 kW-h/ g as the operating pressure increases, i.e., a thermal like high pressure microwave discharge improves the specific yield by a factor of two and one half. (4) The volume in the input variable space when the diamond deposition occurs was increased considerably by changing the operating pressure of the MCPRS from a 20-80 Torr non-equilibrium plasma regime, to a 80-140 Torr thermal-like plasma regime. (5) Carbon conversion efficiency was not significantly affected by the shift of operating the MCPRS from a non-equilibrium plasma regime to a thermal-like plasma regime. CHAPTER FIVE Reactor Experimental Output Performance- Film Characteristics (Y2) 5. 1 Introduction Chapter 4 presented an overview of the high pressure MCPR’s output reactor performance Y1 versus five dimensional input variable space, i.e., Y1: f (0, ft, Ts, t, Pt-p). As pointed out in the summary of Chapter 4, there exists an optimum diamond growth zone in this five dimensional input variable space, i.e., the maximum diamond deposition rate takes place under the experimental input conditions of c~3%, ft~600 sccm, Ts~1080°C, Pt~3.3 kW, and p~135 Torr. This optimum diamond growth zone can be visualized versus the two dimensional input variable space shown in Figure 4.6. In this chapter the reactor output performance Y2 (referred as the diamond film property) is evaluated by taking cross sectional experimental data slices from these multidimensional plots. In particular, a series of experiments were performed by observing the change in output film properties Y2 versus changes in one independent input variable. Experiments are varied around the maximum deposition conditions shown in Figure 4.6. 136 137 Independent experimental variables U1 such as methane concen- tration, total flow rate, and substrate temperature are the selected experi- mental input variables. The absorbed microwave power and pressure were fixed at the optimized conditions of 3.3 kW and 135 Torr, respec- tively. Reactor geometry variables U2 were held constant and deposition process variables U3 were also held constant except the deposition time, which is considered as the fourth input variable in this chapter. The plasma volume, Vd, was held approximately constant at 144 cm3, while the volume power density as described earlier in Figure 4.9, also remains constant, i.e., =Pt/Vd~23 W/cm3. Film output properties Y2 that were investigated include film morphology, film texture, and film struc- tural quality. Thus output film properties Y2 were investigated as a func- tion of methane concentration, total flow rate, substrate temperature, and deposition time, i.e., Y2 =i(c, ft, Ts, t). The first set of experiments, described in Section 5.2, contain eight separate experimental runs using different methane concentrations (c=1- 8%) while holding other independent experimental variables U1 constant such as ft~600 sccm, TS~1060°C, Pt: 3.22 kW, and p=135 Torr. Reactor geometry variables U2 and deposition process variables U3 are held con- stant (deposition time =5h) for these eight experimental runs. Thus Y2 is investigated as a function of methane concentration, i.e., Y2=flc). The sec- ond set of experiments, described in Section 5.3, contain six separate 138 experimental runs using different total flow rates (ft: 400, 600, 800, 1000, 1200, 1400 sccm) while holding other independent experimental variables U1 constant such as c=3%, Ts~1080°C, Pt: 3.22 kW, and p=135 Torr. Reactor geometry variables U2 and deposition process variables U3 (deposition time=10h) are held constant for these six experimental runs. Thus Y2 is investigated as a function of total flow rates, i.e., Y2=f(ft). The third set of experiments, described in Section 5.4, contain six separate experimental runs using different substrate temperatures (Ts=950, 1025, 1050, 1075, 1 100, l 125 °C) by varying the number of insulation discs in water-cooled substrate holder setup (see Section 3. 6) while holding other reactor geometry variables 0; such as the quartz dome size and the sub- strate material constant. Independent experimental variables U1 such as c, ft, Pt, and p are held constant at 3%, 618 sccm, 3.22kW, and 135 Torr, respectively. Deposition process variables U3 were also held constant (deposition time=10h) for these six experiments. Thus Y2 is investigated as function of substrate temperature, i.e., Y2=f(TS). The fourth set of experiments, described in Section 5.5, contain six separate experimental runs using different deposition time (t= 5, 10, 20, 30, 55, 100 h) while holding other deposition process variables 03 such as starting and shut- down procedures and seeding procedures constant. Independent experi- mental variables Ul such as 0, ft, Ts, Pt, and p are held constant at 3%, 618 sccm, 3.22kW,1080°C, and 135 Torr, respectively. Reactor geometry 139 variables U2 were held constant. Thus Y2 is investigated as function of deposition times, i.e., Y2=f(t). 5.2 Effect of. Methane Concentration (c) 5.2.1 Introduction This section presents an experimental investigation of the film out- put properties Y2 as a function of methane concentration. Eight experi- ments were performed by starting at optimized film growth conditions and then carefully varying methane concentrations from 1% to 8%. This section first presents the experimental conditions for the eight experi- ments that were investigated and then the influence of methane concen- tration on film growth rate, film morphology, film texture, and film structural quality is described. 5.2.2 Experimental Variables (A) Input variables(U): (1) Controllable input variables (U1): a) Gas mixture: Variable, c=l% to 8%, b) Total flow rate~600 sccm, c) Substrate temperature=1058°C, d) Absorbed microwave power=3.22 kW, e) Deposition pressure=135 Torr (18 KPa), 5.2.3 140 (2) Reactor geometry variablewz): a) Quartz dome size: 5" in diameter, b Water cooling state configuration: fixed (see Section 3.6.3.5) c) Substrate: 2” (100) silicon substrate (3) Deposition process variables(U3): a) Deposition time=5 h, b) Substrate seeding procedure: fixed(see Section 3.3.1) c) Start-up and shut-down procedures: fixed (see Section 3.3.2), (B) Internal variables(X): (1) Power Density: 23 W/cm3, (2) Area Power Density: 0.16 kW/cmz, (3) Plasma volume~ 144 cm3. Growth Rate=f(c) The dependence of diamond film weight gain and growth rate on methane concentration is shown in Figure 5.1. As might be expected, the growth rate increases with increasing methane flow rate and drops at high methane concentrations (>3.5%). The possible explanation is that as CH4 concentration is increased for a constant hydrogen flow rate, the number of atomic hydrogen is no longer sufficient to etch sp2 graphite. 141 Thus the high-order growth of diamond can not be continued on the layer of graphite, resulting in a lower growth rate and the formation the black cauliflower -like microcrystalline gains. 200 190 180 170 A160 ““150 S140 423 E 110 £100 r (390. H .380 13 7o 3 60 so 40 30 20 1o 0 %CH4/H2 (%) 142 0 1 l l l l l 1 1 l 5.80 5.51 5.22 4.93 4.64 4.35 4.06 3.77 3.48 3.19 2.90 2.61 H- (D 2.32 2.03 \ 1.74 E: 1.45 1.16 0.87 0.58 0.29 mean tumors 4 6 8101214161820222426283032343638404244464850 CH4 Rlow Rate (sccm) Figure 5. 1 Linear growth rates (weight gain) versus methane concen- trations. The experimental conditions are shown in Section 5.2.2 143 5.2.4 Film Morphologye=f(c) The Scanning Electron Microscopy photographs of the films deposited for 5 hours using different methane concentrations, c, in hydrogen are shown in Figure 5.2(a) — 4(h). At c=1% [see Fig.5.2(a)], only small diamond crystals were formed. These crystals have (100) and (l 10) cubo-octahedron morphologies. The discontinuous diamond films that were obtained at c=1% deposition condition, indicate the characteristic source gas boundary (lowest methane concentration) for the high pressure MCPR. When c=2% [see Fig.5.2(b)], the film surfaces consist of densely populated diamond grains. As shown in Fig.5.2(b), the diamond films consist of well-faceted grains having square (100) faces and some (1 10) roof-like facets. When c=3%[see Fig.5.2(c)], the film surfaces show both (1 10) roof-like and (l l 1) triangular morphologies. At c=4%[see Fig.5.2(d)] (l 10) roof-like facets become less and (1 l l) triangular facets dominate. When c=5%, the films completely exhibit (1 l l) triangular morphologies. For 6% < c <8%, the film became microcrystalline and there is no crystallographic planes of diamond observed. The surface of Fig.5.2(f) to (h) has the so-called “cauliflower” structure. — 10p 1 I — 1011 Figure 5.2(a) c=1% Figure 5.2(b) c=2% — 1011 _ 10.. Figure 5.2(0) c=3% Figure 5,2(d) c=4% 145 Figure 5.2(g) c=7% Figure 5.2(h) c=8% Figure 5.2(a)-(h) Surface morphology vs. %CH4/H2 varied from 1% to 8%. 146 5.2.5 Film Texture=flc) The resulting XRD spectra of different methane concentrations are shown in Figure 5.3(a) to (g). Figure 5.4 shows the ratio of intensities of <220> and <1 1 1> peaks from XRD. The positions of the lines at 26~43.9° and 29~75.3° are the indication of the <1 1 l> film texture and <220> film texture. In order to analyze the degree of diamond grain orientation, the intensity of each line was recorded from the original computer control system with the same sensitivity of x-ray detection for all spectra. At low methane concentrations of c=1% and 2%, the I(220) / I(l l 1) ratio is less than one, indicating that the films have no preferred film texture. As methane concentration increased to c=4% and 5%, the I(220) / 1(1 1 1) ratio becomes larger than one, indicating that the degree of <1 10> orientation was further enhanced and the films were textured in the <1 10> direction. When 6% < c < 8%, the I(220) / I(l l 1) ratio decreases to the value of 0.4, indicating the presence of the cauliflower films showing no preferred fihn texture. 147 l l I I I I I l l l l I l I l [ Diamond(111) Silicon(111) 350000 ' ' ( )c-8°/ Diamndmo) Diamond(311) 1.9.; -1: _ w11L - a. LBEWMEOO)- _ 300000 - ‘ A (1) 0:770 g -u42-4L ~44 A44 44» A O - O ‘ in : Art .2. w“ 3 :‘ ‘L‘! 3 250000 ‘ B Q. g 4 2L.- _ _ ~ 0 9. 200000 ' .4? - A. - 4+ - - ~— (I: 5 E q, 150000 - ' E 5.." 6’2 100000 ' ' Jersar/o..- 1.- it. - .. 50000 - ' (a) c=2% J) 0 - J 1 JLA. l 2--.]- L 1.; l i . 12 1 l .1 40 45 50 55 60 65 70 75 80 85 90 95100105110115120125 29 (Degree) Figure 5.3 The X-ray diffraction spectra as a function of methane concen- trations of (a) 2%, (b) 3%, (c) 4%, (d) 5%, (e) 6%, (f) 7% and (g) 8%. Maxi- mum ratio of I(220)/[(1 1 l) was observed at c=4% and 5%. They are summarized in Figure 5.4. 148 r 2.8 2.6 - 2.2 - 1.8 - 1.6 - 1.4 1.2 I (220) /I (111) I l 0.8 - 0.6 B - 0.4 0.2 - a a 1 012345678 %CH4/I-12 (%) Figure 5.4 The ratio of the XRD peak height of (220) to (l l 1) versus the methane concentration. The experimental conditions are shown in Section 5.2.2. 149 5.2.6 Film Structural Qualityaflc) 'lypical Raman spectra of the diamond films deposited on Si(100) at eight different concentrations of methane, ranging from 1% to 8%, are shown in Figure 5.5. The Raman spectra shown in Figure 5.5 exhibit three main features: (1) the appearance of the peak at 1332 cm'1 from 1% to 8%, which is the characteristic line of crystalline diamond, (2) the appearance of a broad peak centered at about 1580 cm’1 for the methane concentration above 6%, which is characteristic of polycrystalline graphite or amorphous carbon with graphite bonding, (3) the 6% and 8% films show a broad peak at about 1280 cm"1 whose origin is not yet fully understood. For 6% < c <8%, the films exhibit a relatively low diamond peak at 1332 cm'1 and relatively strong contribution from nondiamond carbon. The films that contain certain amount of nondiamond carbon exhibit the cauliflower morphology(shown in Figure 5.2(fl-(h)). Compared to 6% < 0 <8%, the Raman spectra of the 1% < c< 5% films display a sharp diamond peak at 1332 cm'1 and no contribution from nondiamond carbon. The films show well-faceted crystalline diamond morphologies as shown in Figure 5.2(a)-(e). Figure 5.6 displays the full width at half maximum (FWHM) of the 1332 cm’1 peak as a function of methane concentration. A FWHM of 2.0 cm"1 is plotted as a reference for natural diamond. As seen in Figure 5.6, the FWHM decreases from 17.43 cm’1 for the c=l% to 8.77 cm"1 for c=2% 150 and remains approximately constant up to 5% methane concentration. 80000 I I 1 1 l I I (h)8% CH4/H2 70000 ' (9)770 CH4/H2 (06% CH4/H2 50000 - 40000 - (9)50/0 CH4/H2 30000 - (d)4% CH4/H2 Relative Intensity( Count per second) 20000 ' (c)3% CH4/Hz ‘ 10000 - - (a)1% CH4/H2 l l l l l l 0 l 1200 1250 1300 1350 1400 1450 1500 1550 1600 Wavenumber ( cm'l) Figure 5.5 Raman spectra of various methane concentrations.The Experimental conditions are shown in Section 5.2.2. 151 The FWHM then increased abruptly from 8.42 cm“1 for c=5% to 72.14 cm’1 for c=8%. 75 T I T I I r I l 0 70 '- ~ 65 " .4 60 ~ 1 55 r '1 A 50 b O .1 '8 45 r 1 e... 40 - i E 35 - 4 25 ’- J 20 r- 1 o 15 F ° - 1° " o 0 ° 0 1 5 r a 0 l l l l l l l l 0 1 2 3 4 5 6 7 8 9 o/OCH4/H2 Figure 5.6 FWHM of Raman spectra as a function 01" {Jig/1‘42 concentration. The experimental conditions are shown in Sec tion 5.2.9.. 152 5.3 Effect of Total Flow Rate (ft) 5.3. 1 Introduction This section presents an experimental investigation of the film out- put properties Y2 as a function of total flow rate. Six experiments were performed by varying the flow rate from 412 sccm to 1442 sccm around optimized deposition conditions. The section first describes the experi- mental conditions for these six experiments and then the variation of film growth rate, film morphology, film texture. and film structural quality ver- sus total flow rates are described in the later section. 5.3.2 Experimental Variables (A) Input variables(U): (l) Controllable input variables (U1): a) Gas mixture: 3%Cl—I4/H2, b) Total flow rate: Variable from 412-1442 eccm. c) Substrate temperature=1080°C, d) Absorbed microwave power=3.22 kW, e) Depositionpressure=135 Torr (18 KPa), [2) Reactor geometry variable(U2): a) Quartz dome size: 5" in diameter, b Water cooling state configuration: fixed (see Section 3.6.3.5 153 c) Substrate: 2” (100) silicon substrate (3) Deposition process variables(U3): a) Deposition time=10 h, b) Substrate seeding procedure: fixed(see Section 3.3.1), c) Start-up and shut-down procedures: fixed (see Section 3.3.2), (B) Internal variables(X): (1) Power Density: 23 W/cm3, (2) Area Power Density: 0.16 kW/cmz, (3) Plasma volume~l44 cm3. 5.3.3 Growth Rate=f(£t) The effect of the total flow rate (denoted as ft) on film growth rate is shown in Figure 5.7. The experimental conditions for different flow rate (CH4/H2=12/400, 18/600, 24/800, 30/1000,36/1200 and 42/1400 sccm) are shown in Section 5.4.2. It is noted that CH4/H2 concentration is kept at 3% for different flow rates. As shown in Figure 5.7, the growth rate increased from 3.92 um/ h at CH4/ H2=12 / 400 sccm and increases to a maximum of 5 um/ h at CH4/H2=30/ 1000 sccm and then decreases to a relatively low growth rate of 2.61 um/h at CH4/H2=42/ 1400 sccm. Therefore growth rate of the diamond film is affected by flow rate while CH4/H2 concentration, pressure, substrate temperature, microwave 154 input power and deposition time remain unchanged. 50 l I l l ' 7.25 45 — -6.53 40 ~ ~5.80 E 35 #- ° 0 0 '5.08 g \ no 31 5 30 - ° «4.35 S? {3‘ . § (5 - 4 3 63 3‘3) 25 '33 20 - -2.90 3’: a . g 15 - -2.18 \ E". 10 - -1.45 5 ~ 10.73 0 l 4 l l l 400 600 800 1000 1200 1400 Total Flow Rate(sccm) Figure 5.7 Linear growth rate as a function of total flow rate at fixed methane / hydrogen concenctration(3%). 5.3.4 Film Morphology=flft) The film morphologies that resulted as the total flow rate was varied are shown in Figure 5.8(a)-(f). As shown diamond films exhibit mixed (1 10) roof-like and (l 1 1) triangular morphologies when ft=412 sccm(CH4/ 155 H2: 12/400). 618 sccm(CH4/H2=18/600 sccm), and 824 sccm (CH4/ H2=24/ 800 sccm) are used. As ft increases to 1030 sccm (CH4/H2 =30/ 1000 sccm) the film exhibits completely (1 1 1) triangular morphology. When flow rate further increases to 1236 sccm(CH4/H2 =36/1200 sccm), (1 10) roof-like facets are again observed in the film. As flow rate increases to 1442 sccm(CH4/H2=42/ 1400 sccm), the film shows (100), (110) and (l 1 l) morphologies. It is noted that the maximum growth rate again occurs when the film completely exhibits (1 1 l) triangular morphologr at flow rate of CH4/H2 =30/ 1000 sccm. Also (100) facets are observed in the film that produced at CH4/H2 =42/ 1400 sccm, which has the minimum film growth rate. 156 .1“ Figure 5.8(a) ft=412 sccm Figure 5.8(b) ft=618 sccm lOu Figure 5.8(c) ft=824 sccm Figure 5.8(d) ft=10305ccm 157 : — — Figure 5.8(e)ft=1236 sccm Figure 5.8(f) ft=412 sccm Figure 5.8(a)-(f) Film morphologies vs. total flow rates. The experimental conditions are shown in Section 5.3.2. 158 5.3.5 film Texture=flft) The XRD spectra of diamond films synthesized by different total flow rates are shown in Figure 5.9. Ratios of I(220) to [(1 1 l) for different ft are shown in Figure 5.10. As shown in Figure 5. 10 the relative height of I(220) with respect to H] l 1) reaches its maximum at ft=618 sccm (CH4/ H2 =18/ 600) .The films synthesized for 10 h under the conditions shown in Section 5.4.2 show no preferred textures at ft=412 sccm (CH4/H2 =12/ 400) and ft=1442 sccm(CH4/H2 =42/ 1400 sccm) and exhibit <110> texture at ft=618 sccm (CH4/H2=18/600), 824 sccm (CH4/H2 =24/ 800), 1236 sccm (CH4/H2 =36/ 1200 sccm). 159 450000IIIIIIIIIIIIIIIj Diamond(111) 400000 - Diamond(zzo) (f) CH / H -42/1400 sccm. 350000 LJL “My 1 A4 2 A # 300000 A (e) CH4/H2 =36/ 1200 scorn: r if I I:— I 1 L 1 J 1 Relative Intensity (Count per second) 250000 0 (f) CH4/H2 =30/ 1000 sccm 200000 ’ 150000 ~ 4 (c) CH4/H2 -24/ 800 sccm .a-J ML n. “Li L 100000 - - I (b) CH4/H2 =18/600 sccm 50000»J “ “ “ - 1 ’L (a) CH4/H2 =12/400 sccm 0 l L l I l J l 40 45 50 55 60 65 70 75 80 85 90 95100105110115120125 29 (Degree) Figure 5.9 XRD spectra of diamond films synthesized by CH4/H2 flow rates of (a)l2/ 400 sccm, (b)18/ 600 sccm, (c) 24/800 sccm, ((1) 30/1000 sccm, (e) 36/1200 sccm and (f) 42/1400 sccm, using the experimental conditions described in Section 5.3.2. 160 3 I l l l l I l l l I l 2.8 ' 2.6 - 2.4 '- 22 - 2 _ a 1.8 - 1.6 - 1.4 - E] 1.2 - 1 . 0.8 - 0.6 ' B 0.4 - 0 0.2 - a O I l l l l l l l l l l 350 450 550 650 750 850 950 1050 1150 1250 1350 1450 Total FLow Rate (sccm) I (220) / I (111) Figure 5.10 The XRD height of (220) relative to that of (l 1 1) as function of total flow rate in microwave plasma cavity system. The experimental con- ditions for these experiments are described in Section 5.3.2. 161 5.3.6 Film Structural Quanty=iug Raman spectra of diamond film produced by different flow rates are displayed in Figure 5.1 1. The FWHM of these Raman spectra are in Figure 5.12. As shown in Figure 5.12, FWHM remains approximately constant at 5.85 cm“1 for flow rates of ft=412, 618, 824 sccm and decreases to 4.79 cm‘1 at ft=1236 sccm and further decreases to 3.50 cm'1 at ft=1442 sccm, which has the FWHM that is very close to that of nature diamond, i.e., 2.3 CHI-1. 162 (03%CH4/ H2=42/ 14005ccm (e)3%CH4/H2=36/ 1200sccm 'AM (d)3%CH4/H2=30/ IOOOSccm Relative Intensity(Count per second) (c)3%CH4/ H2=24 / 8005ccm (b)3%CH4/ H2: 1 8/600sccm (a)3%C 4/H2=36/ 400sccm I l l l l J l 1220 1260 1300 1340 1380 1420 1460 1500 1540 1580 Wavenumber(cm‘ 1) Figure 5.1 1 The effect of total flow rate on Raman spectrum. The experimental conditions are listed in Section 5.3.2. 163 10 I I I I I I I I I I I FWHM (cm'l) 0 FWHM of natural diamond=2.3cm" 0 I I I I I l I I I I l 350 450 550 650 750 850 950 1050 1150 1250 1350 1450 Total Flow Rate(sccm) Figure 5.12 The effect of total flow rate on FWHM of Raman spectrum. The experimental conditions are shown in Section 5.3.2. 164 5.4 Effect of Substrate Temperature (1}) 5.4. 1 Introduction The six experiments presented in this section were performed by varying TS from 950 to 1 125°C while holding the other experimental variables constant.This section first describes the experimental conditions for the six experiments that were investigated and then the influence of substrate temperatures on film growth rate, film morphology, film texture, and film structural quality are presented in the following section. 5.4.2 Experimental Variables (A) Input variables(U): (1) Controllable input variables (U1): a) Gas mixture: 3%CH4/H2(18/600 sccm), b) Total flow rate: 618 sccm, c) Substrate temperature: Variable horn 950-1 125°C, (1) Absorbed microwave power=3.22 kW, e) Deposition pressure=135 Torr (18 KPa), (2) Reactor geometry variable(U2): a) Quartz dome size: 5" in diameter, b Water cooling state configuration: fixed (see Section 3.6.3.5 165 c) Substrate: 2" (100) silicon substrate (3) Deposition process variables(U3): a) Deposition time=10 h, b) Substrate seeding procedure: fixed(see Section 3.3.1) c) Start-up and shut-down procedures: fixed (see Section 3.3.2), (B) Internal variables(X): (1) Power Density: 23 W/cm3, (2) Area Power Density: 0. 16 kW/cmz, (3) Plasma volume~ 144 cm3, 5.4.3 Growth Rate=fITs) The growth rate as a function of substrate temperatures is shown in Figure 5.13. The rest of experimental conditions are shown in Section 5.4.2. A water cooling stage, as shown in Figure 3.12, is utilized to control the substrate temperature by adding or removing the graphite holder layers. Therefore the substrate temperature can be independently studied without changing the experimental input variables such as deposition pressure, MW input power, substrate position which could also influence the substrate temperature. Experiments with substrate temperatures of 950C, 1025C, 1050C, 1075C, 1 100C and 1 125C were performed at c=3% while holding the other experimental variable approximately constant. As shown, growth rate increases from Ts=950C and reaches a maximum at Ts: Weight Gain in 10 h( mg) 166 1 100C and then decreases with increasing Ts. 400 , . r , 1 5.80 375 ' 3% CH4/H2 +— ‘ 5-44 350 . . 5.08 325 - - 4.71 300 _ - 4.35 275 - - 3.99 250 - . 3.63 225 - - 3.26 200 - - 2.90 175 - - 2.54 150 _ - 2.18 125 - - 1.81 100 - . 1.45 75 - - 1.09 50 - . 0.73 25 - 40.36 0 . . . . . 900 950 1 000 1050 1 100 1 150 Substrate Temperature (°C) Figure 5.13 The effect of Ts on the film growth rate. (II/uni) emu ammo mean 167 5.4.4 Film Morphologyd'fl‘g The film morphologies inspected by the SEM were shown in Figure 5.14(a) -(fl. The experimental conditions are shown in Section 5.4.2. As shown, the film exhibits both (100)square and (1 10) roof-like morphologies at TS=950C and then shows mostly (1 1 1) triangular and some (1 10) roof-like facets at Ts=1025, 1050C and 1075C. When Ts increases to l 100C the film exhibits only (1 1 1) triangular facets. Then when Ts increases to 1 125C, the film mostly exhibits (l l l) triangular facets and begins to shown the (1 10) roof-like morpholog'. Since (1 1 1) triangular faces are highly defective, the structural defects of (l 1 l) triangular faces will promote nucleation and growth of diamond[56],[65]. It is believed that (1 l l) triangular facets are more favored under high growth rate conditions than (1 10) roof-like or (100) square facets. It is noted that crystal size also increases as substrate temperature increases. 168 _'_10u 3 Figure 5.14(a) Ts=950C Figure 5.14(b) Ts=10250 —1011 7 I" - I _1ou - Figure 5.14(c) Ts=10500 Figure 5.14(a) Ts=1075C 169 — 1011 — 1014 Figure 5.14(e) Ts=1 1 10C Figure 5.14m Ts=1 1250 Figure 5.14(a)-(f) Film morphologies vs. substrate temperatures. The experimental conditions are described in Section 5.4.2. 170 5.4.5 Film Texture=ftT5) The X-ray diffraction spectra for different substrate temperatures are shown in Figure 5.15(a)-(f). The comparison of the intensity ratio of <220> to <1 1 1> for the diamond films deposited at different substrate temperate is shown in Figure 5.16. As seen in Figure 5.16, the I(220)/ Hi 10) ratio increased from 0.3 (at Ts=10250) to 0.78 (at TS=1050C) and 0.8 for T8: 1075C. The I(220)/I(1 10) ratio shows a maximum value of 1.6 at Ts=l 100C. This indicates that the film texture in the direction of <1 10> was enhanced as substrate temperature increased to l 100C and then suppressed with further increase in the substrate temperature. 171 I I I I I I Dilarnorl‘d(2éo) I I I I I I I Diamond(111) 300000 ' Silicon(111) ' #0191125? 21.11 _ . 033T”). . - 1 - 250000 ' ‘ “g (e) 1100c L211“ 1. _ 1.111 ‘ A -_ 200000 - - I-I Q) 5‘ (d) 1075 c g 2... x A L_._ #4“ hhhxuh 5 1 7' 150000 ' I ‘ 3‘ . "(3 I (C) 1050 C I g -M 21-- 1.1- . Ax.“ m_...mh- E: i: E 100000 ' - 84’ ___JLIbI}O?5_E -J _ i -- - _1 50000 I ‘ (a) 950 C II I 0 40 45 50 55 60 65 70 75 80 85 90 95100105110115120125 20 (Degree) Figure 5.15 X-ray diffraction s ectra of diamond films synthesized for (a) 9 0C. (1)) 1025C, (c) 1050C. (d 1075C. (e) 1100C and (f) 1125C. Maxi- mum ratio of I(220)/I(1 1 l) was observed at (e) l 100C. They are summa- rized in Figure 5.16. 172 2 I I I I I I I I I I 1.9 - 1.8 - 1.7 - 1.6 - 1.5 - 1.4 - 1.3 - 1.2 - 1.1 - 1 . 0.9 - 0.8- G 0.7 - 0.6 - 0.5 ~ 0.4 - 0.3 ' a 0.2 ' n 0.1 - 0 I I I l l I I I I I 900 925 950 975 1000 1025 1050 1075 1100 1125 1150 1175 Substrate Temperature (°C) I (220) / I (111) Figure 5.16 The XRD peak height of (220) relative to that of (l l l) as a function of substrate temperature. The experimental conditions for these experiments are shown in Section 5.4.2. 173 5.4.6 Film Structural Quality=f(T5) The FWHM of diamond peak lines from Raman spectra using different substrate temperatures are displayed in Figure 5.17. The Raman spectra are shown in Figure 5.18. As shown in Figure 5.17, the FWHM is not affected significantly by the substrate temperature. As seen in Figure 5.18, the strong diamond peak at 1332 cm‘1 is observed and the sp2 graphite at 1580 cm"1 and nondiamond carbon background are not observed fiom these Raman spectra. 10 I I I I T I '- q 0’1 1- — FWHM (cm'l) )- 2.3 cm“ FWHM of natural diamond P 10 Q h 1 - _ J l l J o l l 1 l l I 900 925 950 975 1000 1025 1050 1075 1100 1125 1150 1175 Substrate Temperature(°C) Figure 5.17 The effect of substrate temperatures on FWHM of diamond peak lines from Raman spectra. Their Raman spectra are shown in Figure 5.18. 174 70000 - (f) 11250 60000 - (e) 1100C ’6 1: Se; 50000 P g (CI) 10750 E 40000 - 8 (c) 1050C 9.3.“ 8 13 30000 - g (b) 1025 0 B “ 20000 - 10000 - (a) 9250 0 I I I I I I I 1200 1250 1300 1350 1400 1450 1500 1550 1600 Wavenumber ( cm‘l) Figure 5.18 The effect of substrate temperature on Raman spectra. The experimental conditions are listed in Section 5.4.2. 175 5.5 Effect of Deposition Time (t) 5.5.1 Introduction The six experiments presented in this section were performed by varying the deposition time, t, from 5 to 100 hours while holding the other experimental variables constant.This section first presents the experimental conditions for six experiments and then influence of the deposition time on film growth rate, film morphology. film texture, and film structural quality are presented. 5.5.2 Experimental Variables (A) Input variables(U): (1) Controllable input variables (U1): a) Gas mixture: 3%CH4/ H2(18/ 600 sccm) b) Total flow rate: 618 sccm. c) Substrate temperature: 1080°C d) Absorbed microwave power=3.22 kW, 0) Deposition pressure=135 Torr (18 KPa), (2) Reactor geometry variable(U2): a) Quartz dome size: 5” in diameter, b Water cooling state configuration: fixed (see Section 3.6.3.5 c) Substrate: 2" (100) silicon substrate 176 (3) Deposition process variables(U3): a) Deposition time: Variable from 5-100 h, b) Substrate seeding procedure: fixed(see Section 3.3. l) c) Start-up and shut-down procedures: fixed (see Section 3.3.2), (B) Internal variables(X): (1) Power Density: 23 W/cm3, (2) Area Power Density: 0.16 kW/cmz, (3) Plasma volume~ 144 cm3, 5.5.3 Growth Rate=f(t) The experiments performed in this section have identical experi- mental input variables except that different deposition times (denoted as t) were used. The experimental conditions and the effect of deposition time on film growth rate are shown in Section 5.5.2 and Figure 5.19, respectively. As shown in Figure 5.19, the film growth rate slightly increases as deposition time increases. It is believed that the slowly increasing growth rate with the increasing deposition time is attributed to the success of cyclically higher-order diamond growth[3]. 177 co _1 .1 -1 .1 .1 .. .. _1 I 3% CH4/H2 0 _. 7‘ N 9° P 9' 9’ .‘1 mammmwmbmmmmmum I [III I [11 Linear Growth Rate (um / h) C I I I I I I I I I I I 0 10 20 30 40 50 60 70 80 90 100 110 12 Deposition Time(h) O Figure 5.19 Linear growth rate as a function of deposition time. The experimental conditions are described in Section 5.5.2. 5.5.4 Film Morphology=fm The film morphologies produced from different deposition times are shown in Figure 5.20(a)-(f). At t=5h, the fihn exhibits (110) roof-like and some (1 l 1) triangular morphologies. These morphologies are also shown at t=10h. At t=20h the film mostly develops the (l l 1) triangular facets and the number of (l 10) roof-like facets has decreased. At t 30h, the films completely exhibit (1 1 l) triangular morphologies. On the other hand, no (1 10) roof-like facet is observed when t=30h. The average of dia- 178 mond grain size increases from 8mm at t=5h to 30mm at t=30h and then reaches its maximum of more than 250 pm at t=100h. It is noted that 630 um film thickness with 250 um crystal sizes are synthesized in the MCPR operating at the pressure of 135 Torr. — 1011 — 1011 Figure 5.20(a) t=5h Figure 5.20(b) t=10h 179 10011 ' — 10011 Figure 5.20e) t=55h Figure 5.20“) t=100h Figure 5.20 (a)-(f) Film morphologt vs. deposition time. The experimental conditions are shown in Section 5.5.2. 180 5.5.5 Film Textured“) X-ray diffraction spectra of diamond film synthesized for 5h to lOOh are shown in Figure 5.21. As shown the peak height in the <11 1> direction, which has 20:43.90, and in the <1 10> direction, which has 20:75.30, vary with the deposition time. At t=5h and 10h ratio of I(220) to 1(1 1 1) is less than 1, indicating that the films do not have preferred orientation. As film is synthesized longer the ratio of I(220) to I(l l 1) increases. At t 30h the peak height of <1 1 l> direction becomes very small while that of <1 10> becomes larger. This indicates that the films which were synthesized longer than 30h are highly textured in the direction of <1 10>. The XRD height of (220) relative to that of (1 l l) as function of deposition time is shown in Figure 5.22. 181 3e+06 - - 2.Se+06 - . 28+06 r - 1.56+06 - -* (fl t=100h “-13 1e+06 - . . JL g - g j o e t=55h 3 500000 ~ ( ) - - 1.. o _ Q 0 AI (dl) t-I3OP I I A I I I I I I l l H g 40 45 50 55 60 65 70 75 80 85 90 95 10010511011512012! 0 300000 I I I I I I I I I I I I I I I 9. Diamond(220) g Silicon(111) 8 250000 - - 4.3 E: (u _ . E) | (C) t—20h L p Diamond(gfl) # 1 50000 (b10101. 501 - w ‘ 'L; 1 00000 50000 - I I t=5h f._JLLi-1..l.i- 0 40 45 50 55 60 65 70 75 80 85 90 9510010511011512012! 20(Degree) I Fi e 5.21 X-ray diffraction s ectra of diamond films synthesized for (a) 5 . (b) 10 h, (c) 20 h, (d) 30 h, e) 55 h and (i) 100 h, using the experimen- tal conditions described in Section 5.5.2. 182 70 1 1 1 1 1 1 1 1 1 1 65- 60- 55- 50- 45- 40- D 35- 1 30- 25- 20- 15- 10- 0 5.. 01101111111111 0 1020 3040 5060 708090100 Deposition Time(h) I (220)/I (111) El Figure 5.22 The XRD height of (220) relative to that of (l l 1) as a function of deposition time in high pressure MCPR. 183 5.5.6 Film Structural Quality=f(t) The influence of deposition times on FWHM is shown in Figure 5.23. Raman spectra of the diamond films produced by different deposition time is displayed in Figure 5.24. As shown in Figure 5.23, FWHM is reduced as the deposition time is increased. The narrowest FWHM of 2.84 cm'1 is measured from a diamond film synthesized for 100h. This value is very close to the value of natural diamond(2.3 cm’l). As shown in Figure 5.24, a sharp peak of 1332 cm'1 is observed and no graphite or nondiamond carbon background is observed in these Raman spectra. 1 o I I I I I T 9 I- -i O 8 L -4 7 >— O —1 :1" 18 6 ,_ _1 O o E 5’ ‘ 4 _ ’ .. E . 3 " . -1 0 - FWHM of natural diamond=2.30m 2 P a 1 _ c- o I I I J + I I L I o 10 20 so 40 so so 70 so so 100 110 12o Deposition Time(h) Figure 5.23 FWHM of the diamond peak lines vs. deposition times. Their Raman spectra are shown in Figure 5.24. 184 60000 I I I I l I I I I 50000 - (f) t=100h 40000 - 30000 (d) t=30h (0) t=20h 20000 Relative Intensity (Count per second) (b) t=10h 10000 - 0 1220 1260 1300 1340 1380 1420 1460 1500 1540 1580 Wavenumber (cm'l) Figure 5.24 The effect of deposition time on Raman spectrum. The experi- mental conditions are listed in Section 5.5.2. 185 5.6 Summary Diamond films that were deposited within the optimized diamond growth zone (centralized at c~3%, ft~600 sccm, Ts~1080°C, Pt~3.3 kW, p~135 Torr, t= 100 h) showed various film morphologies, film textures, and film structural qualities as one of these independent experimental input variables was varied around the optimized deposition conditions. When 0 was varied between 1% and 8% while holding ft~ 600 sccm, Ts~10800C, Pt~3.3 kW, p~135 Torr, t=5 h), diamond films partially exhibited (100) square morphology at c=2%, (l 10) roof-like and (l 1 1) triangular morphologies at c=3%~5%, and showed cauliflower morphology at c 6%. Raman spectra showed the stnng diamond peaks at 1332 cm"1 when 0 was between 2% and 5%. These methane concentrations produced the diamond films with crystalline diamond morphologies. However, Raman spectra showed a large graphite peak at c=6-8%. These methane concentrations produced the diamond films with microcrystalline morphologies. When ft was varied between 412 sccm and 1442 sccm, the resulting films mostly exhibited (1 10) roof-like and (l 1 1) triangular morphologies and strong Raman diamond peaks. When T8 was varied from 950 to 1 125°C, diamond films showed partially (100) square morphology at TS=950°C and films exhibited (110) roof-like and (111) triangular morphologies when TS 950C. The films showed strong Raman 186 diamond peaks in this temperature region. As deposition times were varied from 5 to 100 hours, the resulting films displayed both (1 10) roof- like and (l 1 1) triangular morphologies at t<30h. When films synthesized with deposition times longer than 30 h, the films completely exhibited (l 1 1) triangular morphology. Films that exhibited both (1 l l) triangular morphology and <1 10> film texture were synthesized under conditions of maximum growth rate. CHAPTER SIX Development and Optimization of the Microwave Plasma Jet Reactor 6.1 Introduction This chapter presents the development and optimization of the microwave plasma jet reactor (MPJR) for the deposition of diamond films on irregular and conducting substrate materials. The MPCR apparatus described in Chapter 3-5 was used to create a uniform coating on flat substrate surfaces. This apparatus was difficult to be reproducibly used when coating surfaces with complex geometries such as ring seals and drills. The problem of uncontrollable and nonuniform deposition was especially evident if the surface to be coated was electrically conductive. The conductive surface interfered the electromagnetic fields within the cavity and made it very difficult to form a uniform and controllable plasma around the substrate surface. Often intense, nonuniform dis- charges were formed on the edges of the substrate. It was therefore an objective of the present research to design an apparatus and develop pro- cesses for the treatment of a surface with a complex geometry such as drill bits and or even substrates with very small and fragile structure 187 188 such as carbon fibers. The experimental work in this chapter carries on from the earlier research by J. Zhang[9],[l 10]. The early and basic versions of the MPJR, where originated from the work of Zhang and Asmussen [91,[110], are described and the historical development of this reactor concept is pre- sented. Then the optimization of these MPJRs for diamond deposition on round tools and carbon fibers and its final version are described. The experimental work described in Chapter 6 and 7 involved over 100 exper- iments and over 1,000 hours of reactor operation. The experimental work involved the process of the reactor design, experimental evaluation and test. 6.2 Development of the MPJR 6.2.1 Experimental Systems The MPJR described in this chapter utilizes the same experimental systems as those of the high pressure MCPR. These common experimen- tal systems include microwave power supply/waveguide transmission systems, flow control/vacuum pump systems, computer monitoring sys- tem, microwave tuning assembly, microwave cavity, and gas baseplate assemblies. They have been described in chapter 3. Indeed the major dif- ference between high pressure MPCR (described in Chapter 3-5) and MPJR (described in Chapter 6-7) systems is the reactor configuration 189 inside the bell jar (i.e., substrate holder setup). These different reactor configurations can be loaded and unloaded from the systems manually by the system operator. Thus the operation procedures such as starting and shut down procedures are very similar for all of the configurations described in this thesis except that the cavity length (Ls) is approximately 20 cm for the high pressure MCPR and is approximately 19 cm for the MPJR. 6.2.2 Early Version of MPJR The MPJR rector concept originated from the earlier work of Zhang and Asmussenll 10]. The early version of MPJR is shown in Figure 6.1[9]. As shown, the substrate (10) and substrate holder (16) are located in a region separated from the plasma discharge (14). The metal stage (7) and quartz tube (17b) are connected together with a tight fit in order to force the input gas to flow through the jet grid (23). This jet grid is an electri- cally conducting plate (22) in which one or more holes are present to allow the input gas to flow from above the plate, onto the deposition region below the plate. A plasma discharge is created above the conducting plate and the input gases are then forced through the plasma, through the grid plate (22), and onto a substrate located below the grid plate (22). This design reduces the input gas by passing of the discharge, i.e., all the input gas must flow through the discharge and thus improves the gas uti- lization efficiency. Since the input gases are flowed through the discharge 190 and through one or more small orifices, this configuration is called a plasma jet. It was expected that the jet-like design would result in a high deposition rates similar to other jet reactors[41]-[42],[93]-[96]. The design also has the advantage of having very low microwave fields in the sub- strate deposition region. Ionization takes place above the grid plate or in the grid hoes. The deposition volume is essentially microwave field free where the excited species and radicals were created above the grid, flowed/ diffused onto and around the substrate. Therefore this configura- tion also allows the diamond deposition on irregular and conducting sub- strate materials. 191 (B) Substrate Holder Setup Lama) (6) Resonant Breaker (7) Metal Stage (10) Substrate (14) Plasma Discharge (16) Substrate Holder (17a) Quartz Tube (17b) Quartz Tube (2) Grid Plate (23) Jet Grid (320 Laser Port Figure 6.1 Cross sectional view of the early version of MPJR for high rate diamond film deposition on flat substrates[9]. 192 6.2.3 Basic Configuration of the MPJR for Diamond Film Deposi- tion on Round Tools Figure 6.2 schematically displays the basic version of the MPJR developed by Zhang[l 10] that was first used to deposit the diamond film on irregular shaped objects such as drill bits. This reactor configuration utilized a plasma discharge that was created below the jet plate (22) instead of above the jet plate. As shown the input gases were forced downstream from the nozzle (23), through the plasma (14), and onto the substrate surfaces (10). The deposition area were shielded from the elec- tromagnetic fields by the jet plate and thus the downstream discharge can cover many irregular shaped substrates. The substrates(10), which were two 2" long, %" diameter WC-6%Co round tools, were placed on the metal stage (7) and supported by a graph- ite holder (16). This graphite holder was mounted on a 10 mm long, 4” diameter quartz tube (17a). A jet nozzle plate (22) with one %" diameter nozzle(23) at its center was placed on a 77 mm long, 4" diameter quartz tube (17b). This jet plate was used to force the source gas mixtures of CH4 and H2 to flow through the nozzle and onto the substrates. A plasma dis- charge (14) was created in the substrate region where the electromagnetic field strength is low. With this substrate holder configuration, uniform but weakly adhering diamond films were deposited on the surfaces of the two round tools. However when the number of the tools was increased to 193 six, nonuniform diamond film deposition on the tool surface was observed for all of the six tools. (B) Substrate Holder Setup 13350001131 (7) Metal Stage (10) Substrate (14) Plasma Discharge (16) Substrate Holder (17a) 10mm Quartz Tube (17b) 77mm Quartz Tube (22) Jet Nozzle Plate (23) Jet Nozzle (32) Laser Ports Figure 6.2 The cross sectional view of the basic version of MPJR for diamond film coating on two WC-6%Co round toolsll 10]. 194 6.3 Optimization of the MPJR for Diamond Film Deposition on Round Tools 6.3.1 Introduction While the first prototype MPJR could deposit films uniformly on one or two round tools, as the number of the tools increases the deposition became nonuniform. Thus reactor concept was redesigned and optimized to deposit films on very many tools. This study was primarily experimen- tal where different geometries were tested, evaluated and redesigned. Some of the results of this investigation are described in the section below. 6.3.2 Optimization Figure 6.3 displays the first reactor geometry that was modified from the basic configuration of MPJR. This first variation was used to deposit diamond films on six 1 g " long, %” diameter WC-6%Co round tools. It is noted that this reactor variation used the different size of round tools from previous deposition experiments. These 1%" long, %” diameter WC-6%Co round tools were used as the benchmark substrate in these 1.. 1 optimization. As shown, substrates(10), which were six 1 2 long, §0 diameter WC—6%Co round tools, were placed on a graphite holder (16). 195 This graphite holder was mounted on a 17 mm long, 4" diameter quartz tube (17a). A jet nozzle plate (22) with one ‘11” diameter nozzle(23) at its center was placed on a 77 mm long, 4" diameter quartz tube (17b). This jet plate was used to force the source gas mixtures of CO and H2 to flow through the nozzle and onto the substrates. A plasma discharge (14) was created in the substrate region. With this substrate holder configuration, uniform and adherent diamond films were deposited on the surfaces of the six round tools. However nonuniform films were deposited on the tool surfaces when the number of the tools was increased to eighteen. 196 (B) Substrate Holder Setup mammal (6) Resonant Breaker (7) Metal Stage (10) Substrate (14) Plasma Discharge (16) Substrate Holder (17a) 17mm Quartz Tube (17b) 77mm Quartz Tube (22) Jet Nozzle Plate (23) Jet Nozzle (32) Laser Ports Figure 6.3 The cross sectional view of the first variation of MPJR for diamond film coating on six WC-6%Co round tools. 197 Figure 6.4 shows the second variation of MPJR that was used to deposit diamond films on eighteen 1%" long, %" diameter WC-6%Co round tools. As shown, eighteen round tool substrates(lO) were placed on a graphite holder (16) and supported by a graphite shielding plate (21). The graphite holder and the shielding plate were mounted on a 17 mm long, 4" diameter quartz tube (17a) and a 10 mm long, 4” diameter quartz tube (17b), respectively. A jet nozzle plate (22) with one :1" diameter noz- zle(23) at its center was placed on a 77 mm long, 4" diameter quartz tube (170). This jet plate was used to force the source gas mixtures of CO and H2 to flow through the nozzle and onto the substrates. A plasma dis- charge (14) was created in the region between two conducting plates (21.22). With this substrate holder configuration, uniform and adherent diamond films were deposited on the surfaces of the eighteen round tools. Uniform diamond film deposited on the tool surface was observed when the number of the tools was increased to thirty six. This substrate holder setup demonstrated the repeatability and resulted in uniform diamond film deposition over the thirty six round tool substrates. 198 7 l ' I 17a To Pump (B) Substrate Holder Setup LeuandnLtBi (6) Resonant Breaker (7) Metal Stage (10) Substrate (14) Plasma Discharge (16) Substrate Holder (17a) 17mm Quartz Tube (17b) 10 mm Quartz Tube (17c) 77mm Quartz Tube (22) Jet Nozzle Plate (23) Jet Nozzle (32) Laser Ports Figure 6.4 The cross sectional view of the first variation of MPJR for diamond film coating on eighteen and thirty six WC-6%Co round tools. 199 Figure 6.5 displays another design of MPJR that was used to deposit diamond films on 7 um diameter carbon fibers. As shown, sub- strates(10), which consisted of two thousand 4.5 inch long, 7 pm diameter carbon fibers, were placed on a graphite holder (16). This graphite holder was mounted on a 4" diameter graphite shielding plate (21). This shield— ing plate was placed on a 40 mm long, 4" diameter quartz tube (17a). A jet _1_,. nozzle plate with one 4 diameter nozzle(23) at its center was placed on a 20 mm long, 4" diameter quartz tube (17b). This jet plate was used to force the source gas mixtures of CO/CH4/ H2 to flow through the nozzle and onto the substrates. A plasma discharge (14) was created in the region between these two conducting plates (21,22). With this substrate holder configuration, uniform and adherent diamond films were deposited on the surfaces of carbon fibers. This substrate holder setup demon- strated repeatable experiments and resulted in uniform diamond film deposition on thousands of carbon fibers. 200 To Pump (B) Substrate Holder Setup Legend (6) Resonant Breaker (7) Metal Stage (10) Substrate (14) Plasma Discharge (16) Substrate Holder (17a) 40mm Quartz Tube (17b) 20mm Quartz Tube (22) Jet Nozzle Plate (23) Jet Nozzle (32) Laser Ports Figure 6.5 The cross sectional view of the thirdth variation of MPJR for diamond film coating on many carbon fibers. 201 6.3.3 Operation of the Optimized MPJR for Diamond Film Deposi- tion on Round Tools The final configuration of MPJR optimized for diamond film deposition on round tools is shown in Figure 6.6. The baseplate assembly consists of a water-cooled and air-cooled baseplate(2), an annular input gas feed plate(3), a gas distribution plate(4), a 12.50m i.d. quartz dome(5), the MPJR substrate holder setup assembly (6, 7, 16, 17, 21. 22, 23) and the substrate itself (10). After the reactor chamber was evacuated to ~5 mTorr, source gas mixtures of CH4/CO/H2 were introduced from the baseplate (2), the annular gas plate (3), and the gas distribution plate (4) into the quartz dome (5) volume. A plasma discharge(l4) is created at pressures between 25-50 Torr by varying LD and L8 to the position of ~3 and ~ 19 cm, respectively, which matches and excites the TM013 plasma- loaded resonant mode. By adjusting the cavity length to ~19 cm, the plasma discharge was positioned in the substrate region below the jet plate (22) and can be viewed through the laser diagnostic ports (32). The substrate temperature was measured by focusing an optical pyrometer through these ports. A metal tube (6) which served as an electromagnetic field resonance breaker was placed inside the quartz tube (17a) and prevents the plasma discharge from forming underneath the substrate by reducing the electric field underneath the substrate. The metal tube (6) and quartz tube (17a) are placed on a metal stage (7) which has 30 mm diameter hole in its center to pass the hot gases to the pump. 202 7 17a To Pump Leusud (1) Cavity Walls (2) Baseplate (3) Gas Annular Plate (4) Gas Distribution Plate (5) Quartz Bell Jar (6) Resonant Breaker (7) Metal Stage (10) Substrate (14) Plasma Discharge (16) Substrate Holder (17a) 10mm Quartz Tube (17b) 77mm Quartz Tube (22) Jet Nozzle Plate (23) Jet Nozzle (32) Laser Ports Figure 6.6 The cross sectional view of the final optimized configura— tion of MPJR which was used for diamond film coating on eighteen and thirty six WC-6%Co round tools. 203 6.4 Summary The MPJR is a new microwave plasma reactor concept, which has several advantages over the more conventional MCPR. First all the input gases are forced to flow through the discharge and thus gas bypassing is eliminated. The MPJR also utilizes a different configuration of creating a plasma jet discharge from the microwave discharge formed in MCPRs. In MPJR, the reactive gases are forced to flow through a nozzle or a group of nozzle-like holes ((23) in Figure 6.2 - Figure 6.4). Then they are dissoci- ated in the plasma (14) which is located in the holes and in between two conducting plates (21,22). The dissociated reactive species then flow/dif- fuse onto the substrate (lO). In other CVD jet reactors[891-[97], significant amount of volume and surface recombinations of dissociated reactive spe- cies are observed when they are forced through a nozzle. Hence, the gas flow and power efficiencies of the reactors are greatly reduced. In addi- tion, when hot gases are forced through the nozzles in these jet reactors, the problems of erosion, deposition, and melting on the nozzle are also found. On the other hand, the plasma in MPJR is located downstream from the nozzle (23). Thus the problem of recombination in the nozzle is significantly reduced. The input gas flow serves as a natural cooling agent to the nozzle (23) and thus also drastically reduces the nozzle erosion and hence film contamination. The excited radical species are located in a microwave field free region and thus allows deposition on irregular shaped and conducting substrates. This new reactor configuration pro- 204 duces the discharge adjacent to the substrate, thus the discharge is not any larger than necessary, resulting in efficient use of microwave energy. The plasma created in the MPJR and CVD process are relatively independent of the geometry and material of the substrates. Therefore it becomes possible to use this apparatus to coat the substrates with com- plex geometries and different materials, such as cylindrical cemented car- bide inserts (1/8" in diameter) and fine carbon fibers (7 pm in diameter). The experimental parameter space and experimental results for diamond film deposition using MPJR are presented in the next chapter. CHAPTER SEVEN The Experimental Performance of the Microwave Plasma Jet Reactor 7.1 Introduction This chapter presents the experimental performance of the opti- mized MPJR. The experimental input/ output variable space and experi- mental procedures are first described. Then the optimized experiments that produced uniform and adherent, approximately 1-5 um diamond films on round tools and carbon fibers are presented. It is expected that the optimum deposition conditions presented in this chapter will be use- ful when evaluating process scale up for round tools and carbon fibers deposition. 7.2 Experimental Input/ Output Variable Space During this thesis research over 100 experimental runs represent- ing over 1000 hours of reactor operation were performed. The resulting experimental parameter space that was used to deposit uniform and adherent diamond films on the benchmark substrates is displayed in Fig- ure 7.1. Table 7 .1 summarizes the ranges of the experimental variables 205 206 that have achieved excellent diamond film deposition on the benchmark substrates. In Section 3.4 it was pointed out that the microwave plasma deposition process is very complex. It consists of a large number of input (U=[U1,U2,Usl), internal (X), and output variables (Y). The output vari- ables (Y) are dependent upon the input and internal variables, i.e., Y=g(U,X) and the internal variables are dependent upon the input vari- ables (X=flU)). As shown in Figure 7.1 and Table 7.1, the controllable input vari- ables (U 1) include the gas mixture, total flow rate, deposition pressure, and absorbed microwave power. No gas concentrations, i.e., the meth- ane concentration (cl=%CH4/H2) between 0.5% and 2% and the carbon monoxide concentration (02=%CO/H2) between 2.5% and 10°/o were used to achieve useful diamond film deposition. Total flow rates, denoted as ft, were varied between 100 and 600 sccm. As indicated by the dashed curve in Figure 7 .1, deposition pressure (p=13-50 Torr), absorbed microwave power (Pt=0.8-1.5 kW), and substrate temperature (Ts=600-1050°C) are directly related since the MPJR is operated in a thermally floating config- uration[51]. Thus these three input variables are not independent of each other and for the experiments presented in this chapter are referred to as a triad input variable, namely, p-Pmc-TS. Using the ranges of these exper- imental controllable input variables, good to excellent diamond films were deposited on the surface of benchmark substrates. 207 The reactor geometry variables (1.12) were (1) a quartz dome with inner base diameter of 5” (12.7 cm), (2) the MPJR substrate holder setup,(3) the substrate. The quartz dome size is fixed for the experiments presented in this chapter. Two optimized MPJR substrate holder setups (Figure 6.6 and Figure 7.14) were used to investigate diamond film depo- sition on the irregular shaped substrates. Note also that the substrates themselves were variables, i.e., the number and placement of the round tools and carbon fibers could be changed from rrrn to run. The deposition process variables (1.13) such as start-up/shut-down procedure and substrate seeding procedure were held constant except the deposition time, which was varied from 4h to 30h. The experiments that were investigated in this chapter did not utilize any substrate pretreat- ment procedure (including substrate cleaning, etching and seeding). The output performance variables that were of interest were film growth rate, film uniformity, film structural quality, and film morphology. Note that only experimental runs where the films were observed to stick to the sub- strates after the experimental run was completed were included as useful experimental data. 208 .250meng 25 .8.“ 282.5» Bwfiw a ma 00.506 1:8 on 58 339:3 no 0mg 95 55 @5805 835.5, mp. 05 0m d 05 88.028 255 00:30 05. .moumbmnam xumanocon 05 no 2506030 BE 55 0.5520 .8.“ :5:wa «.003 made: a N 0.53m MD— mEE. conmmommn— 0m 8508on @0003 8.5 00.5 02601035 05 magnum ~> $29352 80m 0 1 N c .wumw Bwfi End. 5:50 Eufiosbm SE _ AITQOUH 0 H1: EU" 3 s 1..., _ N 50355959 III I I ImWBszzE £0 5 aho E ”m I I I I > u. 3.5 mm Imumbmhsm um amgom >32 0353‘s.“ I I I _ I I L @800 000.80 fi .56 a: a 95 258 50 555.03 was L 085590 .3 z n 1055655 magnum Seam EB 3535 ~52 5600 55:0 _ _ _ _ a: 209 Table 7.1 Reactor input/ intemal/ output variables investigated in this and next chapters. a) Deposition Pressure. Variable: p=13-50’I‘orr Controllable b) Absorbed Microwave Power, Variable: Pt=0.8-1.5 kW Input Variables, 0) Gas Chemistry, Variable: cl=0.5-2% CH4/H2,02=2.5— U1 10% CO/H2 c) Total Flow Rate, Variable: ft=100-600 sccm a)Quartz Dome Size, fixed at 5" diameter b) Reactor Configuration, Variable: Substrate Holder Input Reactor Setups Variable, Geometry c) Substrate Material and Size, variable: WC-6% Round U Variables. U2 Tools and Carbon Fibers d) End-feed Excitation. Fixed e) Electromagnetic Mode and Cavity Tuning, Fixed at TM01:3 gleopcoessistion a) Starting and Shut-down Procedures, fixed Variables II b) No Substrate Seeding Procedure ' 3 c) Deposition Time, Variable: t=4-30 Hours Internal Variable, a) Substrate Temperature. Variable: T,=600-1050C X Output Reactor Variable, Performance. Y1: a) Linear Growth Rate Y Film 3) Film Uniformity Characteristics. b) Film Structural Quality Y3: 0) Film Morphology 7.3 Experimental Procedures 7.3. 1 Substrate Pretreatment Procedure The experiments presented in this chapter used gas chemistries that were different from the CH4/H2 experiments presented in Chapter 3- 210 5. As shown in Figure 7.1 and Table 7 .1, two different carbon containing gases, CO and CH4, were mixed with the hydrogen gas. The mixtures of these two gases are denoted as cl=CH4/H2 and 02: CO/H2. The addition of CO/H2 was found to increase the deposition rate three to five times over the deposition rate achieved by using CH4/H2. The CO/H2 gas mix- tures also were able to produce films without any special substrate seed- ing. Thus no substrate seeding procedure was necessary for the experiments presented in this chapter. 7.3.2 Start-up and Shut-down Procedures Since MPJR used the same experimental systems that were used in high pressure MCPR, the start-up and shut-down procedures are referred to Section 3.6.2. The reactor operation is referred to Section 6.3.3. 7.3.3 Measurement of Substrate Temperature The substrate temperature was measured by an optical pyrometer. Optical pyrometer temperature measurements are performed through the side laser ports ((32) in Figure 6.1 - Figure 7.2) focusing directly upon the rod tools or the carbon fibers. 7.3.4 Measurement of Reactor Output Variables Y 7.3.4. 1 Measurement of Reactor Performance Variable (Y 1) 211 The linear growth rate is the only reactor performance variable that was investigated for MPJRs. The thickness of the resulting film was mea- sured fiom the cross sectional view of the SEM photo. The linear growth rate is defined as the thickness divided by the deposition time. 7.3.4.2 Measurement of Film Characteristic Variables (Y,) The film uniformity over the substrate was determined by two thickness measurements from SEM cross sectional photos over the bro- ken tools or fibers. The two measured spots were located in the opposite position on the broken surface. One thickness measurement, referred as the minimum thickness measurement, and the other thickness measure- ment, referred as the maximum film thickness measurements, were recorded for each tool cross section. Film uniformity was calculated as: Film Uniformity: (Maximum Thickness Measurement-Minimum Thickness Measurement/Maximum Thickness X 100%. Raman Spectroscopy was performed on the deposited diamond to assess the diamond structural quality. Film morphology was identified by SEM. 7.4 Diamond Film Deposition on W06%Co Round Tools 212 7.4.1 Reactor Configuration Figure 6.6 displays a cross-sectional view of the MPJR that was used to deposit uniform diamond films on round tools (10). As shown thirty six 1 /8" diameter, 1% " long, WC-6% wt Co round tools were used as the substrates (10). These tools were held in a vertical position by a graphite holder (16) and supported by a shield plate (21). This substrate holder (16) and shielding plate (21) were mounted on the quartz tubes (17a) and (17b), respectively. A plasma (14) is created in the region between the jet pattern plate (22) and the shielding plate (2 1) below a noz- zle (23) by tuning cavity length and probe length at 19 cm and 3 cm, respectively. The metal stage plate (7) and quartz tube (170) are placed together so that the source gas is forced to flow through the nozzle (23) of the jet pattern plate (22), through the plasma (14), and flows out through a ring of holes in the shielding plate (21) and finally through the center hole of the substrate holder (16). The plasma discharge (14) was confined between two conductive parallel plates (21),(22). The one inch top sur- faces of the round tools were placed between (21) and (22) so that the dis- charge and the free radicals could freely diffuse over these irregular surfaces. The following sections present the optimized experimental input variables and the reactor output performance. 7.4.2 Experimental Controllable Input Variables (U) and Internal Input Variables (X) 213 The optimized independent experimental input variables (U) and internal input variables (X) that were used to produce the uniform (4% or better), thin (~3 urn) diamond films over 36 round tools per run were: (1) Controllable input variables (U1): a) Power input: i) Power source: a 2.45 GHz microwave generator, ii) Pabs= 1.35 KW. b) Gas input: i) Gas mixture: CO/H2 =10/ 400 sccm, CH4/H2=O, ii) Total flow rate=410 sccm, c) Deposition pressure=30 Torr (4 kPa), (2) Reactor geometry variables(Uz): a) Reactor Configuration: - A 5" diameter(12.7 cm) quartz bell jar, - Substrate holder setup for tool coating (Figure 7.1) was applied, b) No external substrate heating, cooling or biasing, (3) Deposition process variables (U2): a) Deposition time: 8 h, b) No Substrate pretreatment procedures such as cleaning, seeding, etc. were applied, 0) Substrate size and material=l /8” diameter, 1% " WC-6%Co 214 round tools, (B) Internal variables(X): (1) Substrate temperature=754-800°C(1027- 1073°K). Using these experimental input conditions, the reactor performance (Y 1) and film characteristics (Y 1) are presented in the sections below. 7.4.3 Reactor Performance (Y 1) - Linear Growth Rate Figure 7 .2 shows two thickness measurements for each round tools (totally 36 round tools) that were produced from a single run using the experimental conditions described in Section 7 .4.2. Two measured spots were chosen as the thickness measurement for each tool using the SEM. The first spot was picked on one edge of the breaking tool surface and the second spot was on the other side. Film thickness of 2.0 um-3.6 um were deposited on these 36 tools after a 8 hour experimental run. The average linear growth rate is 0.25-0.45 um/h. 215 4.5 — - a ll ? 1 4 5 6 7 8 9 10111213141516171819202122232425 26272829303132 38843536 Tool Number 9’ 01 h I I r r 0) I r .I -‘ 01 N I I I Thickness Measurement um) IO 01 O U! l Figure 7.2 Thickness measurements of 36 WC-6%Co round tools. 7.4.4 Film Characteristic (Y9) 7.4.4.1 Film Uniformity Figure 7.3 shows the film uniformity of thirty six round tools that were diamond coated in a single experimental run. Film uniformity is cal- culated from the thickness data shown in Figure 7.2 according to the for- mula described in Section 7.3.4.2. One representative cross sectional 216 SEM photo of the diamond film deposited on the round tools is shown in Figure 7.4. dde-‘MNNNNW oroommorobarmo Uniformity(%) 310th 123 4 S6 7 8 91011 I2131415161718192021222324252627282930313233343536 Tool Number Figure 7.3 Film Uniformity of diamond films for thirty six WC-6%CO round tools. 217 7.4.4.2 Film Morphology Film morphologies of the WC-6%Co surface that were deposited with diamond films at different times were shown in Figure 7 .4-Figure 7 .12. The experimental input variables shown in Section 7 .2 were held constant except that the deposition time was varied fi'om 0.5 h(Figure 7.5), l h(Figure 7.6), 2 h(Figure 7.7), 3 h(Figure 7.8), 4 h(Figure 7.9), 5 h(Figure 7.10), 6 h(Figure 7.11), and 8 h(Figure 7.12). Figure 7.4 shows the surface morphology of a WC-6%Co before deposition. As shown in Figure 7 .5, the WC surface was nucleated by the CO/ H2 plasma in the first 0.5 hour. At t=1h(Figure 7.6), some octahedron dia- mond crystals were deposited on the etched surfaces. At t=2h(Figure 7 .7), more octahedron diamond crystals covered the surfaces. However the round tool surface was not fully covered by the diamond film at this moment. It was observed that these 1 urn big octahedron diamond crys- tals were believed to be the first layer of the diamond film growth on the WC tools. They offered the nuclearation of high-order growth. At t=3h(Fig- ure 7.8), the WC round tool surfaces were fully covered by cauliflower dia- mond films with some square diamond films. At t=4h(Figure 7 .9), the situation described in t=3h became more obvious that more square dia- mond films were observed. At t=5h(Figure 7 .10), square and triangular facets were both observed. At t=6h(Figure 7 .1 l), the grain size increased with the same morphologies described at t=4h. At t=8h(Figure 7.12), same morphologies were observed with increasing grain sizes. 218 — “I: Figure 7 .4 SEM photo of the pre-deposited WC-6%Co round tool sur- face. Figure 7.5 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 0.5 h. 219 Figure 7 .6 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 1 h. _ ' 1” Figure 7.7 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 2 h. 220 — 11 Figure 7 .8 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 3 h. s»; _ “1‘11 Figure 7 .9 SEM photo of a diamond film deposited WC-6%Co surface. The deposition time is 4 h. 221 Figure 7.10 SEM photo of a diamond film deposited WC-6%Co sur- face. The deposition time is 5 h. Figure 7 . 1 1 SEM photo of a diamond film deposited WC-6%Co sur- face. The deposition time is 6 h. 222 Figure 7 . 12 SEM photo of a diamond film deposited WC-6%Co sur- face. The deposition time is 8 h. 7.4.4.3 Film Structural Quality Figure 7.13 shows the Raman spectrum of the diamond film depo- sition experiment that was described in Section 7 .2. Raman Spectros- copy, which has approximately 30 um spot size, shows the diamond peak at 1332 cm'l. 223 16000 I I I I I I I 1.111. I :5 Count per Second 5 12000 110m I I I I I J I 1200 1250 1300 1350 1400 1450 1500 1550 1600 Wavenumber (cm' 1) Figure 7.13 Raman spectrum of the diamond film that was deposited on a WC-6%Co round tool surface. 224 7.5 Diamond Film Deposition on Carbon Fibers 7.5.1 Reactor Configuration Figure 7.14 shows schematic cross-sectional view of an apparatus for diamond thin coating of the carbon fiber substrate (10), wherein the substrate (10) being coated is supported by a graphite holder (16). This substrate holder (16) is mounted on a graphite shielding plate (21) which stands on a quartz tube (17) used to change the position of the substrate (10). A plasma (14) is created in the region between the jet pattern (22) and the graphite shielding plate by coupling microwave power into the plasma (14). The metal stage (7) and quartz tube (17) are sealed together to force the source gas to flow through the orifice (23) of jet pattern plate. The rest of the components of this apparatus are identical to those of the MCPR for diamond film synthesis on silicons and the MPJR for diamond film coating on cemented carbide round tools, which were described in Section 3.5 and Section 7.4.1, respectively. 225 . v O'O'Q'O'Q ’ ’020202020. AVA'AVA'AOAOA (1) Cavity Side Wall (2) Baseplate (3) Annular Gas Plate (4) Gas Distribution Plate (5) Quartz Dome (6) Resonant Breaker (7) Metal Stage (10) Carbon Fiber Substrate (11) Sliding Short (12) Finger Stock (13) Excitation Probe (14) Plasma Discharge (15) Screen Side View (16) Substrate Holder (17) Quartz Tube (21) Shielding Plate (22) Jet Nozzle Plate (23) Jet Nozzle (32) Laser Ports Figure 7.14 The cross sectional view of the microwave plasma jet reac- tor for diamond thin film coating on carbon fibersll4]. 226 7.5.2 Experimental Controllable Input Variables (U) and Internal Input Variables (X) The optimized independent experimental input variables (U) and internal input variables (X) that were used to produce the uniform, thin (~l um) diamond films over thousands of carbon fibers per run were: (1) Controllable input variables (U1): a) Power input: i) Power source: a 2.45 GHz microwave generator, ii) Power level: 1 kW, b) Gas input: i) Gas mixture: CH4/H/2 =4/4OO sccm ii) Total flow rate=404 sccm, c) Deposition pressure=35 Torr (4.66 kPa), (2) Reactor geometry variables (U2): a) Reactor size: - A 5" diameter(12.7 cm) quartz bell jar, - Substrate holder setup for fiber coating (Figure 7.14) was used, b) No external substrate heating, cooling or biasing, (3) Deposition process variables (U2): a) Deposition time=28 h, b) Substrates were pretreated in a CO/ H2(10/ 400 sccm) plasma for 20 minutes, 227 c) Substrate size and material=7 um diameter carbon fibers (B) Internal variables(X): (l) Substrate temperature=1000°C(1273°K). 7.5.3 Film Characteristic (Y2) Figure 7.15 and Figure 7. 16 show the SEM photos of diamond thin film deposition on carbon fibers. Since the film thickness is approxi- mately 1.5 um after a 28 hour run, the average linear growth rate is approximately 0.05 ml h. From both figures, the diamond films were uniformly deposited on the carbon fiber surface. Figure 7.17 shows the Raman spectrum of the diamond film that was deposited on the surface of carbon fiber. It is noted that the spot size of this Raman measurement is approximately 5 um in diameter, which is referred as micro-Raman measurement. 228 7.900 t I~ - ’ .25)" Philip Figure 7.15 The SEM picture of a diamond film coated carbon fiber. The experimental conditions are shown in Section 7.3.2. 229 Figure 7.16 The SEM picture of the diamond film on carbon fibers. This figure is an enlarge of Figure 7.15 that contains more carbon fibers. SE30 j l 7 r l l I I 1580 1344 150- 4 2702 'U 8 o 1w~ I1 0) U) H 0) CL ‘5 O C) 2923 4276 50- 3232 I o l l l A l 1 i 0 500 1000 1500 2M 2500 3M 3500 4W Wavenumber (cm’l) Figure 7 . 17 Raman spectmm of the diamond film on the carbon fibers. 231 7.6 Summary The configuration of the microwave plasma jet reactor and its opti- mized experimental input conditions that were used for the uniform dia- mond film coating on irregular shaped objectives are presented in this chapter. Rather, the original experiments intelligently chose several start- ing points, i.e., the substrate setups, CO/CH4/H2 chemistries, jet pattern plates, and then varied the experimental input variables in promising directions to achieve deposition uniformly over the irregular objects with reasonable adhesion and deposition rates. Achieving excellent adhesion was not an objective of this study. However, experimental exploration of the original experimental variable space was achieved with the exception of following variables: (1) experiments were limited to less than 1,500Watt input power, (2) total gas flow rates were held under 600 sccm, (3) no experiments were performed with 002 input gas, (4) high packing capacity was achieved, (5) no substrate pretreatment was included, (6) uniform, well—adhered diamond thin films(0.5~5 mm) were coated irregular objects such as WC-6%Co round tools and car- bon fibers, (7) diamond quality was identified(by Raman Spectroscopy), (8) smooth diamond thin films were observed(by SEM). CHAPTER EIGHT Conclusions 8.1 Introduction The experiments described in this dissertation have demonstrated the successful operation of two new prototype reactor configurations. The first, identified as an improved microwave cavity plasma reactor (MCPR) concept has demonstrated considerable improvement of performance when compared with other lower pressure microwave plasma thin film deposition reactors. This reactor operates at pressures between 80 to 150 Torr and produces a thermal-like plasma which is able to deposit uniform films over two to three inch substrates. This new reactor modified an existing low pressure microwave plasma reactor by incorporating a water cooled substrate holder and increasing the gas air cooling of the reactor. These design improvements allowed operation at higher pressures, higher input power levels and also achieved independent control over substrate temperatures. The output performance of the redesign, as measured by film growth rate, deposition efficiency (specific yield, carbon conversion effi- ciency, etc.), demonstrated considerable improvement over the published 232 233 performance of other microwave reactors. Deposition rates are of the order 50 mg/ h with energy efficiencies of 70 kW-h/ g (or 14 kW-h/ ct) while still achieving carbon conversion efficiencies of 5- 10%. These per- formance measures compare well with the best overall performance of any CVD diamond deposition reactor. Clearly, an original objective of this thesis research which was to increase the deposition rate has been achieved. A scaled up version of the MPCR, i.e., a thirteen inch diameter discharge reactor, excited at 9 1 5 MHz has been constructed and is in operation at Norton Diamond Film Corporation. While the performance measures of this larger reactor remain unpublished, if these performance measures are similar to those presented in this thesis then the larger scale up reactor is capable of high deposition rates, expressed in weight gain per hour, over large areas. This thesis research has demonstrated the ability to improve the deposition performance (deposition rate and deposition area) of microwave reactors. Thus this reactor concept itself opens up the possibility of applying microwave plasma reactors to both thin film and thick film deposition applications. The second prototype reactor, i.e. , microwave plasma jet reactor (MPJ R), is operated at pressures between 10 to 45 Torr. This reactor con- cept has the specific design advantages of (l) preventing gas bypassing, (2) providing a forced flow of input gases into the deposition region and even sometimes onto the deposition surfaces, and (3) reducing the elec- tromagnetic field in the deposition volume. The MPJ R utilizes a new pat- 234 tromagnetic field in the deposition volume. The MPJR utilizes a new pat- ented force-flow concept to create a microwave plasma discharge in a region where the electromagnetic field strength is low. This ensures that only the plasma is produced adjacent to the substrate which can be con- ductive and can have a complex geometry. CO/CH4/H2 source gases together with this force-flow concept are utilized to enhance the diamond film growth on irregular shaped, conducting materials with sharp edges. and on fragile materials. In the experiments presented in chapter six and seven demonstrate that the MPJR is able to deposit uniform diamond films on many of 1/8” diameter tungsten carbide round tools and thin films on 5-7 mm diameter carbon fibers. 8.2 Operational Characteristics of a High-Pressure MPCR 8.2. 1 Introduction Using CH4/H2 chemistries the MPCR was experimentally characterized in the high-pressure 80-150 Torr regime. In this pressure regime the microwave absorbed power density increases to 20-30W/cm3 and the discharge produces a thermal-like plasma where the translational temperatures and hydrogen electronic temperature are both within the 2000-2200k ranges. Under these conditions the maximum uniform film deposition rates increase to 4.6-7um/h(45mg/h on 5.08 cm diameter substrate). Thus the deposition rate have increased by a factor 235 of five over the deposition rates of low pressure nonequilibrium microwave discharge reactors. Using the approach of J. Angus [15], where the performance of different reactors are compared by relating growth rate vs. reactor area power density (i.e. Pabs divided by deposition area), the specific energy of this reactor can be calculated. The reactor performance is displayed as point C in Figures 2.19 and 2.20. The area power density is 0.15-0.25 kW/cm2 yielding a specific deposition emery of 14-20 kW-h/ ct or specific yield of 70-100 kW-h/ g. Optimum growth conditions are CH4/H2=3-4%, ft=600-700sccm, Ts=1060-l lOOC, p=120- 135Torr, and =30 W/cm3. Deposited films exhibit (l l l) morpholog' and <1 10> film texture under high growth rate conditions. The reactor performance is very repeatable and controllable with respect of the growth rate, morphology, etc. Both thick films and thin films can be deposited using many substrate materials. Using reactor scaling laws described earlier [1 12-1 15], this reactor concept with 915MHz excitation can create much larger thermal microwave discharges and thus has the ability to deposit films uniformly at high growth rates over much larger areas. 8.2.2 Diamond Film Deposition on 2" Silicon Wafers An experimental study of diamond film deposition on 2" silicon wafers was investigated by performing a series of experiments where dif- ferent CH4/H2 concentrations, substrate temperatures, deposition times 236 and total flow rates were important variables and measured various film growth rate, film uniformity, film Raman Spectrum, film morphologies, and film textures were measured as the output variables. The volume in input variable space. i.e., a volume in CH4/H2, T8, and ft space, where diamond films with excellent quality were synthesized was identified and compared to analogous volume for lower pressure microwave reactors. This three dimensional diamond deposition volume is much larger for the MCPR than the associated volume for the lower pressure, nonequilibrium microwave plasma reactors. This significant increasein the input variable space for good diamond deposition is believed to be caused by the shift of the discharge chemistry from a nonequilibrium cold plasma regime to a thermal-like plasma discharge. Films with excellent. well-defined crystal facets could be synthesized at high growth rate under much higher tem- perature (>IOOOC) and higher CH4/H2 concentrations (>3%). Other results of this study are briefly summarized as below: (1) The optimum growth conditions in the high-pressure MPCR system are 3%~4% CH4/H2 concentrations with total flow rate between 600 sccm and 800 sccm, substrate temperature of ~1050C, deposition pressure of 120-135 Torr, and absorbed MW power ~3.4 kW. The resulting growth rate is approximately 5.5 m/ h (or 46 mg/ h). (2) High quality diamond films of thickness between 10 to 100 um with uniformity better than 10% have been routinely deposited on 2" diameter silicon substrates. The high quality of the diamond film has 237 been demonstrated by its linewidth of Raman Spectra (3-8 cm'l). (3) 2" thick (> 100 um) free-standing diamond films with uniformity better than 15% have been routinely produced. High quality of the free-standing diamond film has been demonstrated by its linewidth of Raman Spectra (~3 cm'l) and thermal conductivity (IO-15 W/ cm-K)[1 16]. (4) Diamond films show different morphologies as CH4/H2 concentration varies while other experimental conditions are kept constants. These constant experimental controllable input variables include the substrate temperate (10600), deposition pressure (135 Torr), total flow rate (600 sccm), substrate temperate (5 h), and absorbed microwave power (3.4 kW). (1) diamond films exhibit mixed (100) square, (1 10) roof-like, and (l 1 1) morphologies at CH4/H2=2%. (ii) Diamond films mostly exhibit (1 l l) triangular morphologf at CH4/H2=3%~5%. (iii) Diamond films show cauliflower morphology at CH4/H2 6%. (4) Diamond films show different morphologies versus variations in substrate temperature while other experimental controllable input variables are kept constant. These variables include CH4/H2 (3%), deposition pressure (135 Torr), total flow rate (600 sccm), deposition time (10h), absorbed microwave power (3.4 kW). (i) At substrate temperatures between 8000 and 950C 238 diamond films show mixed (100) square, (110) roof-like and (11 l) triangular morphologies. (ii) At substrate temperatures between 950C and 1125C diamond films mostly exhibit (1 l 1) triangular morphologr. (5) Diamond films display different morphologies vs. the deposition time. These constant input variables include CH4/H2 (3%), substrate temperature (1060C), deposition pressure (135 Torr). total flow rate (600 sccm), absorbed microwave power (3.4 kW). (1) At deposition times between 4h and 30h the resulting films show both (1 10) roof-like and (1 l l) triangular morphologies. (ii) When films are synthesized longer than 30h, the films completely exhibit (l l l) triangular morphology. (6) Diamond films display different morphologies vs. the total flow rates. These constant input variables include CH4/H2 (3%), substrate temperature (10600), deposition pressure (135 Torr), deposition time (10 h), absorbed microwave power (3.4 kW). (1) At total flow rates between 600 and 800 sccm, the resulting films show (1 1 l) triangular morphologr. (ii) When films synthesized with 200

sample Us") (05) (1:11) (050) (scfctm) (It!) “m/h ($31k) (“I/(2,1113) (“r/£1.12) ST79 120 1010 3.52 2.54 564 55 4.06 126 24.4 173 ST231 135 1015 3.73 2.92 617.5 10 5.05 107 25.9 185 31225 135 1025 3.71 3 618 10 5.20 103 25.8 184 ST230 135 1010 3.71 3 618 10 4.52 1 19 25.8 184 $1249 135 965 3.67 1.5 609 10 1.24 429 25.5 181 $1247 135 1053 3.67 2 612 8 2.23 239 25.5 181 TK38 135 1055 3.72 2.5 615 10 2.64 204 25.8 184 TK37 135 1053 3.71 3 618 14 4.80 112 25.8 184 TK33 135 1096 3.66 3.5 621 10 4.41 120 25.4 181 TK40 135 1010 3.66 4 624 10 4.86 109 25.4 181 $183 135 990 4.13 2.8 514 4 5.34 1 10 28.7 203 ST64 135 990 4.10 2.33 614 4 5.01 1 19 28.5 202 ST48 135 1060 4.16 2 357 8 3.57 1 69 28.9 205 ST44 135 1040 4. 18 1.75 407 8 3.42 177 29.0 206 31227 135 1018 3.71 3 721 10 5.17 104 25.8 184 ST65 120 916 4.00 2.33 614 4 4.53 128 27.8 197 TKQ 120 1085 3.62 3 618 8 4.63 1 13 25.1 178 ST199 120 1045 3.12 3 618 10 3.84 119 21.7 154 ST200 120 1055 3. 19 3 618 10 3.50 132 22.2 157 ST188 120 1055 3.12 3 618 10 3.08 147 21.7 154 S'1‘198 120 1025 3.15 3 618 10 2.95 155 21.9 155 TK7 130 1110 3.59 3 618 8 4.48 116 24.9 172 TK31 135 1010 3.73 3 618 8 5.04 107 25.9 184 TK32 135 1061 3.66 3 618 8 4.78 111 25.4 181 TK14 135 1006 3.62 3 618 8 4.44 118 25.1 179 TK34 135 1020 3.61 3 618 8 4.00 131 25.1 178 W355 135 1 113 3.66 3 618 8 3.46 153 25.4 181 ST221 120 1050 3.76 3 618 100 4.58 1 19 26.1 185 $1220 120 1045 3.76 3 618 100 5.85 93.2 26.1 185 31215 135 1035 3.72 3 618 100 4.91 l 10 25.9 183 Table A. l (cont’d) 245 T p c f t Specific

sample (l‘grr) 1°01 (kw) (%) (scctm) (h) “In/h mfiffiig, (W/cm3) (W/:m2) 81233 135 1015 3.74 3 618 100 6.25 86.8 26.0 184 M05 135 1080 3.35 3 618 100 6.27 77.5 23.3 165 ST190 135 1065 3.76 3 618 58 6.10 89.4 26.1 185 ST192 120 1000 3.19 3 618 72 4.83 96 22.5 157 81204 120 1060 3.12 3 618 75 4.30 105 21.7 154 A126 135 1056 3.22 1 606 10 0.43 1086 22.4 159 A125 135 1057 3.22 2 612 10 1.77 264 22.4 159 A116 135 1057 3.22 3 618 10 3.94 1 19 22.4 159 A127 135 1054 3.22 4 624 10 4.31 108 22.4 159 A128 135 1057 3.22 5 630 10 3.86 121 22.4 159 A156 135 1056 3.22 6 636 10 3.52 132 22.4 159 A164 135 1056 3.22 7 642 10 3.39 137 22.4 159 A165 135 1055 3.22 8 648 10 3.32 141 22.4 159 AT35 135 950 3.17 3 618 10 3.18 145 22.0 156 A129 135 1025 3.10 3 618 10 3.56 126 21.5 153 A141 135 1050 3.02 3 618 10 3.02 145 21.0 149 AT43 135 1080 3.02 3 618 10 4.59 95.4 21.0 149 AT33 135 1100 3.10 3 618 10 4.46 101 21.5 153 AT31 135 1125 3.15 3 618 10 4.02 114 21.9 155 A116 135 1057 3.22 3 618 5 3.94 1 19 22.4 1 59 A179 135 1132 3.10 3 412 10 3.97 113 21.5 153 A167 135 1110 3.12 3 618 10 4.50 101 21.7 154 A174 135 1128 3.12 3 721 10 4.41 103 21.7 154 A170 135 1133 3.14 3 824 10 5.13 88.8 21.8 155 A172 135 1132 3.14 3 927 10 5.17 87.8 21.8 155 A173 135 1114 3.12 3 1030 10 5.24 86.3 21.7 154 A175 135 1133 3.12 3 1133 10 4.70 96.3 21.7 154 A182 135 1130 3.12 3 1236 10 5.06 89.4 21.7 154 A180 135 1140 3.14 3 1339 10 2.37 192 21.8 155 A181 135 1145 3.14 3 1442 10 2.60 175 21.8 155 Table A. 1 (cont'd) 246 1 p C 1' t spec‘fic

SamP‘° 115m 1°01 1kvtv1 1%) (945m) on “W“ 0.31%, (W/cm3) (WI:mZ) AT‘88 100 745 2.69 4 624 10 1.67 234 11.5 132 ST138 100 1031 3.33 3 721 20 3.42 141 23.1 164 ST138 100 165 3.33 3 824 20 3.27 147 23. 1 164 ST140 100 1013 3. 19 3 721 20 3.28 141 22.2 157 ST141 100 1018 3.22 3.5 621 20 3.09 151 22.4 159 ST142 100 994 3.19 2.5 615 20 2.86 162 22.2 157 ST126 100 1028 3.26 3 412 20 3.8 124 22.6 160 ST134 100 1042 3.21 3 515 20 3.14 148 22.3 158 ST135 100 1050 3.21 3 412 20 2.99 156 22.3 158 ST137 100 1018 3.19 3 618 20 3.28 141 22.2 157 ST132 110 1088 3.26 3 515 20 3.60 131 22.6 160 ST133 110 1067 3.28 3 515 20 4.04 l 18 22.8 161 TK15 110 1015 3.29 5 630 5 1.74 274 22.8 161 AT89 90 703 2.69 4 624 10 1.67 234 18.7 132 AT105 90 945 2.48 3 618 8 2.60 138 17.2 122 AT110 90 945 2.48 2 612 8 1.48 242 17.2 122 BCl 80 650 2.37 4 624 19 0. 174 1975 16.4 1 17 AT104 80 887 2.48 3 618 8 1.96 183 17.2 122 AT109 80 845 2.50 2 612 8 1.64 221 17.4 123 BC2 70 601 2. 10 4 624 8 0.55 526 14.6 104 AT103 70 803 2.29 3 618 8 1.35 246 15.9 112 AT108 70 801 2.29 2 612 8 1. 11 300 15.9 112 J21 33 750 1.47 2 204 8 0.32 290 8.65 32 J22 45 850 2.05 2 204 8 0.64 203 12. 1 45 J23 57 950 2.51 2 204 8 0.58 274 14.8 55 J24 51 900 2.34 1.5 50.75 8 0.32 463 13.8 51 J25 51 900 2.34 1.5 101.5 8 0.50 292 13.8 51 J26 51 900 2.34 1.5 203 8 ' 0.61 243 13.8 51 J27 51 900 2.34 1.5 304.5 8 0.65 228 13.8 51 J28 51 900 2.34 1.5 406 8 0.68 218 13.8 51 Table A. l (cont’d) 247 1‘ SPCC‘fiC <1) >

samP‘c (151-r) 60) (1:11 (:0 (scfctm) at) “m/ h (1.352111% (W/cvm31 (W/cAm2) J29 51 900 2.34 1.5 406 8 0.55 163 13.8 51 JZlO 40 836 2.28 1.5 406 8 0.47 179 13.4 49 J21 l 40 825 2.24 1.5 406 8 0.46 180 13.4 49 JZl2 37 813 2.20 1.5 406 8 0.42 194 13.4 49 JZ13 40 845 2.24 1.25 400 8 0.37 224 4.6 28 JZ14 40 845 2.24 1.38 400 8 0.42 197 4.6 28 JZlS 40 845 2.24 1.5 400 8 0.40 207 4.6 28 JZIG 40 845 2.24 1.63 400 8 0.40 207 4.6 28 JZl7 40 845 2.24 1.75 400 8 0.35 237 4.6 28 LIST OF REFERENCES LIST OF REFERENCES [1] M. Kamo, Y. Sato, S. Matsumoto, and N. Setaka, “ Diamond synthe- sis from gas phase in microwave plasma,” J. Crystal Growth,Vol. 62. pp.642-644. 1983. [2] A. R. Badzian. T. Badzian. R. Roy, R. Messier, and K. E. Spear, “ Crystallization of diamond crystals and films by microwave assisted CVD (Part 11),” Mat. Res. Bull.,Vol. 23, pp.53l-548 , 1988. [3] K. Kobashi, K. Nishimura, Y. Kawate, and T. Horiuchi, “ Synthesis of diamonds by microwave plasma chemical-vapor deposition: morphol- ogy and growth of diamond films,” Physical Review B, Vol. 38, pp4067-4084. 15 August1988. [4] D. Pickrell, W. Zhu, A. R. Badzian. R. Messier, and R. E. Newnham, “ Downstream plasma-enhanced diamond film deposition.” Appl. Phys. Lett.. Vol. 56. pp.2010-2012, 14 May 1990. [5] R. A. Bauer, N. M. Sbrockey, and W. E. Brower. Jr., “ Quantitative nucleation and growth studies of PACVD diamond film formation.” J. Mater. Res.,Vol. 8, pp.2858-2869, November 1993. [6] N. Lee and A. Badzian. “ Eflect of methane concentration on surface morphologies and surface structures of (001) homoepitaxial diamond thin films,” Appl. Phys. Lett.. Vol. 67, pp.201 1-2013, 2 October 1995. [7] H. Kawarada, T. Suesada, and H. Nagasawa, “ Heteroepitaxial growth of smooth and continuous diamond thin films on silicon sub- strates via high quality silicon carbide buffer layers,” Appl. Phys. Lett.. Vol. 66. pp.583—585, 30 January 1995. [8] H. C. Barshilia, B. R. Mehta, and V. D. Vankar," Growth of diamond thin films by microwave plasma chemical vapor deposition process," J. Mater. Res., Vol. 11. pp. 1019-1024, April 1996. [9] J. Zhang. “ Experimental development of microwave cavity plasma 248 [10] [11] [12] [13] [14] [15] [16] [l7] [18] [19] 249 reactors for large area and high rate diamond film deposition,” Ph.D dissertation, Michigan State University(1993). G.S. Yang, M. Aslam, K. P. Kuo, D. K. Reinhard, and J. Asmussen, “ Effect of ultrahigh nucleation density on diamond growth at difier- ent growth rates and temperatures,” J. Vac. Sci. Tech.,Vol. 13, pp. 1030- 1036(May/June 1995). K. P. Kuo and J. Asmussen, " An experimental investigation of high pressure synthesis of diamond films using a microave plasma cavity reactor," accepted in Diamond and Related Materials. to be pub— lished. K. P. Kuo and J. Asmussen, “ An experimental investigation of high pressure synthesis of diamond films using microwave plasma cavity reactor,” 23rd IEEE International Conference on Plasma Science, Boston. Massachusetts, June 1996. K. P. Kuo and J. Asmussen, “ A parametric study of high pressure synthesis of diamond films using microwave plasma cavity reactor,” 7th European Conference on Diamond, Diamond-like and Related Materials, jointly with 5th International Conference on the New Diamond Science and TechnologIUCNDST-S), Tours, France(Sep- tember, 1996). K. P. Kuo and J. Asmussen,” A method for uniform diamond coating on the high packaging density drill tools using a CO/Hz microwave plasma jet discharge,” U.S. Patent in process. J. C. Angus and C. C. Hayman. " Low-pressure, metastable growth of diamond and diarnondlike phases," Science, Vol. 241, 19 August 1988. R. F. Davis, " Diamond films and coatings," Noyes Pubications, pp. 224. D. S. Dandy, M. E. Coltrin," Relationship between diamond growth rate and hydrocarbon iryector location in direct-current arcjet reac- tors,"App1. Phys. Lett.. Vol. 66, pp.391-393, 16 January 1995. S. Matsumoto, Y. Sato, M. Kamo. and N. Setaka, “ Vapor deposition of diamond particles from methane,” Jpn. J. Appl. Phys., Vol. 21, pp. L183-L185(March 1982). S. Matsumoto. Y. Sato, M. Tsutsumi. and N. Setaka, “ Growth of diamond particles fiom methane-hydrogen gas,” Jpn. J. Appl. Phys., [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [301 250 Vol. 17. pp. 3106-31 12(1982). C. J. Chu, R. H. Hauge, J. L. Margrave, and M. P. D’Evelyn, " Growth kinetics of (1 00), (1 10), and (1 1 1) homoepitaxial diamond films," Appl. Phys. Lett., Vol. 61, pp. 1393-1395, 21 September 1992. A. Sawabe and T. Inuzuka." Growth of diamond thin films by elec- tron assisted chemical vapor deposition."App1. Phys. Lett., Vol. 46, pp. 146-147, 15 January 1985. K. Kukmagai, K. Miyata. K. Nishmura. and K. Kobashi, “ Growth of (1 10)-oriented diamond films by electron-assisted chemical vapor deposition.” J. Mater. Res., Vol. 8, pp. 314- 320(Feb 1993). K. Suzuki, A. Sawabe, H. Yasuda, and T. Inuzuka, “ Growth of dia- mond thin films by dc plasma chemical vapor deposition, ” Appl. Phys. Lett, Vol. 50. pp. 728-729(March1987). K. Suzuki, A. Sawabe, and T. Inzuka," Growth of diamond thin films by dc plasma chemical vapor deposition and characteristics of the plasma." Jpn. J. Appl. Phys., Vol. 29, pp. 153-157. January 1990. K. Kurihara. K. Sasaki, M. Kawarada, and N. Koshino, “ High rate synthesis of diamond by dc plasma jet chemical vapor deposition, ” Appl. Phys. Lett. Vol. 52, pp. 437-438, Feb. 1988. S. W. Reeve, W. A. Weirner, and D. S. Dandy. " On the optimization of a dc arcjet diamond chemical vapor deposition reactor," J. Mater. Res., Vol. 11, pp. 694-702. March 1996. S. K. Baldwin, T. G. Owano, and C. H. Kruger. “ Increased deposi- tion rate of chemically vapor deposited diamond in a direct-current arcjet with a secondary discharge, ” Appl. Phys. Lett, Vol. 67, pp. 194-196(July 1995). Y. Matsui. A. Yuuki, M. Sahara, and Y. Hirose, “ Frame structure and diamond growth mechanism of acetylene torch,” Jpn. J. Appl. Phys., Vol. 28, pp. 1718-1724, September 1989. R. C. Aldredge and D. G. Goodwin, “ Influences of ambient atmo- sphere on diamond synthesis using an oxygen-acetylene torch.” J. Mater. Res., Vol. 9, pp. 80-84. January 1994. K. Bang, A. J. Ghajar, and R. Komanduri," The effect of substrate surface temperature on the morphology and quality of diamond films [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] 251 produced by the oxyacetylene combustion method." Thih Solid Films, Vol. 238, pp. 172-183. 1994. H. S. Shin and D. G. Goodwin," Diamond growth in premixed propy- lene-oxygenflames," Appl. Phys. Lett. Vol. 68, pp.2909-291 1. 22 May 1995. S. J. Harris, H. S. Shin, and D. G. Goodwin," Diamond films fiom combustion of methyl acetylene and propadiene." Appl. Phys. Lett., Vol. 66. pp. 891-893, 13 February 1995. J. S. Kim and M. A. Cappelli, “ An experimental study of the temper- ature and stoichiometry dependence of diamond growth in a low pressure flat flames,” J. Mater. Res., Vol. 10, pp. 149-157. January 1995. J. S. Kim and M. A. Cappelli," Diamond film synthesis in low-pres- sure premixed methane-oxygenflames." Appl. Phys. Lett., Vol. 67, pp.1081-1083. 21 August 1995. S. Matusmoto, “ Chemical vapour deposition of diamond in RFng discharge,” J. Mater. Sci. Lett., Vol. 4, pp. 600-602, 1985. S. Matusmoto, M. Hino, and T. Kobayashi, “ Synthesis of diamond films in a if induction thermal plasma,” Appl. Phys. Lett., Vol. 51, pp. 737-739(September 1987). M. Kohzake, K. Uchida, K. Higuchi. and S. Node," Large-area high- speed diamond deposition by if induction thermal plasma chemical vapor deposition method.” Jpn. J. Appl. Phys., Vol. 32, pp. L438- L440, 15 March 1993. J. J. Chang, T. D. Mantei, R. Vuppuladhadium, and H. E. Jackson, “ Effects of oxygen and pressure on diamond synthesis in a magneto- active microwave plasma,” J. Appl. Phys.,Vol. 59, pp.2918, 1992. 2. Ring, T. D. Matei, S. Tlali, and H. E. Jackson." Plasma synthesis of diamond at low temperature with a pulse modulated magnetoac— tive discharge," Appl. Phys. Lett., Vol. 66, pp. 3380-3382, 12 June 1995. 2. Ring and T. Mantei. " Low-temperature diamond growth in a pulsed microwave plasma," J. Vac. Sci. Technol. Vol. A 13, pp. 1617-1618, May/June 1995. Y. Mitsuda, T. Yoshida, and K. Akashi, “ Development of a new [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] 252 microwave plasma torch and its application,” Rev. Sci. Instrum.,Vol. 60, pp.249-252, February, 1989. K. Takeuchi and T. Yoshida, “ The effect of oxygen on diamond syn- thesis is a microwave plasma jet.” J. Appl. Phys.,Vol. 71, pp.2636- 2639. 15 March 1992. Y. Liou. A. lnspektor. R.Weimer, and R. Messier, “ Low-temperature diamond deposition by microwave plasma-enhanced chemical vapor deposition,” Appl. Phys. Lett.. Vol. 55, pp.631-633, 14 August 1989. Y. Liou, A. lnspektor, R. Weirner, D. Knight. and R. Messier, “ The effect of oxygen in diamond deposition by microwave plasma enhanced chemical vapor deposition.” J. Mater. Res.,Vol. 5, pp.2305-2312, November 1990. P. Bachmann, D. Leers, and H. Lydtin, " Toward a general concept of diamond chemical vapour deposition," Diamond and Related Materials, Vol. 1, pp.1-12, 1991. Y. K. Kim, K. Y. Lee, J. Y. Lee. “ Texture-controlled diamond films synthesized by microwave plasma enhanced vapour deposition.” Thin Solid Filrns.Vol. 272. pp.64-70. 1996. M. C. Salvadori. J. W. Ager III, I. G. Brown, and K. M. Krishnan. “ Diamond synthesis by microwave plasma chemical vapor deposition using graphite as the carbon source.” Appl. Phys. Lett., Vol. 59, pp.2386-2388, 4 November 1991. 2. Feng, K. Komvopoulos, I. G. Brown, D. B. Bog. “ Efiect of graph- ite carbon films on diamond nucleation by microwave-plasma- enhanced chemical-vapor deposition.” J. Appl. Phys., Vol. 74, pp.2841-2849, 15 August 1993. Z. Feng, M. A. Brewer, K. Komvopoulos, I. G. Brown, D. B. Bog‘y, “ Diamond nucleation on unscratched silicon substrates coated with various non-diamond carbon films by microwave plasma-enhanced chemical vapor deposition.” J. Mater. Res.,Vol. 10, pp. 165-174. Jan— uary 1995. M. A. Brewer, I. G. Brown, M. R. Dickinson, J. E. Galvin, R. A. MacGill, and M. C. Salvadori, “ Simple, safe. and economical micro- wave plasma-assisted chemical vapor deposition facility,” Rev. Sci. Instrum., Vol. 63, pp3389-3393, June 1992. S. Khatami, Ph.D Dissertation. Michigan State University, 1997. 253 [52] C. F. M. Borges, S. Schelz, L. St.-Onge, M. Moisan, L. Martinu, “ Silicon contamination of diamond films deposited on silicon sub- strates in firsed silica based reactors.’ J. Appl. Phys., Vol. 79, pp.3290-3298, 15 Marchl996. [53) C. F. M. Borges, V. T. Airoldi, E. J. Corat, M. Moisan, S. Schelz, and D. Guay, “ Very low-roughness diamond film deposition using a sur- face-wave sustained plasma.’ J. Appl. Phys., Vol. 80, pp.6013- 6020, 15 November1996. [54] C. F. M. Borges, L. St.-Onge, M. Moisan, A. Gicquel, “ Influence of process parameters on diamond film CVD in a surface-wave driven microwave plasma reactor:’ Thin Solid Fihns, Vol. 274, pp.3-l7, 1996. [55] X. Zhang, T. Shi, J. Wang, X. Ahang," Oriented growth of a diamond film on Si(100) by hot filament chemical vapor deposition," J. Crystal Growth. Vol. 155. pp. 66-69, 1995. [56] R. F. Clausing, L. Heatherly, L. L. Horton, E. D. Specht, G. M. Begun and Z. L. Wang, " Textures and morphologies of chemical vapor deposited (CVD) diamond," Diamond and Related Materials. Vol. 1, pp. 411-415, 1992. [57] G. Popvici, C. H. Chao, M. A. Prelas, E. J. Charlson and J. M. Meese, " Smooth diamond films grown by hot filament chemical vapor deposition on positively biased silicon substrates," J. Mater. Res., Vol. 10, pp. 2011-2016, August 1995. [58] Q. Chen, J. Yang, and 2. Lin, " Synthesis of oriented textured dia- mond films on silicon via hot filament chemical vapor deposition," Appl. Phys. Lett., Vol. 67, pp. 1853-1855, 25 September 1995. [59] J. M. Olson and M. J. Dawes.” Examination of the material proper- ties and performance of thin-diamond film cutting tool inserts pro- duced by arcjet and hot-filament chemical vapor deposition,” J. Mater. Res. Vol. 11, pp.1765-1775 (Jul 1996). [60] V. J. Trava-Airoldi, E. J. Corat, E. Del Bosco, N. F. Leite, " Hot fila- ment scaling-up for CVD diamond burr manufacturing." Surface and Coating Technologr, Vol. 76-77, pp. 797-802, 1995. [61] A. Johnson, “ CVD Diamond Inserts Stick.” Cutting Tool Engineer- ing. pp. 34-39, Feb 1996. [62] P. K. Bachmann and R. Messier, " Emerging Technology of diamond [63] [64] [651 [66] [67] [68] [69] [70] [71] [72] [73] 254 thin films," C&EN, Vol. 24, 15 May 1989. R. B. Jackrnan, J. Beckrnan, J. S. Foord, " Diamond chemicak vapor deposition fiom a capacitively coupled radio frequency plasma," Appl. Phys. Lett.,Vol. 66, pp. 1018-1020, 20 February 1995. S. P. Bozeman, D. A. Tucker, B. R. Stoner, J. T. Glass, and W. M. Hooke. “ Diamond deposition using a planar radio fiequency induc- tively coupled plasma,”Appl. Phys. Lett, Vol. 66. pp. 3579- 3581(June 1995). W. Zhu, B. R. Stoner, B. E. Willams, and J. T. Glass, " Growthand Characterization of diamond films on nondiamond substrates for electronic applications." Proceedings of IEEE, Vol. 79, pp. 621-646, May 1991. A. Masood, M. Aslam, M. A. Tamor and T. J. Potter, J. Electrochem. Soc.. Vol. 138. pp.L67 (1991). J. Mossbrucker, Ph.D Dissertation, Michigan State University, 1997. P. Hendra, C. Jones, and G. Warnes,_EQuneLILan§fQ:m_Baman_ W, Horwood, Chichester(1991). P. W. Morrison, A. Somashekhar, J. T. Glass, and J. T. Prater, “ Growth of diamond films using an enclosed combustion flame,” J. Appl. Phys.,Vol. 78, pp.4144-4156(1995). H. C. Barshilia, B. R. Mehta, and V. D. Vankar, “Growth of Diamond Thin films by Microwave Plasma Chemical Vapor Deposition,” J. Mater. Res., Vol. 11, pp.1019-1024(April 1996). S. L. Flegler. W. R. F. Clausing, L. Heatherly, L. L. Horton, E. D. Specht, G. M. Begun, and Z. L. Wang, “ Textures and morphologies of chemical vapor deposited(CVD) diamond,” Diamond and Related Materials, pp.4l 1-415(1992). H. Maeda, K. Ohtsubo, M. Irie, N. Ohya, K. Kusakabe, and S. Morooga, “ Determination of diamond [100] and [1 1 1] growth rate and formation of highly oriented diamond films by microwave plasma assisted chemical vapor deposition,” J. Mater. Res. Vol. 10, pp.31 15-3123(Dec 1995). [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] 255 M. J. Ulczynski, D. K. Reinhard, M. Prystajko and J. Asmussen, WW. Ed. S. Sa1t0. N. Fujirnori, O. Fukunagu and M. Kamo, pp.4l(1994). S. Khatami and J. Asmussen, 23rd IEEE International Conference on Plasma Science, Boston, Massachusetts. June 1996. A. Gucquel, E. Anger, M. F. Ravet, D. Fabre, G. Scatena, and Z. Z. Wang, “ Diamond deposition in a bell jar reactor: influence of the plasma and substrate temperatwe on the microstructure and growth rate,” Diamond and related Materials,Vol. 2, pp.417-424 (1993.) W. Zhu, “ Microwave plasma-enhanced chemical vapor deposition and structural characterization of diamond films,” Ph.D Disserta- tion, Penn. State University, 1989. S. J. Harris and A. M. Weiner, “ Pressure and temperature effects on the kinetics and quality of diamond films,” J. Appl. Phys.,Vol. 75, pp.5026-5032(May1994). M. P. Everson and M. A Tamor, “ Investigation of growth rates and morphology for diamond growth by chemical vapor deposition,” J. Mater. Res., Vol. 7, pp.1438-1443(June 1992). N. Setake, “ Low pressure gas phase synthesis of diamond,” Hyo- men (Surface),Vol. 22, pp.1 10 (1984). J. Wei, H. Kawarda, J. Suzuki, and A. Hiraki, “ Growth of diamond films at low pressure using Magneto-microwave plasma CVD,” Journal of Crystal Growth,Vol. 99, pp. 1201-1205(1990). A. Gicquel, K. Hassouni, S. Farhat, Y. Breton. C. D. Scott, M. Lefe- bvre, and M. Pealat, “ Spectroscopic analysis and chemical kinetics modeling of a diamond deposition plasma reactor” Diamond and Related Materials. Vol. 3, pp.581-586(1994). Y. K. Kim, K. Y. Lee, J. Y. Lee. “ Texture-controlled diamond films synthesized by microwave plasma enhanced vapour deposition,” Thin Solid Films,Vol. 272, pp.64-70 (1996). C. J. Chu, R. H. Hauge, J. L. Margrave, and M. P. D’ Evelyn, “ Growth kinetics of (1 00), (1 10), and (1 1 1) homospitaxial diamond films,” Appl. Phys. Lett,Vol. 61, pp.1393-1395 (September 1992). L. K. Bigelow and N. M. henderon, 2nd Int. Conf. on the Diamond [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] 256 Sci. and Tech., ICNDST-2. Washiongton, D.C., Sep. 23, 1990. R. A. Rudder, R. E. Thomas, R. J. Markaunas, 179th Meeting of Electrochemical Society, Washington, D.C., May 5-10, 1991. R. F. Davis. WW5. edited by R. F. Davis. Noyes publications. P. K. Bachmann,” Emerging Technology of diamond thin films,” Sci- ence,pp.24-39 (15 May 1989). G. Johnson and O. L. Sparkman, “ The future of small hole drill- ing?” Drilling Advances, Vol. 19, No.3, pp.40-44(3 March 1996). B. Schulte, H. Juergensen, P. Blackborow, and E. Sevillano,” Flexi- bility of microwave plasma chemical vapour deposition for diamond growth,” Surface and Coatings Technology,Vol. 74-75, pp.634- 636 (1995). ' M. Portland, “ Diamonds could be high tech’s best fiiend,” Business Week. pp.90-91(February 12 1990). J. T. C. Lee, “ A comparison of HDP sources for polysilicon etching,” Solid State Technology, cover article, pp. l-4(August 1996). K. Takeuchi and T. Yoshida,” The effect of oxygen on diamond syn- thesis in a microwave plasma jet,” J. Appl. Phys..Vol. 71, pp.2636- 2639 (March1995). S. W. Reeve, W. A. Weirner, and D.S. Dandy,” On the optimization of a dc arejet diamond chemical vapor deposition reactor,” J. Mater. Res., Vol. 11, pp.694 (March 1996). A. Higa, A. Hatta, T. Ito, T. Maehama,” Homoepitaxial diamond syn- thesis by dc arc plasma jet chemical vapor deposition,” Jpn. J. Appl. Phys..Vol. 35. pp.L577-L580 (May 1996). A. Higa, A. Hatta, T. Ito, M. Toguchi, and A. Hiraki, “ Efl’ect of C02 addition on diamond growth by DC arc plasma jet chemical vapour deposition,” Jpn. J. Appl. Phys., Vol. 35. pp.216-220(l996). F. Deuerler, M. Pies, H. Van Den Berg, R. Tabersky, “ Production, characterization, and wear behaviour of plasma jet CVD diamond films on hard metal cutting tools,” Phys. Stat. Sol., Vol. 154, pp.403-423(1996). [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [l 10] [111] 257 M. Nesladek, K. Vandierendonck, C. Quaeyhaegens, M. Kerlchofs, and L. M. Stals, “ Adhesion of diamond coatings on cemented car— bides,” Thin Solid Films, Vol. 270, pp.184-188(1995). A. K. Mehlrnann, A. Fayer, and S. F. Dirnfeld, “ nucleation and growth of diamond on cemented carbises by hot-filament chemical vapor deposition,” Diamond and Related Materials, Vol. 2, pp.3l7- 322(1993). J. Ting and M. L. Lake, “ Diamond-coated carbon fiber,” J. Mater. Res., Vol. 9, pp.636-642(March 1994). J. Ting, A. G. Lagounov, M. L. Lake, “ Chemical vapour infiltration of diamond into a porous carbon,” J. Mater. Sci. Lett., Vol. 15(1996). Y. Saito, I. Isozaki. A. Masuda, K. Fukumoto. M. Chosa, and T. Ito, “ Adhesion strength of diamond film on cemented carbide insert,” Diamond and Related Materials, Vol. 2, pp. 1391- 1395(1993). Y. Muranaka, H. Yamashita, K. Sato, and H. Miyadera, “ The role of hydrogen in diamond synthesis using a microwave plasma in a CO/H2 system,” J. Appl. Phys., Vol. 167. pp.6247-6254(May 1990). purchased from Travel Tool Company, CT. purchased from BFGoodrich Super Temp., Santa Fe Springs, CA. Macro-Raman Spectroscopic equipment. Environmental Scanning Electron Microscopy, Model 2020 # 101, purchased from ElectroScan, Wilmington, MA. Any thing related to the graphite signal stronger than diamond in Raman. J. Zhang and J. Asmussen, U.S. Patent Number 5,31 l, 103. J. Zhang and J. Asmussen, U.S. Patent Number 5,571.57. G. L. King, “ Temperature and concentration of ionic and meutral species in resonant microwave cavity plasma discharge,” Ph. D dissertation, Michigan State University (1994). 258 [1 12] J. Asmussen and R. Garard,“ Precision Microwave Applicators and Systems for Plasma and Material Processing” Microwave; W, Ed. W.A. Sutton, M. H. Brooks and I.J. Chabinsky. [1 13] J .Asmussen, “Microwave Plasma Disk Processing Machine” in , Ed. O.A Popov, 251-311 (Noyes Publ., NJ. 1995). [1 14] J. Asmussen and J. Root, “ Ion generating apparatus and method for the use thereof,” U.S. Patent No. 4,507,588 ( 26 March, 1985). [1 15] J. Asmussen and D. K. Reinhard, “ Method for treating a surface with a microwave plasma and improved apparatus,” U.S. Patent No. 4.585.688 (29 April 1986). [l 16] H. M. Relyea, F. Breidenich, J. V. Beck, and J. J. Mcgrath, “ Mea- surement of thermal properties of thin films using infrared ther- mography,” pp. 183-194, Thermal Conductivity, Vol. 23, Edited by K. E. Wilkes, R. B. Dinwiddie, R. S. Graves, Oak Ridge National Laboratory, Oak Ridge, Tennessee. (Technomic Publishing Com- pany, Inc., 1996).