Stuart M. Leeds.
A thesis submitted to the University of Bristol in accordance with the requirements for the degree of Doctor of Philosophy in the Faculty of Science, Department of Physical Chemistry.
This thesis describes the construction and use of a molecular beam mass spectrometer to obtain gas-phase compositional measurements from a microwave plasma chemical vapour deposition (CVD) reactor used for diamond growth. Molecular beam mass spectrometry (MBMS) allows simultaneous, in situ detection of stable and highly reactive species (such as radicals) from the plasma region. The design was such that the mass spectrometer sampled gas from the side of the plasma region, and therefore differs in approach from previous studies. In addition, diamond growth experiments have been performed under the same conditions as the MBMS experiments, and these deposits have been characterised by Scanning Electron Microscopy (SEM), Laser Raman Spectroscopy (LRS), and Secondary Ion Mass Spectrometry (SIMS).
The MBMS system has been used to examine the effect of the hydrocarbon/H2 gas mixtures CH4/H2, C2H2/H2, and C2H4/H2 on gas-phase composition. Experiments have been performed using a constant input C:H2 ratio of 2%, while varying the microwave power used to sustain the plasma. It was found that the gas-phase composition became independent of the nature of the hydrocarbon source species provided the microwave power level was above ~ 700 W. Below this power level, the measured composition was dependent on the nature of the hydrocarbon source used. The effect that the input mole fraction of each of these three hydrocarbons (in an H2 feed) has on gas-phase composition was studied, and it was found that at higher microwave power levels the C2H2 mole fraction scales as the square of the CH3 mole fraction for all three gases. The dependence of the gas-phase composition on the other CVD deposition parameters (pressure, and substrate holder temperature and position) has also been elucidated. All MBMS results showed that the total measured carbon did not add sum to the total input, but was instead a factor of 3-5 times smaller. This observation has been ascribed to mass dependent thermal diffusion, in common with previous mass spectrometric studies of diamond CVD.
The effect of nitrogen addition to a typical 1% (carbon containing gas)/H2 CVD gas mixture has been studied. Nitrogen was added as N2, NH3, CH3NH2, and HCN. In the case of N2 and NH3, CH4 was used as the carbon source. It was found that N2 and HCN are very stable under the plasma conditions employed here, whilst NH3/CH4/H2 and CH3NH2/H2 gas mixtures lead to the formation of large quantities of HCN at high microwave powers. At low microwave powers, the gas-phase composition was dependent on the nature of the feed gas constituents, as found using hydrocarbon/H2 feed gas mixtures. The difficulty in dissociating N2 in the present type of plasma system has been related to the low N incorporation ratio found when using N2 additions to p-dope CVD diamond during growth. A radical species, with a mass-to-charge ratio of 28 was observed between 200-600 W applied microwave power, and was believed to be HCºNH+ formed in the mass spectrometer by dissociative ionisation of an intermediate species in the formation of HCN. Diamond films grown from an N2/CH4/H2 gas mixture were examined for nitrogen incorporated into the film bulk by the technique of SIMS. In common with other such studies, a very low incorporation ratio was found.
The MBMS system has been used to determine neutral gas temperatures in the plasma region by using the well characterised mass spectrometer response to argon, as a function of temperature, in a 2% Ar/H2 gas mixture. Temperatures obtained by this method are in good agreement with a variety of non-intrusive (optical) methods, and intrusive methods (thermocouples). The plasma temperature was found to be ~1400 K at 1000 W microwave power. A second sampling probe (10 mm shorter) has been used to repeat this experiment and the temperatures obtained were significantly lower (400-600 K), indicating a large temperature gradient in the region of the visible plasma edge. This temperature gradient is the likely cause of the observed mass dependent thermal diffusion effects.
I would like to thank the following people for their help over the last three years. Their individual contributions are many and varied, and without the help of many of them, this project could not have proceeded.
I must firstly thank my advisor, Dr Paul May, for his help and advice on all aspects of the project. In addition to being a most excellent advisor, Paul has been a great friend and a shaft of shining wit, his humour making the lab a pleasant place to work. Dr Roland Tsang deserves a huge thank you for teaching me to use the mass spectrometer when it was a part of his experiment. Roly also deserves a great ‘Chhuuuuurrrs then’ for his sense of humour, his friendship, and legendary cocktail-making skills. Professor Mike Ashfold has provided invaluable help and discussion during the project, and I thank him for the time he has taken to give a critical reading of the first draft of this thesis. Mr Keith Rosser deserves thanks for his wisdom and help with the experiment over the last three years, particularly when things went wrong! Many thanks to Liz Bartlett for her help with the MBMS experiment over the previous year.
I could not forget the contemporaries with whom I have shared lab. space over the last three years - Rob Lade, Kevin Kuo, Matt Latto, Marcus Elliot, and Stefan ‘Sorry ’bout that’ Höhn.
Thanks to the staff of the School of Chemistry workshops who have had to fabricate a range of esoteric objects for me, some successful, some not, over the last three years. Mr John Dimery deserves credit for running the SEM, and Mr David Jones for all his effort in printing the electron micrographs herein.
I would like to thank all the Physics BUDGies, past and present, for joining in on all the crazy BUDGie capers, showing that diamond research at Bristol is truly multidisciplinary. Dr Tim Davis and Dr Dave Pickard, in particular deserve especial thanks for their help with the laser Raman work.
I would also like to thank all the friends I have made in Bristol over the last three years, especially those ex-‘12 Wav’-ers, fellow housemates at 8 Warwick Rd, Louise, Joel & Helen, Chris & Karen, and Chris H.
I would like to thank the owners and patrons of the many Bristol alehouses within whose hallowed portals I have slaked my thirst on many an eve. An especial mention must go to Mick & Jan at the Coronation Tap. Few people go to the pub and find God making an Exhibition of himself, but at the Corrie it’s an everyday occurrence (he collects the glasses).
Last but certainly not least, I would like to thank my parents for supporting me throughout my career thus far.
Stuart M. Leeds, April 1999.
“Ah Mr. Gibbon, another damned, fat, square book.
Always scribble, scribble scribble, eh?”
The Duke of Gloucester, on being presented with volume 2 of
‘The Decline and Fall of the Roman Empire’.
“A final point to consider is an aspect of molecular beam research that is often overlooked - the basic interdisciplinary nature of the activity normally occurring in a typical molecular beam laboratory. The molecular beam scientist needs to be a skilled engineer in designing the machines, a refined physicist in setting up the technology of each experiment, and often an expert chemist in carrying it out. Because few people are all of these, it follows that nowhere is the mixing of people of different backgrounds so immediately profitable as in the molecular beam laboratory. Unfortunately, the rigid departmental structure of our educational institutions makes this mixing difficult to achieve.”
Giacinto Scoles, in the introduction to ‘Atomic and Molecular Beam Methods’, Vol. I.
“Græcum est, non legitur”
For my family.
The research described in this thesis was carried out by the author in the School of Chemistry at the University of Bristol under the supervision of Dr. P. W. May and Professor M. N. R. Ashfold.
The work reported herein is original to the author, except where acknowledged by reference or special recognition. No part of this work has been submitted previously for any degree.
Stuart M. Leeds.
1.1 The Diamond in History 1
1.2 Diamond: The Material 1
1.3 Chemical Vapour Deposition 4
1.4 Microwave Plasma CVD 6
1.4.1 Waveguide Transmission of Microwaves 6
1.4.2 Plasma Generation Process 7
1.4.3 Plasma Physical Properties 8
1.4.4 ASTeX-type Microwave Plasma CVD Reactor 9
1.5 CVD Diamond Films 12
1.5.1 Growth Rates 14
1.5.2 Film Quality 14
1.6 The Substrate Material 14
1.7 Nucleation Processes 15
1.8 Characterisation of Diamond CVD 16
1.9 Studies of HFCVD Gas Phase Chemistry 18
1.9.1 Early Studies 18
1.9.2 Molecular Beam Mass Spectrometry 19
1.9.3 Summary of early HFCVD gas phase composition studies 22
1.9.4 The Role of Radical species 23
1.9.5 Temperature Effects 23
1.9.6 The Diamond Growth Species 26
1.10 Studies of Microwave Plasma Chemistry 26
1.10.1 Optical Emission Spectroscopy studies 27
1.10.2 Actinometry 28
1.10.3 Fourier Transform Infra-Red Spectroscopy studies 28
1.10.4 Mass Spectrometry studies 29
1.10.5 Absorption Spectroscopy studies 30
1.10.6 MBMS studies 30
1.11 Summary 31
1.12 References 31
2.1 Introduction 39
2.2 Laser Raman Spectroscopy (LRS) 39
2.3 Scanning Electron Microscopy (SEM) 45
2.4 Secondary Ion Mass Spectrometry (SIMS) 46
2.5 References 47
3.1 Introduction 50
3.2 Experimental System used for Diamond Film Growth 52
3.2.1 Microwave Circuitry 52
3.2.2 Microwave Plasma Reactor: Vacuum System Design 54
3.2.3 Substrate Holder / Heater Design 55
3.2.4 Exhaust and Pressure Control 59
3.2.5 Gas Supply and Handling 60
3.2.6 Preparation of Samples for Diamond Deposition 63
3.2.7 Procedure for Diamond Deposition 63
Problems encountered during Diamond Growth
184.108.40.206 Secondary Plasma Formation 65
220.127.116.11 Plasma Instability 66
3.3 Experimental System used for MBMS 67
3.3.1 Sampling Probe Design,
and Coupling to Microwave Plasma Reactor 68
3.3.2 Molecular Beam Mass Spectrometer: First Stage 70
3.3.3 Molecular Beam Mass Spectrometer: Second Stage 70
3.3.4 Considerations for MBMS Studies 71
18.104.22.168 Temperature Dependence of the
Sensitivity Factor, SEXP 72
22.214.171.124 Mass Discrimination 74
3.4 Hiden Analytical HAL/3F PIC
Quadrupole Mass Spectrometer 74
3.4.1 HAL/3F PIC 100 QMS: Vacuum components 75
126.96.36.199 Source 76
188.8.131.52 Quadrupole Mass Filter 76
184.108.40.206 Detector 77
3.4.2 HAL/3F PIC 100 QMS:
Computerised Operating System 78
220.127.116.11 STATUS menu 79
18.104.22.168 TRIP menu 80
22.214.171.124 BAR menu 80
126.96.36.199 MID (Multiple Ion Detection) menu 82
188.8.131.52 SETUP menu 83
184.108.40.206 TUNE menu 83
3.5 Mass Spectrometer Characterisation 86
3.5.1 Energy Scale Calibration 87
3.5.2 Threshold Ionisation Technique 88
3.5.3 Mass Spectrometer Pressure Considerations 90
3.6 Experimental Procedure for using the MBMS 92
3.7 Mass Spectrometric Data Collection 94
3.7.1 Notation 94
3.8 Data Analysis Procedure 95
3.8.1 Background Subtraction 96
3.8.2 Temperature Correction 99
3.8.3 Correction for Cracking Products 100
3.8.4 Room Temperature Species Calibration 101
3.8.5 Improved Calibration Procedure 102
3.8.6 Calibration of Radical Species 106
3.9 Reaction of Plasma and Reactor Operating Parameters
to the Presence of the Sampling Probe 108
3.10 Mass Spectrometer Problems Encountered 110
3.11 References 113
4.1 Introduction 115
4.2 Experimental Considerations 116
4.3 Effect of Deposition Time 118
4.4 Effect of Hydrocarbon Mole Fraction in H2 120
4.4.1 Growth Rates 120
4.4.2 Film Quality 122
4.4.3 MBMS Gas Phase Studies 126
4.5 Effect of applied Microwave Power 138
4.5.1 Growth Rates 139
4.5.2 Film Quality 141
4.5.3 MBMS Gas Phase Studies 141
4.6 Effect of Reactor Pressure 148
4.6.1 Growth Rates 148
4.6.2 Film Quality 150
4.6.3 MBMS Gas Phase Studies 150
4.7 Effect of Substrate (Vertical) Position 154
4.7.1 Growth Rates 156
4.7.2 MBMS Gas Phase Studies 158
4.8 Effect of Substrate Temperature 159
4.8.1 Growth Rates 160
4.8.2 Film Quality 161
4.8.3 MBMS Gas Phase Studies 161
4.9 Effect of Probe Length - Probe 2 MBMS measurements 163
4.10 References 165
5.1 Introduction 169
5.2 The source of Plasma Heating 170
5.3 Gas Temperature Determination by MBMS 173
5.4 Gas Temperature as a Function
of Microwave Power - Probe 1 175
5.5 Gas Temperature as a Function
of Microwave Power - Probe 2 177
5.6 Vertical Plasma Temperature Profile 178
5.7 Effect of Substrate Temperature
on Gas Phase Temperature 181
5.8 References 183
6.1 Introduction 185
6.2 Effect of applied Microwave Power 187
6.2.1 1% CH4 / 0.5% N2 / H2 Gas Mixture 190
6.2.2 1% CH4 / 1% NH3 / H2 Gas Mixture 192
6.2.3 1% CH3NH2 / H2 Gas Mixture 195
6.2.4 0.5% HCN / H2 Gas Mixture 197
6.3.1 Effect of Nitrogen in the gas-phase
using N2 as Nitrogen source 199
6.3.2 Effect of Nitrogen in the gas phase
using NH3 as Nitrogen source 201
6.4 References 204