close

Вход

Забыли?

вход по аккаунту

?

Extraction of bio-flocculant from okra using hydrothermal and microwave extraction methods combined with a techno-economic assessment

код для вставкиСкачать
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UM I a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize m aterials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.
U niversity Microfilms International
A Bell & Howell Information C o m p an y
3 0 0 N orth Z e e b R o ad . Ann Arbor. Ml 4 8 1 0 6 -1 3 4 6 USA
3 1 3 /7 6 1 -4 7 0 0 8 0 0 /5 2 1 -0 6 0 0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O rd e r N u m b e r 9326790
E xperim ental developm ent o f microwave cavity plasm a reactors
for large area and high rate diam ond film deposition
Zhang, Jie, Ph.D.
Michigan State University, 1993
Copyright ©199S b y Zhang, Jie. All rig h ts reserved.
300 N. Zeeb Rd.
Ann Arbor, MI 48106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
EXPERIMENTAL DEVELOPMENT OF
MICROWAVE CAVITY PLASMA REACTORS FOR
LARGE AREA AND HIGH RATE DIAMOND FILM DEPOSITION
By
Jie Zhang
A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Electrical Engineering
1993
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
EXPERIMENTAL DEVELOPMENT OF
MICROWAVE CAVITY PLASMA REACTORS FOR
LARGE AREA AND HIGH RATE DIAMOND FILM DEPOSITION
By
Jie Zhang
Diamond, w ith its unique mechanical, therm al, chemical, optical
and electrical properties, is an attractive m aterial in applications ranging
from w ear-resistant coatings for m echanical and optical components to
substrates for advanced semiconductor devices. This research concerns
the experimental investigation and development of microwave cavity
plasm a reactors for large area and high rate diam ond film deposition.
The research and development have lead to th e successful creation
of a new improved prototype reactor. Diamond films with uniformities
b etter th an 2% have been deposited on 3” and 4” silicon wafers with this
prototype reactor. The linear growth rates obtained are 0.89 p m /h o u r on
2 ” silicon wafers, 0.67 p m /h o u r on 3" silicon wafers and 0.43 p m /h o u r
on 4 ” silicon wafers. The performance “figures of merit" were developed to
quantitatively compare diamond film deposition reactors in term s of
linear growth rate, weight gain rate, deposition area, energy efficiency,
gas flow efficiency an d carbon conversion efficiency. Three generations of
microwave cavity plasm a reactors have been investigated a n d /o r
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
developed for diamond film deposition over larger su b strate areas. The
new prototype reactor employs an end feed coupling concept which
enables the reactor to accept high microwave power in p u t (from 1.5 kW
to 4.5 kW) and create discharges up to 12.5 cm in diameter. Three
generations of microwave cavity je t reactors have also been developed
a n d /o r designed. These reactors have th e potential to deposit diamond
films w ith higher growth rates. A method to obtain the operational
characteristics of th e microwave cavity plasm a reactors was developed.
The reactor was characterized by a set of input, internal and output
experim ental param eters. The relationship betw een th e internal and
in p u t experimental param eters w as established to locate the required
experim ental conditions w hen diamond films w ith desired properties,
su ch as film thickness, morphology, and uniformity, etc. are deposited.
The diamond films deposited in the prototype reactor were investigated
w ith respect to their growth rate and morphology a s th e in p u t
experim ental param eters were varied. The electric fields in the reactor
were also m easured to develop a better understanding of the
electromagnetic field/plasm a interactions during th e diamond film
deposition process. It w as found th a t the tangential com ponent of the
electric field is the m ain discharge excitation field.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Copyright by
JIE ZHANG
1993
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To my parents
Zu-Di Zhang and Xin-Xuan Wu
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
The author wishes to th a n k Dr. J e s A smussen, J r. for his
encouragem ent, guidance and su p p o rt throughout this research. Special
th a n k s is extended to Dr. Donnie Reinhard, Dr. Timothy Grotjohn and
Dr. Chow Ling Chang of Norton for th eir valuable discussions. Additional
th a n k s is given to Mr. Jero n Campbell for h is experimental laboratory
help an d to Dr. Aslam for his com m ents on th e m anuscript.
This research was supported in p a rt by grants from Norton, Corp.,
Wavemat, Corp., and Michigan Research Excellence fund.
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
LIST OF T A B L E S.............................................................................................. xiii
LIST OF F I G U R E S .........................................................................................xiv
Chapter One
Introduction
1.1
1.2
1.3
1.4
I n t r o d u c t i o n .....................................................................................
Research O b je c tiv e s..........................................................................
R esearch H is to r y ...............................................................................
D issertation O u tlin e ..........................................................................
1
5
5
8
Chapter Two
Diamond Film Deposition Reactors: A Review
2.1
I n t r o d u c t i o n ...........................................................................................10
2.1.1 A Generic Diam ond Film Deposition Reactor
. . . .
10
2.1.2 Reactor C a te g o riz a tio n ................................................................... 12
2.1.3 “Figures of M erit” of Reactor P e rfo rm a n c e..................................12
2.1.4 Notes on Literature R e v i e w ........................................................ 13
2.2 A Review of Diamond Film Deposition R e a c t o r s .............................15
2.2.1 Thermally Activated CVD R e a c t o r ............................................. 15
2.2.2 Direct C u rren t Discharge CVD R e a c t o r ..................................17
2.2.2.1 Direct C urrent Discharge R e a c t o r ..................................17
2.2.2.2 Direct C urrent Discharge J e t R e a c t o r ............................20
2.2.3 Com bustion Flame R e a c t o r ........................................................ 22
2.2.4 Radio Frequency (RF) Plasm a CVD R e a c t o r s ............................24
2.2.4.1 RF Glow Discharge R e a c t o r .............................................24
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.2.4.2 RF Induction Thermal Plasm a R e a c to r ...........................26
2.2.5 Microwave Plasm a Enhanced CVD R e a c t o r s .......................... 28
2.2.5.1 I n t r o d u c t i o n ....................................................................... 28
2.2.5.2 T ubular Microwave R e a c t o r ............................................ 29
2.2.5.3 Magneto-microwave R e a c t o r ............................................ 31
2.2.5.4 Magneto-active Microwave Reactor
...........................33
2.2.5.5 Bell-jar Microwave R e a c t o r ............................................35
2.2.5.6 Microwave J e t R e a c t o r .......................................................37
2.3 Com parison of Diamond Film Deposition Reactors . . . .
40
2.4 B achm ann C-H-O Phase D i a g r a m ................................................... 43
Chapter Three
The Microwave-Cavity Plasm a Reactors (MCPR)
3.1
3.2
I n t r o d u c t i o n ......................................................................................... 45
Experim ental S y s t e m s .........................................................................46
3.2.1 I n t r o d u c t i o n ...................................................................................46
3.2.2 Microwave Power Delivery S y s t e m s ............................................ 47
3.2.3 Flow Control and Vacuum Pump S y s t e m s ........................... 49
3.2.4 Com puter Monitor System
........................................................53
3.3 The First G eneration Seven Inch M C P R ....................................... 57
3.3.1 I n t r o d u c t i o n ................................................................................... 57
3.3.2 Reactor G e o m e t r y ........................................................................57
3.3.3 Reactor O p e r a t i o n ........................................................................59
3.3.4 S u b strate Holder and Quartz D o m e s .......................................63
3.3.5 Reactor P e r f o r m a n c e .................................................................. 64
3.4 The Second G eneration Seven Inch M C P R ..................................67
3.4.1 I n t r o d u c t i o n ...................................................................................67
3.4.2 Reactor G e o m e t r y ........................................................................67
3.4.3 T./Ioin Modes E x c i t a t i o n .............................................................69
3.4.4 End Feed A s s e m b ly ........................................................................ 71
3.4.5 Reactor P e r f o r m a n c e .................................................................. 72
3.5 Microwave Coupling M e th o d s..............................................................75
3.5.1 I n t r o d u c t i o n ...................................................................................75
3.5.2 Loop Coupling and Probe C o u p lin g ............................................ 75
3.5.3 End Coupling and Side C o u p l i n g ............................................ 78
3.5.4 Mode E x c i t a t i o n ............................................................................. 79
3.6 The Third G eneration Seven Inch M C P R ....................................... 80
3.6.1 I n t r o d u c t i o n ..................................................................................80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.6.2 Reactor C o n f ig u r a tio n .................................................................. 80
3.6.2.1 I n t r o d u c t i o n ........................................................................ 80
3.6.2.2 The Basic C o n f i g u r a t i o n ..................................................80
3.6.2.3 The Forced Flow C o n f ig u r a tio n .......................................84
3.6.2.4 The J e t C o n f ig u r a tio n ....................................................... 87
3.6.2.5 The Lower Cavity Resonance Configuration . . .
87
3.6.2.6 The Down-stream C o n fig u ratio n .......................................90
3.6.3 Reactor O p e r a t i o n ........................................................................ 90
3.6.4 Reactor P e r f o r m a n c e .................................................................. 90
3.6.4.1 I n t r o d u c t i o n ........................................................................ 90
3.6.4.2 The Basic C o n f i g u r a t i o n ................................................. 92
3.6.4.3 The Forced Flow C o n f ig u r a tio n ...................................... 92
3.6.4.4 The J e t C o n f ig u r a tio n .......................................................93
3.6.4.5 The Lower Cavity Resonance Configuration . . .
95
3.6.4.6 The Down-stream C o n fig u ratio n ...................................... 95
3.7 Comparison of Seven Inch M C P R s ....................................................97
3.8 Fourteen and Eighteen Inch M C P R s ..............................................99
3.8.1 I n t r o d u c t i o n ................................................................................... 99
3.8.2 Fourteen Inch M C P R .................................................................. 99
3.8.3 Eighteen Inch M C P R ................................................................ 102
Chapter Four
The Microwave-Cavity J e t Reactors (MCJR)
4.1
4.2
I n t r o d u c t i o n ........................................................................................ 106
Experimental S y s t e m s ....................................................................... 107
4.2.1 I n t r o d u c t i o n ................................................................................. 107
4.2.2 Microwave Power Delivery S y s t e m ...........................................107
4.2.3 Flow Control and V acuum Pump Systems
..........................108
4.3 The First Generation M C J R ........................................................ 110
4.3.1 I n t r o d u c t i o n ................................................................................. 110
4.3.2 Reactor G e o m e t r y ...................................................................... 110
4.3.3 Reactor O p e r a t i o n ...................................................................... 113
4.3.4 Reactor P e r f o r m a n c e .............................................................114
4.4 The Second Generation M C J R ...................................................116
4.4.1 I n t r o d u c t i o n ................................................................................. 116
4.4.2 Reactor G e o m e t r y .................................................................. 116
4.4.3 Reactor P e r f o r m a n c e .............................................................119
4.5 The Third G eneration M C J R ............................................................121
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.5.1 I n t r o d u c t i o n ................................................................................. 121
4.5.2 Reactor G e o m e t r y ...................................................................... 121
4.5.3 R e m a r k s ....................................................................................... 124
C hapter Five
Microwave Electric Fields in the MCPR
5.1 I n t r o d u c t i o n ......................................................................................... 125
5.2 Experimental M easurem ent System
.............................................127
5.3 M easurement R e s u l t s ........................................................................ 133
5.3.1 Electromagnetic Field P a t t e r n s ................................................ 133
5.3.2 Power Balance and Discharge Power D e n s ity .......................... 137
5.3.3 Electric Field S t r e n g t h .................................................................141
5.3.4 Substrate T e m p e ra tu re .................................................................144
5.4 Discussion
.........................................................................................147
5.5 S u m m a r y .............................................................................................. 148
C hapter Six
Diamond Film Deposition in the MCPR
I n t r o d u c t i o n .........................................................................................150
Experimental M e t h o d o l o g y .............................................................152
6.2.1 I n t r o d u c t i o n ................................................................................. 152
6.2.2 Independent, In p u t Experimental Param eters
. . . .155
6.2.3 Internal, D ependent Experimental Param eters . . . .
156
6.2.4 External, O utput Experimental P a r a m e t e r s .......................... 157
6.3 Experimental Operational Characteristics
................................. 159
6.3.1 I n t r o d u c t i o n ..................................................................................159
6.3.2 Experimental C o n f ig u r a tio n ......................................
159
6.3.3 Common Experimental P r o c e d u r e s ...........................................162
6.3.3.1 Seeding P r o c e d u r e s ........................................................... 162
6.3.3.2 Start-up and Shut-dow n P r o c e d u r e s .......................... 164
6.3.4 Experimental Operational C h a ra c te ristic s................................164
6.3.5 S u m m a r y ....................................................................................... 169
6.4 Diamond Film Deposition on 3" Silicon W a f e r s ........................... 173
6.4.1 I n t r o d u c t i o n ..................................................................................173
6.4.2 Film Growth Rate and Growth E ffic ie n c y ................................174
6.4.2.1 Effects of S ubstrate T e m p e r a t u r e ................................174
6.1
6.2
x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.4.2.2 Effects of CH4 C o n c e n tra tio n ........................................... 183
6.4.2.3 Effects of CO2 C o n c e n tra tio n ........................................... 187
6.4.2.4 Effects of Total Flow R a t e ................................................ 190
6.4.3 Film M o rp h o lo g y ............................................................................194
6.4.3.1 I n t r o d u c t i o n ...................................................................... 194
6.4.3.2 Effects of S ubstrate T e m p e r a t u r e ............................... 194
6.4.3.3 Effects of CH4 C o n c e n tra tio n .......................................... 200
6.4.3.4 Effects of CO2 C o n c e n tra tio n .......................................... 207
6.4.3.5 Effects of Total Flow R a t e ................................................207
6.4.4 Typical Ram an S p e c t r u m ...........................................................215
6.4.5 Typical Uniformity P r o f i l e ...........................................................215
6.4.6 S u m m a r y ...................................................................................... 222
6.5 Study of Experimental Set-ups with 4" Silicon Wafers . . . 223
6.5.1 I n t r o d u c t i o n .................................................................................223
6.5.2 Cavity Shell G e o m e tr y ................................................................ 224
6.5.3 Q uartz Dome G e o m e t r y ...........................................................225
6.5.4 S ubstrate L o c a t i o n ..................................................................... 226
6.5.5 Cavity Mode E x c ita tio n ................................................................ 228
6.5.6 Seeding D e n s i t y ....................................................................... 230
6.5.7 S u m m a r y ...................................................................................... 231
6.6
Diamond Film Deposition on 4" Silicon W a f e r s ......................... 232
6.6.1 I n t r o d u c t i o n .................................................................................232
6.6.2 Film Growth Rate and Growth E ffic ie n c y ...............................233
6.6.2.1 Effects of S ubstrate T e m p e r a t u r e ...............................233
6.6.2.2 Effects of CH4 C o n c e n tra tio n .......................................... 235
6.6.2.3 Effects of CO2 C o n c e n tra tio n .......................................... 238
6.6.2.4 Effects of Total Flow R a t e ................................................241
6.6.3 Film M o rp h o lo g y ........................................................................... 241
6.6.3.1 I n t r o d u c t i o n ...........................
241
6 .6.3.2 Effects of S ubstrate T e m p e r a t u r e ...............................244
6.6.3.3 Effects of CH 4 C o n c e n tra tio n .......................................... 244
6.6.3.4 Effects of CO2 C o n c e n tra tio n .......................................... 250
6.6.3.5 Effects of Total Flow R a t e ................................................250
6.6.4 Typical R am an S p e c t r u m ...........................................................256
6.6.5 Typical Uniformity P r o f i l e ...........................................................256
6 .6.6 S u m m a r y ...................................................................................... 256
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C hapter Seven
Conclusions and Recomm endations
7.1
Sum m aiy of Significant R e s u l t s ......................................................261
7.1.1 I n t r o d u c t i o n ................................................................................. 261
7.1.2 MCPR D e v e lo p m e n t...................................................................... 262
7.1.3 Microwave Plasm a J e t Reactor D e v e lo p m e n t......................... 264
7.1.3.1 Microwave Cavity J e t R e a c t o r s .............................264
7.1.3.2 J e t Configuration in M C P R .................................. 264
7.1.4 Reactor O perational C h a r a c t e r i s t i c s .....................................265
7.1.5 Diamond Film Deposition on 3" Silicon W afers . . . .
266
7.1.6 Diamond Film Deposition on 4 ” Silicon W afers . . . .
267
7.1.6.1 Study of Experim ental S e t - u p s .............................267
7.1.6.2 Diamond Film Deposition on 4 ” Silicon Wafers . . 268
7.1.7 Reactor Performance “Figures of M e r i t " ............................... 269
7.1.7.1 Performance “Figures of Merit” of MCPR . . . .
269
7.1.7.2 Com parison w ith other R e a c to rs ............................. 271
7.1.8 M easurem ent of Electric Fields in M C P R ............................... 272
7.2 Recomm endations for F uture R e s e a r c h .....................................273
LIST OF REFERENCES
............................................................................275
xii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table 2.1
Sum m ary of “Figures of Merit" of
Diamond R e a c t o r s ............................................................. 41
Table 3.1
R esonant Caviiy Lengths of TM
Modes in 14” C a v ity ........................................................... 104
Table 5.1
Table 5.2
Diagnostic Probe Power Reading
versus In p u t P o w e r ............................................................130
Absorbed Power an d Probe Power Ratio
. . . .
130
Table
Table
Table
Table
Table
Table
Table
Table
Table
S ubstrate Tem perature Conditions ( I ) ............ 175
S ubstrate Tem perature Conditions ( I I ) ............ 178
S u b strate Tem perature Conditions (III)............ 180
Com parison of Two Cavity Shells
................................. 224
Com parison of Two Q uartz D o m e s ................................. 226
Effects of S u b strate Location ( I ) .......................................226
Effects of S u b strate Location ( I I ) .......................................228
Effects of Cavity Mode Excitation
.................................229
S u b strate Tem perature Conditions (4”)
. . . .
234
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
Table 7.1
Sum m ary of “Figures of Merit” of Diamond
Film Deposition Experiments in M C P R ..........................270
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Figure 1.1
Diamond Properties and Its
Engineering Applications [ 3 ] ............................
2
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
2.11
2.12
2.13
A Generic Diamond Film Deposition Reactor . . .
Thermally Activated CVD Reactor [18,19]. . . .
DC Discharge Reactor [ 2 3 . 2 4 ] ...........................18
DC Discharge J e t Reactor [ 2 5 ] ...........................21
Com bustion Flame Reactor [ 2 6 ] ...........................23
RF Glow Discharge Reactor [ 2 8 ] ...........................25
RF Induction Thermal Plasm a Reactor [29]
. . .
T ubular Microwave Reactor [ 3 0 ] .......................... 30
Magneto-microwave Reactor [31]
Magneto-active Microwave Reactor [32] . . . .
Bell-jar Microwave Reactor [ 3 4 , 3 5 ] ..................... 36
Microwave J e t Reactor [ 3 7 , 3 8 ] ...........................38
Bachm ann C-H-O Phase Diagram [ 3 6 ] ............... 44
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
Microwave Power Supply Circuit
.................................. 48
Flow Control and V acuum System for MCPRs
. . 50
Com puter Monitor System Block Diagram
. . .
54
Monitor Program Flow C h a r t ................................ 56
Schem atic Drawing of M C P R 7 -1 ...........................58
Mode Diagram of an Ideal 7" Cavity
.............................61
Field P attern of Discharge Loaded TM011 Mode . . 62
Ram an Spectrum of a Diamond Film
Deposited in M C P R 7 -1 .......................................... 66
Schematic Drawing of M C P R 7 -2 ........................... 68
Field P atterns of Discharge Loaded
TMqi2 an d TMoi3 M o d e s ..................................... 70
Ram an Spectrum of a Diamond Film
Deposited in M C P R 7 -2 ...........................................74
Figure 3.9
Figure 3.10
Figure 3.11
xiv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
16
27
32
34
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
3.12
3.1 3
3 .1 4
3.1 5
3 .1 6
3.1 7
3.1 8
3.1 9
3 .2 0
3.21
3.2 2
3.2 3
Figure 3 .2 4
Figure 3 .2 5
Figure
Figure
Figure
Figure
4.1
4 .2
4 .3
4 .4
Figure 4 .5
Figure 4 .6
Figure
Figure
Figure
Figure
Figure
Figure
5.1
5.2
5.3
5 .4
5 .5
5 .6
Figure 5 .7
Figure 5.8
Figure 5.9
Field P a tte rn of TE qi i M o d e ........................................76
Schem atic Drawing of a Loop Coupling System .
. 77
The B asic E xperim ental C o n fig u ra tio n ...................... 81
The B ase-plate of M C P R 7 - 3 ....................................... 83
The Forced Flow E xperim ental C onfiguration .
. . 85
Flow P atte rn R e g u l a t o r s ............................................. 86
J e t Experim ental C o n f ig u r a tio n .................................. 88
Lower Cavity R esonance C o n fig u ra tio n ...................... 89
D ow n-stream E xperim ental Configuration
. . .
91
Prelim inary J e t Experim ental C onfiguration . . .
94
Schem atic Drawing of a 3" S u b stra te H eater . . .
96
S u b strate S upported by a
Q uartz T ube in 14" Cavity
............................................ 100
S u b strate Supported by a
Metal T ube in 14" C a v i t y ...........................................101
R esonant Cavity Lengths of
TM Modes in 14" C a v i t y ...........................................103
Flow Control an d V acuum System for M CJRs
. . 109
Schem atic Drawing of M CJR-1 .................................I l l
Schem atic Drawing of M C J R - 2 .................................117
R am an S pectrum of a Film
Deposited in M C J R - 2 ................................................ 120
Schem atic Drawing of M C J R - 3 ................................. 122
Enlarged Drawing of th e Nozzle I n s e r t ..................... 123
Electric Field M easurem ent S y s t e m ........................... 128
Horizontal C ross-sectional V i e w ................................. 129
M easured Electric Field S tren g th D istribution
. 134
M easured Field P atte rn in M C P R ........................... 136
Discharge B oundaries an d Electric Fields
.. 1 3 8
Coaxial Probe Power v ersu s
Microwave Power an d Flow R a t e ............................... 142
Coaxial Probe Power v ersu s
Microwave Power an d P r e s s u r e ............................... 143
S u b strate T em perature vs.
Microwave Power an d Flow R a t e ............................... 145
S u b strate T em perature vs.
Microwave Power an d P r e s s u r e ................................146
xv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure
Figure
Figure
Figure
6.1
6.2
6.3
6.4
Figure 6.5
Figure 6.6
Figure 6.7
Figure 6.8
Figure 6.9
Figure 6.10
Figure 6 .11
Figure 6 .12
Figure 6.13
Figure 6.14
Figure 6.15
Figure 6.16
Figure 6.17
Figure 6.18
Reactor I/O Block D i a g r a m ............................................153
Schematic Drawing of Cavity Shell ( I ) ............................160
Schematic Drawing of Cavity Shell ( I I ) ........................... 161
S ubstrate T em perature versus Pressure
and Absorbed Microwave P o w e r ..................................... 167
S ubstrate T em perature versus
CH4 C o n c e n tr a tio n ........................................................... 168
Substrate T em perature versus
CO2 C o n c e n tr a tio n ........................................................... 170
Substrate T em perature versus Total Flow Rate
and Absorbed Microwave P o w e r ..................................... 171
Linear Growth Rate an d Weight Gain versus
S ubstrate T em perature and Power Flux (I). . .
175
(a) Location of S ubstrate Tem perature Conditions
on Energy Map an d (b) Carbon Conversion
Efficiency and G as Flow Efficiency v ersu s
Substrate T em perature ( I ) ................................................ 176
Linear Growth Rate and Weight Gain versus
Substrate Tem perature and Power Flux (II) . . . 1 7 8
(a) Location of S ubstrate Tem perature Conditions
on Energy Map an d (b) Carbon Conversion
Efficiency and G as Flow Efficiency versus
...........................................179
Substrate Tem perature (II)
Linear Growth Rate an d Weight Gain versus
Substrate Tem perature and Power Flux (III) . . . 1 8 0
(a) Location of S ubstrate Tem perature
Conditions on Energy Map and (b) Carbon
Conversion Efficiency an d Gas Flow Efficiency
versus S ubstrate Tem perature ( I I I ) ................................181
Location of Gas Compositions on th e C-H-0
Phase Diagram ( I ) ........................................................... 182
Linear Growth Rate, Weight Gain an d Carbon
Conversion Efficiency versus CH4
Concentration ( I ) ................................................................. 184
Linear Growth Rate, Weight Gain And Carbon
Conversion Efficiency versus CH4
Concentration (II)
........................................................... 185
Location of Gas Compositions on th e C-H-0
Phase Diagram ( I I ) ........................................................... 186
Linear Growth Rate, Weight Gain an d Carbon
xvi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.19
Figure 6.20
Figure 6.21
Figure 6.22
Figure 6.23
Figure 6.24
Figure 6.25
Figure 6.26
Figure 6.27
Figure 6.28
Figure 6.29
Figure 6.30
Figure 6.31
Figure 6.32
Figure 6.33
Figure 6.34
Figure 6.35
Figure 6.36
Conversion Efficiency versus CO2
Concentration ( I ) .................................................................188
Linear Growth Rate, Weight Gain a n d Carbon
Conversion Efficiency versus CO2
Concentration (II)
........................................................... 189
Location of Gas Compositions on th e C-H-.O
Phase Diagram ( I I I ) ........................................................... 191
(a) Linear Growth Rate, Weight Gain and
(b) Carbon Conversion Efficiency v ersu s Total
Flow Rate and Gas Residence Time ( I ) ..........................192
(a) Linear Growth Rate, Weight Gain and
(b) Carbon Conversion Efficiency v ersu s Total
Flow Rate and Gas Residence Time (II)
. . . .
193
Effects of S ubstrate Temperature on
Film Morphology ( I ) ........................................................... 195
Effects of S u b strate Temperature on
Film Morphology (II)
......................................................197
Effects of CH4 Concentration on
Film Morphology ( I ) ...........................................................201
Effects of CH4 Concentration on
Film Morphology (II)
..................................................... 204
Effects of CO2 Concentration on
Film Morphology ( I ) ...........................................................208
Effects of CO 2 Concentration on
Film Morphology (II)
..................................................... 210
Effects of Total Flow Rate on
Film Morphology ( I ) ...........................................................212
Effects of Total Flow Rate on
..................................................... 216
Film Morphology (II)
Ram an Spectrum of a 3" Diamond Film
Deposited in M C P R 7 -3 ..................................................... 219
Cross-sectional Thickness Profile of
a 3" Diamond F i l m ...........................................................220
Three Dimensional Thickness Profile of
a 3" Diamond F i l m ...........................................................221
Effects of Q uartz Dome Geometry on
Film M o rp h o lo g y ................................................................ 227
Linear Growth Rate and Weight Gain versus
S ubstrate Tem perature and Power Flux . . . .
234
Carbon Conversion Efficiency and G as Flow
xvii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.37
Figure 6.38
Figure 6.39
Figure 6.40
Figure 6.41
Figure 6.42
Figure 6.43
Figure 6.44
Figure 6.45
Figure 6.46
Figure 6.47
Figure 6.48
Efficiency versus S ubstrate Tem perature
and Power F l u x ................................................................ 236
Linear Growth Rate, Weight Gain and
Carbon Conversion Efficiency versus
CH4 c o n c e n t r a t i o n ...........................................................237
Location of Gas Compositions on the
C-H-0 Phase Diagram (IV)
.......................................... 239
Linear Growth Rate, Weight Gain And
Carbon Conversion Efficiency versus
CO2 c o n c e n t r a t i o n ...........................................................240
Location of Gas Compositions on the
C-H-O Phase Diagram ( V ) ................................................242
(a) Linear Growth Rate, Weight Gain and
(b) Carbon Conversion Efficiency v ersus Total
Flow Rate and Gas Residence T i m e ............................... 243
Effects of S ubstrate Temperature on
Film M o rp h o lo g y ................................................................ 245
Effects of CH 4 Concentration on
Film M o rp h o lo g y ................................................................ 247
Effects of CO 2 Concentration on
Film M o rp h o lo g y ................................................................ 251
Effects of Total Flow Rate on
Film M o rp h o lo g y ................................................................ 254
Raman Spectrum of a 4" Diamond Film
Deposited in M C P R 7 -3 ..................................................... 257
Cross-sectional Thickness Profile of
a 4" Diamond Film .
258
Three Dimensional Thickness Profile of
a 4" Diamond F i l m ...........................................................259
xviii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER ONE
INTRODUCTION
1.1
Introduction
The beauty, h ard n ess an d rarity of diam ond have m ade it an object
of fascination for nearly a s long as recorded history. The mechanical,
electrical, optical, chemical, an d therm al properties of diamond make it
attractive in applications ranging from w ear-resistant coatings for
m echanical and optical com ponents to su b strates for advanced
sem iconductor devices. Ever since T ennant discovered th a t diamond is
simply made of carbon in 1797<1>, synthesis of diam ond h a s long been a
goal of num erous research efforts of both individuals an d organizations.
The diamond-cubic lattice consists of two interpenetrating facecentered cubic lattices, displaced by one q u arter of th e cube diagonal.
Each carbon atom is tetrahedrally coordinated, m aking strong, directed a
bonds to its neighbors using hybrid sp 3 atomic orbitals. The lattice can
also be visualized as planes of six-membered satu rated carbon stacked in
an ABCABC... sequence along <111> directions.<2>
Figure 1.1 sum m arizes the diamond properties and its engineering
applications.<3>
Diamond is m ost noted for its mechanical h ard n ess which gives it
great prom ise for a wide range of m echanical coating applications where
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DIAMOND
PRO PERTIES
ENGINEERING
APPLICATIONS
Abrasive Coatings
for C utting Tools
High Thermal Conductivity
H eat Sinks for
Electronic Devices
Electric Insulator
H eat Resistive
Microwave Power
Devices
Large Band Gap
Low Dielectric C o n stan t
High Hole Mobility
Acid R esistant
Radiation R esistant
T ran sp aren t
Figure 1 1
RF Electronic
Devices
High Speed
Electronic Devices
Electronic Devices for
Severe Enviroments
.^ j s u c h a s in Space or
in N uclear Reactors
Electro-optical
Diam ond Properties and Its Engineering Applications [3]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
resistance to abrasive phenom enon is desired. Diamond coated cutting
tools will provide significant productivity improvements for m achining
su per-h ard materials, nonferrous refractory metals and superalloys.
At room tem perature and above, the therm al conductivity of
diam ond is greater th a n th a t of any other known m aterial, including
copper and silver. Thermal tran sp o rt in diamond is accomplished by
phonon m echanism s, w hereas in a metal, therm al conduction is by
electron transport. Diamond is unique in th a t no other known m aterial is
sim ultaneously a superb therm al conductor, optically tran sp a re n t and
electrically insulating. With th is com bination of properties, diamond will
be a excellent m aterial for m any engineering systems. Diamond’s
relatively low therm al expansion coefficient m akes it ideal for
microelectronic circuit boards and sem iconductor su b strates since large
therm al expansion su b strates create micro-cracks in the device circuitry
resulting in open circuits in high-power, high-tem perature electronic
applications.
Diamond offers significant promises for high power and high
tem perature semiconductor device applications. The dielectric
breakdow n electric field is lx l 0 7volts/m , approximately 50 tim es th a t of
conventional semiconductors su ch as GaAs. The theoretical power
handling capability of diamond is a factor of 2500 greater th a n silicon.
The bandgap of diamond is 5.45 eV which is more th a n three tim es th at
of conventional semiconductor su ch a s Si and GaAs. Diamond’s therm al
conductivity is approximately five times th a t of copper an d nearly 20
tim es th a t of silicon. Diamond's dielectric constant is ab o u t h alf th a t of
GaAs. Diamond semiconductors dem onstrate approximately h alf the
capacitive loading of other conventional sem iconductors a t any given
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
frequency. Diamond’s satu rated carrier velocity is 2.7 times th a t of GaAs.
The satu rated carrier velocity of diamond is greater th a n the peak
velocity of GaAs.
U ltra-pure diam ond probably has th e w idest electromagnetic b an d ­
p ass of any known solid material. Diamond is composed of a
homogeneous, tetrahedral network of covalently bonded carbon.
Diamond’s cubic crystal stru ctu re m akes diam ond optically isotropic.
The fact th a t diamond is composed of a single elem ent m eans th a t pure
diamond h a s no fundam ental infrared absorption modes. Diamond
windows can be used in conjunction w ith optical detection devices for
detection of UV, visible an d IR radiation.<3>
In 1955, Bundy and co-workers succeeded in th e reproducible
synthesis of diamond<4> with a molten transition m etal solvent-catalyst
a t pressures and tem peratures where diam ond is th e thermodynamically
stable phase of carbon.
Diamond growth a t low pressures w here graphite is the stable
carbon phase can be traced back to W. G. Eversole<5>, Angus et al.<6>
and Deiyaguin et al.<7>, b u t the low growth rate (less th an 0.1
micrometer per hour) w as not of practical commercial importance, and
th u s a t th a t time it prevented worldwide interest in low pressure
diamond growth. The breakthrough in the synthesis of diamond a t low
pressures cam e in th e late 1970’s and early I9 8 0 ’s, w h en a group of
Soviet researchers<8> and Japanese researchers<9> started publishing a
series of research papers on diamond film growth a t higher growth rates
(several micrometers p er hour) from hydrocarbon-hydrogen gas m ixtures.
Since then, great interest and various techniques have been developed
for diamond film growth a t low pressures.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
1.2
Research Objectives
The objectives of th e research covered in th is dissertation were to:
(1) experimentally investigate the deposition of diam ond films over three
and four in ch diam eter substrates using a microwave discharge, (2)
investigate and develop microwave diam ond film deposition reactors to
achieve large surface area diamond film deposition w ith high growth
rates (> 0.5 pm /hr), (3) investigate the film quality, deposition rate and
uniformity versus various experimental param eters using the prototype
reactor developed, and (4) develop a better understanding of the
electromagnetic field /p lasm a interactions during diamond film
deposition process.
1.3
R esearch History
The research described in th is dissertation w as carried out by the
author over th e period of 1987 to 1992 a t Michigan State University
(MSU) u n d er the direction of Professor J e s A sm ussen. The research work
described in th is dissertation took place prim arily in the Microwave
Plasm a & Microwave Material Processing Laboratories a t Michigan State
University. The funding for this research w as provided by Norton
Company, Wavemat Inc., and Michigan State Research Excellent Fund.
This research builds on previous microwave plasm a reactor
research carried out by Dr. Rogers,<10> Dr. W hitehair<11>, Dr.
Dahimene<12> and Dr. Hopwood<13>. T his p a s t reactor research h as lead
to MSU owned and patented microwave plasm a technologies th a t have
many microwave plasm a and microwave m aterial processing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
applications. The success of th is dissertation depended upon these
earlier experimental an d theoretical studies of microwave discharges
inside microwave cavities. The theoretical an d experim ental knowledge
used to design, build an d understand the diam ond film deposition
reactors presented in th is dissertation relied on knowledge obtained in
these earlier experiments.
The research described in this dissertation reflects co-operation
with Norton Company an d Wavemat, Inc. The first microwave cavity
plasm a reactor (MCPR) developed for diamond film deposition w as
designed and b u ilt for Norton Co. a t MSU by J . A sm ussen in 1986. It was
experimentally evaluated a t Norton Co. in Nov./Dec. of 1987 and
immediately diamond films were successfully deposited. This MSU owned
and patented technology w as exclusively licenced to Wavemat, In c./
Norton Co., and Wavemat, Inc. then developed a commercial version of
this reactor. One of these reactors was p u t into operation a t MSU in
1988. Continued co-operation w ith Norton Com pany and Wavemat, Inc.
h a s contributed to the success of the research described in this
dissertation.
The research conducted for this dissertation experimentally
investigated and developed two different groups of advanced reactors th a t
produce microwave plasm a discharges for diam ond film growth. The first
group of reactors are th e microwave cavity plasm a reactors. The second
group of reactors are th e microwave plasm a je t reactors. The research
described here is prim arily experimental. Since th ere is no theoretical
model for the plasm a assisted deposition of diam ond films, the
improvement of reactors w as carried out through m an y experiments. The
final experimental designs were achieved through num erous
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
experimental iterations and by the investigator’s b est intuition of the
deposition process. The results summ arized in th is dissertation are the
results of at least five h u n d red separate experiments and of thousands of
hours of actual experimental time.
This dissertation sum m arizes the experimental research and
development th a t h as lead to th e successful creation of a working
prototype of a n improved microwave cavity plasm a reactor concept. This
type of reactor is already being used to deposit diamond films over larger
surface areas a t Norton Co.. It h as made an im pact on th e commercial
production of low pressure diamond coated commercial products by
increasing the productivity by more th a n three tim es over earlier
microwave diam ond film deposition reactors. The reactor developments
described in this dissertation were continuously monitored by Wavemat,
Inc. an d Norton Co. and as new reactor concepts were developed and
tested a t MSU, the successful concepts were alm ost immediately
incorporated into commercial diamond film deposition reactor designs.
Some of the research described in this dissertation h a s already been
published in scientific publications and international scientific
conferences’^14' 16> and in p aten t disclosures<17>. However, m uch of the
results on diamond film deposition on three and four inch silicon wafers
using the advanced reactor concept h a s not been subm itted for
publication. It is expected th a t in the n ear future, these resu lts will also
be published.
Diamond films were deposited u n d er various experimental
conditions and characterized w ith respect to the film quality, growth rate
and uniformity. Diamond Film quality w as characterized by Scanning
Electron Microscopy and R am an Spectroscopy. Diamond film growth rate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
was characterized by weight gain and Scanning Electron Microscopy.
Diamond film uniformity was characterized by Laser Interference/
Reflection M easurement. Electromagnetic field p attern m easurem ents
were conducted to develop better understanding of the electromagnetic
field/plasm a interactions.
1.4
D issertation Outline
This dissertation is organized as follows. C hapter two presents a
generic low p ressure diamond deposition reactor concept and briefly
reviews th e state-of-the-art of other diam ond film deposition reactors.
The perform ance of these reactors is com pared w ith respect to the
diam ond film quality, growth rate, growth efficiency, surface area, and
uniform ity whenever possible. C hapter three describes a series of
microwave cavity plasm a reactors th a t were developed and tested as p a rt
of th is dissertation research. C hapter four sum m aries the experimental
development of three generations of microwave plasm a je t reactors for
diam ond film growth a t high growth rates. C hapter five presents an
investigation of the electromagnetic filed p attern s in th e microwave cavity
p lasm a reactor while operating u n d er diam ond film deposition
conditions. The purpose of this stu d y w as to develop a better
u nderstanding of the electromagnetic field /p la sm a interactions during
film deposition process. Chapter six p resents a new method th a t
experim entally characterizes the operational characteristics of prototype
reactors. Param etric studies of diamond film deposition on three and four
inch silicon wafers conducted in the new prototype reactor were also
described. Chapter seven presents the conclusions and some
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
speculations on future research and development of diamond film
deposition using microwave technology.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER TWO
DIAMOND FILM DEPOSITION REACTORS: A REVIEW
2.1
Introduction
2.1.1
A Generic Diamond Film Deposition Reactor
A generic reactor for low pressure diamond film growth by
chemical vapor deposition can be illustrated by the diagram displayed in
Figure 2.1. Here, the reactive gas in p u t is usually a m ixture of hydrogen
(H2), hydrocarbon (CH4, C2H2, etc.) and oxygen or oxides (O2. CO, CO2,
etc.), etc. The energy in p u t is provided by electrical (DC, low a n d /o r high
frequency ac) energy a n d /o r chemical reaction energy. With sufficient
energy input, the input reactive gas m ixture is dissociated, creating a
m ixture of dissociated charged species, electrons and neutral gases. A
su b strate is placed close to th e dissociated reactive gas species an d its
tem perature a n d /o r electrical bias m ay be controlled by external
tem perature controller a n d /o r dc or RF power supply. With proper choice
of in p u t reactive gas mixture, energy input, substrate, su b strate
tem perature a n d /o r bias, etc., a diamond film is deposited on th e
substrate.
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
Mixture of dissociated
charged species,
electrons and neutral
gases.
Reactive gas
input
Energyinput
Substrate
r
i
Substrate temperature
and/or bias controller
Figure 2.1 A Generic Diamond Film Deposition Reactor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
Reactor Categorization
2. 1.2
The reactors m ost frequently used in low pressure diamond film
deposition can be divided into five m ajor categories:
(1)
Thermally activated (filament) chemical vapor deposition
(CVD);
(2)
Direct current (DC) discharge assisted CVD;
(3)
Combustion flame;
(4)
High frequency (RF or microwave) discharge assisted CVD;
(5)
Other and hybrid reactors.
2.1.3
“Figures of Merit” of Reactor Performance
In order to quantitatively compare the performance of the reactors,
it is necessary to define a set of performance “figures of merit”. In this
literature review, th e following performance “figures of merit” are defined:
(1)
Linear growth rate (^m/hour);
It is defined as th e diamond film thickness (jim) gain per u nit
tim e (hour).
(2)
Deposition area
(cm2);
It is defined as the area over which diamond film is
deposited.
(3)
Weight gain u (mg/hour);
It is defined as th e diamond film weight gain (mg) per u n it
time (hour). Since m ost literature ju s t report linear growth rate, weight
gain is usually calculated from the reported linear growth rate,
deposition area and the density of diamond. 3.51 g /cm 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(4)
Uniformity;
It is defined as th e film thickness variation divided by the
average film thickness over the deposition area.
(5)
Growth Efficiency
(i)
Weight gain/pow er in p u t (mg/kW-hr);
It is defined as the diam ond film weight gain (mg) per
energy in p u t (kW-hr).
(ii)
Weight g ain /to tal flow in p u t (mg/liter);
It is defined as th e diam ond film weight gain (mg) per
total gas in p u t (liters). The total flow in p u t u sed here is the su m of all
input gas flow rates.
(iii)
Carbon Conversion Efficiency (%):
It is defined as the percentage of carbon atom s in the
input gases th a t are converted into the diamond film.
2.1.4
Notes on Literature Review
The review described in the following sections is based on
experimental d ata th a t is published in th e open literature. Since
diamond film deposition h as been used in industrial applications, it is
possible th a t m ost of the state-of-art deposition information is n ot
available in th e open literature.
The performance “figures of m erit” described in the following
sections are mostly estimated a n d /o r calculated from the b est d ata
available in th e literature. There are limiting factors on the accuracies of
the perform ance "figures of m erit”. This is due to th a t limited d a ta is
available in th e literature concerning growth rate, deposition a rea and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
uniformity. It is also often n o t clear how these d a ta are obtained.
Film linear growth rate, for example, is m easured differently by
different research groups. Some groups u se th e crystal growth rate (1) as
the linear growth rate, i.e., th e size of an individual crystal divided by
growth time. Some groups determ ine the film thickness from SEM
photographs (2) of the cross-sectional area of a film. The linear growth
rate is th en obtained from dividing the film thickness by the deposition
time. Some groups use weight gain (3) of a su b stra te after a deposition
process along with the deposition area and the density of diamond,
3.51m g/cm 3, to determ ine th e film thickness. The linear growth rate is
then obtained from dividing the film thickness by th e deposition time.
The author believes th a t the first m ethod, w hich is frequently used
in earlier literature, is n o t a n accurate m easure of linear growth rate of a
film since it is often possible to grow large individual crystals without
forming a complete continuous film. The second m ethod can be an
accurate m ethod if the SEM instrum ents are calibrated and th e specimen
are photographed directly onto th e cross-section. The third method is a
simple method and can be a n accurate method of evaluating the film
thickness provided the deposited film is uniform diam ond film. The
m ethods of linear growth rate m easurem ent are indicated whenever
available.
Finally in this review, diam ond film deposition area is often
estim ated since film uniform ity d a ta are mostly n o t available in the
reviewed literature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
2.2
A Review of Diamond Film Deposition Reactors
2.2.1
Thermally Activated CVD Reactor
A typical therm ally activated (hot filament) diam ond film deposition
reactor is shown in Figure 2.2.<18,19> As shown, in th is reactor, th e feed
gas, w hich is a m ixture of hydrocarbon and hydrogen, dissociates when
it is heated by a hot filam ent. The hot filament is heated by a n electric
current. The substrate is placed n ear th e hot filam ent an d diamond film
is formed on the su b strate w hen the h o t dissociated gas species react on
its surface.
Variations of th is reactor have also been u sed for diamond film
growth. For example, su b stra te bias a n d /o r su b strate h eater have been
utilized to enhance th e growth rate and uniformity of th e deposited
diam ond film<20>. Multiple filam ents have been u sed to scale up the
deposition to 4 inch su b strate.<21>
Diamond film deposition experiments have been conducted under
th e following typical experim ental conditions <22>: su b strate = Si, H2 flow
rate = 100 seem, C2H 2 flow rate = 0.5 seem, CO flow rate = 12 seem, gas
pressu re = 50 Torr, total flow rate =100 seem, tem perature of deposition
cham ber = 890 °C, filam ent tem perature = 2400 °C, filam ent voltage = 30
V, filam ent current = 28 A, h eater voltage = 70 V, h ea ter current = 6.5 A
The “figures of m erit” available in the literature are:<22>
(1)
Linear growth rate (pm /hour) = 0.6;
(2)
Deposition a rea (cm2) ~ 6.4;
(3)
Weight gain (m g/hour) ~ 1.35;
(3)
Uniformity = n /a ;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Gas input
Filament
Substrate
Silica holder
Silica
tube •
Thermocouple
To pump
Figure 2.2 Thermally Activated CVD Reactor [18,19]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
(4)
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) ~ 1.04;
(ii)
Growth ra te /g a s in p u t (mg/liter) - 0.2;
(iii)
Carbon conversion efficiency (%) -0.32.
The therm ally activated diamond film deposition reactor is easy to
construct, operate, and scale up. It faces th e problem of filament erosion
and breakage during the diamond film deposition process.
2.2.2
Direct C urrent Discharge CVD Reactor
2.2.2.1.
Direct C urrent Discharge Reactor
A schem atic drawing of a DC discharge diam ond film deposition
reactor is shown in Figure 2.3. <23,24> As shown, in this reactor, the
input gas which is a m ixture of hydrocarbon and hydrogen is dissociated
by a DC voltage across the anode and cathode. A DC power supply is
used to su stain the DC discharge. The su b strate is m ounted on the
anode and both the anode and cathode can be w ater cooled. Diamond
film is formed on the su b strate w hen the dissociated gas species react on
its surface.
Topical experiments have been conducted in th is reactor u n d er the
following experimental conditions: su b strate = Si an d AI2O3, gas m ixture
= H 2 and CH4, CH 4 / H 2 = 0.3 - 4%, p ressu re = over 200 Torr, total flow
rate = 20 seem, anode-cathode separation = 2 cm, su b strate tem perature
= 800 - 850 °C, discharge voltage = 1 kV, discharge current = 4A /cm 2,
su b strate area = 0.25 cm2.
Notes on reactor operation: The initial stages of diamond particle
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Water
I
Anode
Substrate
Power
Supply
DC
discharge
Water
I i
Cathode
Gas input
Figure 2.3
DC Discharge Reactor [23,24]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
growth a t a substrate tem perature of 800 °C and CH 4 concentration of
2% indicate th a t the initial nucleus density of diamond on th e m irror
polished Si (111) surface is about 108/c m 2. This nucleation density is as
large a s th a t on the surface scratched by diamond polishing powder (0.5
pm in diameter). After 30-m in of deposition, a continuous film can be
obtained in both cases. The growth rate of the films is about 20 pm /hr.
Diamond th in films are also formed on AI2O3 su b strate w ithout surface
treatm en t.<23,24>
Diamond thin films can be obtained a t a substrate tem perature
higher th a n 600 °C. At a substrate tem perature below 600 °C,
am orphous carbon films were obtained. If the substrate is m ounted on
the cathode, the growth of diamond can n o t be detected an d am orphous
or graphitic carbon is found over a wide range of experimental
conditions.
The “figures of m erit” available are:
(1)
Linear growth rate (pm /hour) - 20;
(2)
Deposition area (cm2) ~ 0.25;
(3)
Weight gain (mg/hour) ~ 1.76;
(3)
Uniformity = n /a ;
(4)
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) - 0.56;
(ii)
Growth ra te /g a s in p u t (mg/liter) ~ 1.47;
(iii)
Carbon conversion efficiency (%) - 13.7.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
2.2.2.2.
Direct C u rren t Discharge J e t Reactor
The schem atic drawing of a DC discharge je t diamond film reactor
is shown in Figure 2 .4 .<25> As shown, in th is reactor, the input gas
which is a m ixture of Ar, H2 and CH4 is dissociated b y a DC voltage
across the electrodes. A high tem perature discharge je t is created and
sustained by a DC power supply. The su b strate is m ounted down stream
of the je t stream on a w ater cooled su b strate stage. Diamond film is
formed on the su b stra te when the dissociated gas species react on its
surface. The bias voltage
is used to enhance th e film growth rate.
The typical experim ental conditions are: Ar flow rate = 30 1/min,
H 2 flow rate = 10 1/min, CH4 flow rate = 1 1/min, pressure = 140 Torr,
discharge voltage = 70 - 76 V, discharge cu rren t = 133 - 150 A, bias
voltage = 0 - 500 V, b ias current = 0 - 5 A, su b stra te = Mo plate of 20 mm
in diameter, distance between the su b strate an d nozzle = 57 - 102 mm,
su bstrate tem perature = 700 - 1100 °C, deposition time = 10 min,
su bstrate pretreatm ent = scratched w ith 5 - 1 0 pm particle size diamond
paste for about 5 min.
Notes on reactor operation: The deposition rate w as increased by
more th a n twofold w hen positive bias w as applied to th e substrate. The
deposition area was also increased b u t the uniform ity of film thickness
did not improve. The m axim um growth rate obtained w as 900 pm /hour.
“Figures of merit" th a t are available are:
(1)
Linear growth rate (pm/hour) ~ 900;
(2)
Deposition area (cm2) - 2;
(3)
Weight gain (mg/hour) - 632;
(3)
Uniformity = n /a ;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
Ar+H2
CH/
V,b —
J et stream
R
I ) !
Substrate
Stage
I
Cooling
water
Figure 2.4 DC Discharge J e t Reactor [251
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
(4)
2.2.3
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) ~ 57;
(ii)
Growth ra te /g a s in p u t (mg/liter) - 0.26;
(iii)
Carbon conversion efficiency (%) ~ 1.97.
Com bustion Flame Reactor
The schematic drawing of a typical com bustion flame diamond film
deposition reactor is shown in Figure 2.5.<26> As shown, in this reactor,
the feed gas which is a m ixture of acetylene (C2H2) and oxygen
dissociates by the chemical reaction energy of th e gas mixture after a
flame is ignited. The com bustion flame is self-sustained by the energy
released by the chemical reaction of the gas m ixture a t high tem perature.
The substrate is placed in th e C2H2 feather region (see insert in Figure
2.5) an d diamond film is formed on the su b strate w hen the dissociated
gas species react on its surface.
Typical experim ental conditions are as follows<27>: substrate
location = in the C2H 2 feather region of the flame, substrate tem perature
= 650 - 1050 °C, gas flow ratio R = O2/C 2H 2 = 0.9 - 1.2, total flow rate =
2 slm (nozzle dependent). U nder these experim ental conditions, the
feather area is - 6 m m in diameter, th e growth rate is ~ 60 tun/hour.
The “figures of m erit” th a t are available are:<27>
(1)
Linear growth rate (pm/hour) ~ 60;
(2)
Deposition area (cm2) ~ 0.28;
(3)
Weight gain (m g/hour) ~ 5.9;
(3)
Uniformity = N/A;
(4)
Growth Efficiencies
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Inner cone
C2H2 feather
Outer flame
►
CoH,
C2H2
feather
Outer
flame
Substrate
Water
Substrate holder
Figure 2.5
Combustion Flame Reactor [26]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
(i)
Growth rate/pow er input (mg/kW-hr)= N/A;
(ii)
Growth ra te /g a s input (mg/liter) ~ 0.049;
(iii)
Carbon conversion efficiency (%) ~ 0.009
2.2.4
Radio Frequency (RF) Plasm a CVD Reactors
2.2.4.1
RF Glow Discharge Reactor
Figure 2.6 displays th e schematic drawing a RF glow discharge
diam ond film deposition reactor<28>. As shown, in th is reactor, th e input
gas, which is a m ixture of hydrocarbon and hydrogen, is dissociated by a
radio frequency (RF) energy m atched into reaction cham ber through an
im pedance m atching network. A 13.56 MHz RF power supply is used to
su sta in the RF discharge. The substrate is placed in th e middle of the
coil. Diamond film is formed on the substrate w hen the dissociated gas
species react on its surface.
Typical experimental conditions are as follows: su b strate = Si or
Mo or Si02 , in p u t gas m ixture CH4/H 2 = 0.2 - 1%, total flow rate ~ 50
seem, pressure = 3.8 - 22.8 Torr, RF power = 0.5 ~ 1 kW, substrate
tem perature - 950 °C, q uartz tube geometry = 3 cm in diam eter and 1 m
long, coil geometry = 4 cm in diam eter and 13 cm long, coil m aterial = 13
tu rn s of 6.4 mm diam eter copper tube,
Notes on the reactor operation: The su b stra te tem perature could
not be determined accurately a t low tem peratures d u e to radiation from
the discharge. Since there is no external su b strate heating, the
tem peratures of the su b strates were raised by inductive heating and
energy transfer from th e discharge. It is necessary to u se high RF power
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Working coil
To Manometer
OOOOOOO
Feed gas
Substrate
To Pump
0000
Quartz tube
Water
RF generator
13.56 MHz
Figure 2.6 RF Glow Discharge Reactor [28]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
to create the discharge an d the tu b e wall m ay be sputtered a t low
pressures in th is reactor.
The “figures of m erit” are n ot obtainable from the literature.
2.2.4.2
RF Induction Thermal Plasm a Reactor
The schem atic drawing of a RF induction therm al plasm a diamond
film reactor is displayed in Figure 2.7.<29> As shown, in this reactor, the
input gas which is a m ixture of Ar, H2 and CH4 is dissociated by a radio
frequency (RF) energy m atched into reaction cham ber through an
impedance m atching network. A 4 MHz 60 kW RF power supply is used
to su stain the RF therm al plasm a. The substrate is placed w ithin or at
the end of the therm al plasm a on a w ater cooled holder. Diamond film is
formed on the substrate w hen th e dissociated gas species react on its
surface.
Typical experimental conditions are a s follows: cham ber inside
diam eter - 45 mm, su b strate = 20 mm diam eter Mo plate, sh eath (S) gas
flow rate = 35 1/m Ar + 12 1/m H2, plasm a (P) gas flow rate = 17 1/m Ar,
carrier (C) gas flow rate = 8 1/m Ar, reactant gas flow rate = 0.1 - 1.2 1/m
CH4, pressure = 760 Torr, RF power = 60 kW, su b strate tem perature =
N/A, su b strate pre-treatm ent = diamond paste polished.
Notes on reactor operation: Pyrometric m easurem ent of substrate
tem perature was not possible because of th e high reflection of plasm a
em ission from the substrate.
The film thickness w as not uniform. One of th e films w as 12 pm
thick n ea r the edge and 6 Jim thick a t the center of th e substrate. The
growth rate is 60 p m /h o u r for films with well defined crystal planes. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Carrier + Reactant Gas
(Ar+CH^
Plasma
Gas (P)
(Ar) —
(Ar+H2)
Sheath
Gas (S)
O
w
—
R F
Coil
Substrate
Water
Substrate Holder
Figure 2.7 RF Induction Therm al Plasm a Reactor [291
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
essential condition for diamond growth in the present therm al plasm a
CVD seem s to be the cooling of a substrate.
The “figures of m erit” th a t are available are a s follows:
(1)
Linear growth rate (pm /hour) - 60;
(2)
Deposition area (cm2) - 3;
(3)
Weight gain (mg/hour) - 63;
(3)
Uniformity ~ 50%;
(4)
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) ~ 1.05;
(ii)
Growth ra te /g a s in p u t (mg/liter) ~ 0.014;
(iii)
Carbon conversion efficiency (%) ~ 0.2
This reactor h as the advantage of high growth rate. It h a s the
following disadvantages. Due to the high gas tem perature of th e therm al
plasm a, control of substrate tem perature is difficult. The films obtained
have poor adhesion to the substrates an d the film thickness is n ot
uniform.
2.2.5
Microwave Plasm a Enhanced CVD Reactors
2.2.5.1
Introduction
Diamond film deposition using microwave plasm a h as been
achieved by several different reactors. Since this dissertation concerns
w ith th e development of microwave p lasm a reactors for th e deposition of
diam ond film, variations of microwave reactors are reviewed in more
detail.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
2.2.5.2
T ubular Microwave Reactor
A schem atic drawing of a typical tu b u la r microwave diamond film
deposition reactor is shown in Figure 2.8<30>. As shown, in this reactor,
th e in p u t gas which is a m ixture of hydrocarbon an d hydrogen is
dissociated by the microwave energy coupled into the quartz tube
through a set of waveguides, power m onitors and tuners. The plasm a
w as adjusted to be a t the center of th e q uartz tube. Microwave power
generated by a m agnetron is used to su sta in the microwave discharge.
The su b strate is place in the middle of th e quartz tube. Diamond film is
formed on the su b strate w hen the dissociated gas species react on its
surface.
Typical diamond film deposition conditions are as follows: quartz
tube diam eter - 40 mm, substrate = silicon wafer, gas mixture CH4/H 2 =
1 - 3%, total flow rate = N /A pressure = 7.6 - 60.8 Torr, microwave
power = 300 - 700 W, su b strate tem perature ~ 800 - 1000 °C.
The “figures of merit" th a t are available are a s follows:
(1)
Linear growth rate (pm/hour) ~ 3* pm /hour;
(2)
Deposition area (cm2) - 3 - 6 ;
(3)
Weight gain (mg/hour) - 3 - 6 ;
(3)
Uniformity = n /a ;
(4)
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) - 4.3 - 8 .6 ;
(ii)
Growth ra te /g a s in p u t (mg/liter) = n /a ;
(iii)
Carbon conversion efficiency (%) = n /a .
* It is noted th at, in the paper, the au th o r showed the SEM
photograph of a film deposited in the reactor for 3 hours. The film
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
H2 + CH4
Quartz tube
Wave guide
Substrate
Plunger
Sleeve
Magnetron
Vacuum Pump
Figure 2.8 T ubular Microwave Reactor [30]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
showed discrete particles of ~ 2 pm in diam eter and it is not continuous.
It is n o t clear from the paper how the reported growth rate of 3 pm /hour
w as arrived at and w h at m easurem ent technique w as u sed to determine
th a t film thickness.
2.2.5.3
Magneto-microwave Reactor
The schematic drawing of a magneto-microwave discharge
diamond film deposition reactor is shown in Figure 2.9.<31> As shown,
the in p u t gas, which is a m ixture of hydrocarbon an d hydrogen, is
dissociated by the microwave energy coupled into th e cylindrical
waveguide (discharge area) from a rectangular waveguide. The cylindrical
waveguide is excited w ith TEj x mode. The discharge is sustained by the
microwave power from a microwave power source. The m agnetic field
generated by th e Helmholtz-type magnetic coils is u sed to enhance
reactive gas dissociation through electron cyclotron resonance (875
G auss for 2.45 GHz microwave). The substrate is placed n ear the
discharge and the su b strate h eater is used to control su b strate
tem perature. Diamond film is formed on the substrate w hen the
dissociated gas species react on its surface.
Typical experimental conditions are as follows: cylindrical
waveguide inside diam eter = 1 6 0 mm, substrate = Si, su b strate pre­
treatm ent = scratched by 3 m m size diamond paste, gas m ixture CH4/H 2
= 0.5 - 5%, total flow rate = 100 seem, pressure = 0.1 - 50 Torr,
microwave power = 500 - 600 W, su b strate tem perature = 800 - 900 °C.
The “figures of merit" are n ot available in the literature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Coils
Hi
Substrate
3
Heater
Discharge
Gas inlet
Rectangular
waveguide
Cylindrical
waveguide
i
Figure 2.9 Magneto-microwave Reactor [31]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
2.2.5.4
Magneto-active Microwave Reactor
The schem atic drawing of a magneto-active microwave diamond
film deposition reactor is shown in Figure 2.10.<32> As shown, in this
reactor, th e in p u t gas, which is a m ixture of hydrocarbon, hydrogen and
oxygen, is dissociated by the microwave energy coupled into the
cylindrical process cham ber through the quartz window and cylindrical
waveguide section. The discharge is sustain ed by the microwave power
from a microwave power source. Electron cyclotron resonance (ECR) is
used to enhance the gas dissociation. The ECR magnetic field is provided
by th e perm anent magnet. The su b strate is located below the ECR zone
and heated during deposition by a tu n g sten filament located above the
su b strate. Diamond film is formed on th e su b strate when the dissociated
gas species react on its surface.
Typical experimental conditions: cylindrical cham ber diam eter = 25
cm, ECR m agnet = Nd-Fe-B perm anent m agnet (15 x 15 x 9 cm3) with
3250 G au ss a t the lower m agnet face, ECR zone (875 Gauss) = 8 cm from
th e lower m agnet face, substrate location = 6 cm below the ECR zone,
filam ent tem perature = 2040 °C, filam ent-substrate separation = 0.6 cm,
su b stra te pre-treatm ent = scratched w ith 1 jjm diamond paste, pressure
= 2 - 2 0 Torr, microwave power = 950 W, su b strate tem perature = 540 640 °C, CH 4 flow rate = 6 seem, O 2 flow ra te = 2 seem, H2 flow ra te = 400
seem.
Notes on reactor operation: The filam ent is electrically insulated
from th e cham ber wall to eliminate any electron emission current. The
su b stra te tem perature is monitored by a thermocouple embedded in the
su b stra te support. The tem perature depends m ainly on the filam ent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
Permanent
magnet
Microwave
■ power
Gas
inlet
Quartz
window
Filament
Thermocouple
Substrate
Pump
Figure 2.10 Magneto-active Microwave Reactor (32]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
tem perature, b u t there is some microwave heating. 250 - 290 °C, for
typical in p u t microwave levels of 400 - 1000 W. The pressure range 2 20 Torr is low for diam ond growth b u t substantially higher th a n th e subTorr p ressu res typically used w ith ECR plasm as.
The “figures of merit" th a t are available are a s follows:
(1)
Linear growth rate (pm /hour) ~ 2**;
(2)
Deposition area (cm2) - 6.45;<33>
(3)
Weight gain (mg/hour) - 4.5;
(3)
Uniformity = n /a ;
(4)
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) ~ 4.7;
(ii)
Growth ra te /g a s in p u t (mg/liter) - 0.18;
(iii)
Carbon conversion efficiency (%) ~ 2.3.
** It is noted th at, in the paper, th e au th o r showed the SEM
photographs of diamond films which consist of discrete particles of ~ 4.4
- 9.7 pm in diameter. These films are n o t continuous. It is not clear from
the paper how the growth rate of 2 p m /h o u r mentioned in the paper w as
calculated an d w hat m easurem ent technique w as used to determ ine th a t
film thickness.
2.2.5.5
Bell-jar Microwave Reactor
The schem atic drawing of a bell-jar microwave diamond film
deposition reactor is shown in Figure 2.11.<34,35> As shown, in this
reactor, the in p u t gas which is a m ixture of hydrocarbon, hydrogen and
oxygen is dissociated by the microwave energy coupled into the
cylindrical cavity through the waveguide an d antenna. A ball-shaped
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Waveguide
Cylindrical cavity
Quartz bell jar
Plasm a
Substrate holder
Quartz
cylinder
Gas m ixtures
Vacuum pump
Temperature
controller/m eter
Figure 2.11
Bell-jar Microwave Reactor [34,35]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
plasm a is formed inside the quartz bell ja r an d the substrate is placed
n ear the plasm a. Diamond film is formed on th e su b strate w hen the
dissociated gas species react on its surface.
Typical experimental conditions: su b stra te = Si, pressure = 4 0 - 7 0
Torr, su b strate tem perature = 850 - 1030 °C, CH4/H 2 = 0.5%, total flow
rate = 200 - 600 seem.
Notes on reactor operation: Coatings of more th an 65 mm (2.75
inch) diam eter were deposited on silicon w ith a separate substrate
heater. Pressure variations between 40 an d 70 Torr were found to have
little effect on the diamond growth rate in th is reactor. The variation of
the total flow between 200 seem and 600 seem (while keeping the
m ethane concentration constant a t 0.5%) did not affect the growth rate.
The figures of m erit th a t are available in the literature are:<36>
(1)
Linear growth rate (jim/hour) - 3.5:
(2)
Deposition area (cm2) - 12.5;
(3)
Weight gain (mg/hour) - 15.4;
(3)
Uniformity ~ n /a ;
(4)
Growth Efficiencies
2.2.5.6
(i)
Growth rate/pow er in p u t (mg/kW-hr) - 10.3;
(ii)
Growth rate /g as in p u t (mg/liter) ~ 28.5;
(iii)
Carbon conversion efficiency (%) - 4.2.
Microwave J e t Reactor
The schem atic drawing of a microwave je t reactor is shown in
Figure 2.12.<37,38> As shown, the in p u t gas, which is a m ixture of
hydrocarbon, hydrogen an d oxygen, is dissociated n ear the je t nozzle by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Water in
Water out
Quartz
Microwave
power
Gas in
Gas in
Water in
Water out
Substrate
To pump
To exhaust
Water out
t
Water in
Figure 2.12 Microwave J e t Reactor [37,38]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
the microwave energy. Microwave energy is transm itted from th e power
source to the je t nozzle through TEqi mode in the rectangular waveguide,
a transition u n it an d TEM mode in the coaxial waveguide. The plasm a je t
is generated from the end of the center p lasm a flow stabilizer an d blows
into the cham ber where th e substrate is placed on a water-cooled
substrate holder. Diamond film is formed on the substrate w hen the
dissociated gas species react on its surface.
Diamond films have been deposited u n d er the following
experimental conditions: substrate = Si, Ar flow rate = 1 0 1/min, H2 flow
rate - 20 1/min, CH 4 flow rate = 0.6 1/min, O 2 flow rate = 0.15 1/min,
total p ressure = 760 Torr, su b strate tem perature = 887 - 927 °C,
microwave power = 3.8 - 4.2 kW.
Notes on reactor operation: The diam eters of the center and outer
conductor in the coaxial waveguide are 20 an d 57.2 mm, respectively.
The conductors tap er off an d play the roles of plasm a flow stabilizers for
plasm a generation. These stabilizers are m ade of water-cooled copper in
order to prevent therm al evaporation. The edge of th e outer electrode
(plasma je t nozzle) is 22 m m in diameter, which m u st be designed
properly depending upon th e plasm a gas composition.
The “figures of merit" th a t are available are as follows:
(1)
Linear growth rate (pm/hour) ~ 12;
(2)
Deposition area (cm2) ~ 2.5;
(3)
Weight gain (mg/hour) - 10.53;
(3)
Uniformity ~ n /a ;
(4)
Growth Efficiencies
(i)
Growth rate/pow er in p u t (mg/kW-hr) - 2.6;
(ii)
Growth ra te /g a s in p u t (mg/liter) ~ 0.0057;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
(iii)
2.3
Carbon conversion efficiency (%) - 0.055.
Comparison of Diamond Film Deposition Reactors
A sum m ary of performance “figures of m erit” of diamond film
deposition reactors is given in Table 2.1.
Since there are no electrodes present in the microwave plasm a, the
problem of metallic contam ination in the process of diamond deposition
does not inherently exist. Contamination from reactor walls can be
minimized with a proper reactor design. Com pared to th e erosion of
filaments in hot filam ent reactors, erosion of electrodes in direct current
reactors and nozzle erosion in com bustion flame reactors, microwave
plasm a diamond film deposition is a cleaner process. It is also easier to
control and optimize th e deposition process which m akes microwave
plasm a reactors a prom ising technique for growing pure, uniform, and
high quality diamond films.
To successfully commercialize diamond synthesis a t low pressures,
diamond growth at high rates and low deposition tem peratures on large
area su b strates is desirable. Each of the microwave reactors described
above h as its advantages an d disadvantages for diam ond film deposition.
W hat their disadvantages have in common is th a t they can n ot be easily
scaled up for large area diamond film deposition.
The tu b u lar reactor h as two disadvantages, first, th e su b strate size
is limited by the by the inside diameter of the silica tube, which is 40
mm, and the system is n o t easily scalable for diamond film growth on a
larger surface since the diam eter of the silica tube is limited by the size
of th e rectangular waveguide; second, the plasm a generated inside the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
Table 2.1
Sum m ary of "Figures of Merit" of Diamond Reactors
Linear
growth
rate
(lim/hr)
Dep.
area
(cm2)
Weight
gain
(mg/hr)
Energy
effic.
(mg/
kW-hr)
Gas
flow
effic.
(mg/
liter)
Carbon
conv.
effic.
(%)
0.6
6.4
1.35
1.04
0.2
0.32
DC
p lasm a
[23,24]
20
0.25
1.76
0.56
1.47
13.7
DC
je t
[251
900
2
632
57
0.26
1.97
Flame
[271
60
0.28
5.9
n /a
0.049
0.009
RF
therm al
[291
60
3
63
1.05
0.014
0.2
T ubu lar
MW
[301
3
CO
■
3 -5
4.3 - 8.6
n /a
n /a
Mag.active
[321
2
6.45
4.5
4.7
0.18
2.3
Bell-jar
MW
[36]
3.5
12.5
15.4
10.3
28.5
4.2
MW
je t
[37,381
12
2.5
10.53
2.6
0.0057
0.055
CO
Hot
filam ent
[22]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
silica tube is very close to the inside walls, u n d er th e conditions suitable
for diamond deposition, erosion of the silica walls an d hence
contam ination of the diam ond film are likely.
The magneto-microwave reactor uses a narrow , electron cyclotron
resonance region to generate th e high density plasm a. Non-uniformity in
the deposited film can be expected as an inherent resu lt of the nonuniform distribution of the resonant magnetic field across the substrate
surface, especially for diam ond film deposition over a larger surface. The
reported surface area for diamond growth w as 30 m m in diameter. It h as
a narrow pressure region for diamond film growth, nam ely from 4 to 50
Torr. At pressures above 50 Torr, the discharge a re a becomes unstable
and below 4 Torr, the products contain graphite or SiC phase and in
extreme cases th e products are only these phases. The magneto-active
microwave reactor u ses the sam e principle as th e magneto-microwave
reactor. Thus it is expected th a t it has the sam e disadvantages as the
magneto-microwave reactor.
The bell-jar reactor uses a plasm a ball and th e substrate is located
near the lower pole of th e plasm a ball. The reactive species distribution
over the su b strate surface is inherently non-uniform . This is especially
true w hen the su b strate surface extends fu rth er away from the lower
pole of the plasm a ball. The reported coating surface area with the
plasm a is 25 m m in diameter. A separate h eater/cooler is needed to coat
a surface of 65 mm in diameter. The location of th e su b strate and
substrate holder are fixed in order to generate th e ball shaped plasma.
External tuning is needed in order to m in im ize th e reflected power from
the reactor since there is only one internal ad justm ent (i.e. the antenna)
available for microwave coupling.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
The je t reactor u ses a sm all je t nozzle as an exit orifice for the
reactive species. It can be used to deposit diamond films -with higher
growth rates over small areas. B u t it can n o t be used to deposit uniform
diamond film on large su b strates since fine particles of am orphous
carbon were deposited on the su b strate edge. This came from th a t the
su b strate w as not heated uniformly, hotter near th e center and cooler
near the edge.
2.4
Bachm ann C-H-O Phase Diagram
A C-H-O phase diagram w as introduced by P. K. B achm ann et al.
<36>
provide a common schem e for all major diamond chemical vapor
deposition (CVD) m ethods used to date. A schem atic drawing of this
phase diagram is shown in Figure 2.13. Each side of th e equilateral
triangle represents th e atom fractions of th e gas phase composition of
one of the binary system s C-H, H-O and O-C, ranging from 0 to 1. Any
ternary C-H-O gas compositions are located inside th is triangle. There
are three regions in th is phase diagram, i.e., the non-diam ond carbon
growth region, the diam ond growth domain and the no growth region. In
chapter six, the gas compositions used in the deposition experiments are
m apped onto this phase diagram.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
C
0.9
non—diamond
carbon
growth region
0.9
no growth region
0 /( 0 + H )
Figure 2.13
0.9
B achm ann C-H-O Phase Diagram [36]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER THREE
THE MICROWAVE-CAVITY PLASMA REACTORS (MCPR)
3.1
Introduction
In order to deposit diamond films on large su b strate surfaces,
diam ond film deposition reactors th a t can create large area discharges
are needed. These reactors should be designed su ch th a t they are easily
controllable and the deposition processes can be easily repeated. In this
chapter, a series of microwave-cavity plasm a reactors (MCPRs) are
investigated in order to deposit diamond films uniformly on large
su b strate surfaces. Cylindrically symmetric cavity applicators which
utilize internal m atching and w hich are excited in a single
electromagnetic mode are investigated for their ability to deposit diamond
films over large surface areas. Earlier work with this type of reactor
included the development of plasm a ion sources and plasm a etching and
oxidation reactors.<12,13,39,40> This type of microwave applicator h a s the
advantage of easy operation an d adaptability to different su b strate sizes
and types. High quality diamond film can be produced and easily
reproduced in these reactors. The difference between th e different reactor
concepts investigated here is the size of plasm a th a t can be safely created
and the associated substrate area th a t can be uniformly covered with
diam ond films.
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
There are a n u m b er of reactor geometry param eters in the plasm a
reactor design. These param eters include: (1) cavity shape and size, (2)
power coupling m ethod and (3) base-plate design, etc. All these reactor
geometry param eters have a strong influence on th e reactor performance.
In th e research described in this chapter, a n u m b er of these param eters
are investigated. They include: (1) 7", 14" an d 18" inside diameter
cylindrical cavities, (2) probe side and end coupling, and loop side
coupling, (3) different size and shape quartz dom es an d their associated
base-plates. These reactors are described below in the order in which
they were experimentally investigated. The experim ental microwave
power supply, reactive gas supply and vacuum pum p system s used in
the investigation of these reactors are first reviewed in Section 3.2. The
com puter m onitor system used in the experim ents described in chapter 6
is also described in section 3.2.
3.2
Experim ental Systems
3.2.1
Introduction
All th e microwave-cavity plasm a reactors th a t are described in this
dissertation u se microwave power as their energy source and a single
vacuum pum p system w as used for reactive gas supply and operating
pressure control. The microwave power supply, gas flow and vacuum
pum p system u sed in th e experimental evaluation of MCPRs are
described below in sections 3.2.2 and 3.2.3. The com puter monitor
system used in the experim ents described in ch ap ter 6 is described in
section 3.2.4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
3.2.2
Microwave Power Delivery System s
The experim ental microwave circuit used to deliver power into the
microwave applicators is sim ilar to those described in earlier work
< 12, 13.39 .40 > T h u g only a brief description is given here. A schematic
drawing of the microwave power supply circuit is shown in Figure 3.1. It
consists of a 2.45 GHz or 915 MHz, variable power source (1), circulator
(2) and m atched dum m y load (3), dual-directional coupler (4) and
associated incident an d reflected power m eters (5) and (6), and
microwave cavity applicator (7). The microwave power supplied by the
power source (1) propagates through the circulator (2) and the dualdirectional power coupler (4) and is incident on th e microwave cavity’
applicator (7). In th e case of any m ism atch betw een the impedances of
the waveguide an d th e cavity applicator (7), som e of the incident power
will be reflected back from the cavity applicator. This reflected power
travels in th e opposite direction from the incident power. It propagates
through th e dual-directional coupler (4) and is directed by the circulator
(2) into th e m atched dum m y load (3), where it is absorbed and dissipated
as therm al energy. The circulator (2) an d its m atched du m m y load (3)
perm it the propagation of th e microwave energy into th e cavity applicator
(7) and prevent th e propagation of th e reflected power back into the
power source (1) w here it m ay cause dam age to th e power source. The
incident power Pj is m easured by the incident power m eter (5) attached
to the incident power sampling port of th e dual-directional coupler (4)
and the reflected power Pr is m easured by th e reflected power m eter (6)
attached to the reflected power sampling port. The microwave power
absorbed by th e cavity applicator (7) is given by Pt = Pj - Pr.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
(7) Microwave
Cavity
Applicator
3
(4) D ual-directional
Power Coupler
(2) Circulator
r-------
o
(6) Reflected Power Meter
(5) Incident Power Meter
(3) Dummy Load
(1)
2.45 GHz
or
915 MHz
Microwave
Power
Source
Figure 3.1
Microwave Power Supply Circuit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Four microwave power sources have been u sed during the research
described in this dissertation. The first one is th e Chuang (Model No.
CA2450-1200CW) 2.45 GHz, 1.2 kW power source. It was used in th e
experim ents conducted in the first generation MCPR, the 2.45 GHz
experim ents conducted in the 18 an d 14 inch MCPRs and the low power
experim ents conducted in the second generation MCPR. The second
power source used is the Chuang (Model No. MPS915-500) 915 MHz, 500
W power source. It was the only 915 MHz power source available in our
laboratory an d was used in the 915 MHz experim ents conducted in th e
18 inch MCPR. Due to its low power capacity, it did not find m uch u se in
diam ond film deposition process. The th ird power source used is the
Gerling (Model No. GL119) 2.45 GHz, 3 kW power source. It was u sed in
the high power (1.1 - 1.7 kW) experim ents conducted in the second
generation MCPR and the low power ( 1 . 5 - 3 kW) experiments conducted
in th e th ird generation MCPR. The fourth power source used was the
Cober (Model No. S6F/4503) 2.45 GHz, 6 kW power source which w as
used in the high power ( 3 - 4 . 5 kW) experim ents conducted in the third
generation MCPR. Among the 2.45 GHz power sources available, th e
Gerling 3 kW power source h ad the narrow est a n d m ost stable frequency
spectrum and th u s was the “cleanest” power source.
3.2.3
Flow Control and V acuum Pum p System s
A schem atic drawing of the gas flow control and vacuum pum p
system s is shown in Figure 3.2. As shown, th e source gases, H2 (1). CH4
(2) an d CO2 (3), are supplied by cylinder ta n k gases of high purity
(99.99% p urity or better). The gas flows are controlled by three MKS type
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
(10) Cooling
Water
(4) Mass Flow
Controllers
(5) Base-plate
(9) Air
Cooling Port
(7) Microwave
Cavity
Applicators
(8) Discharges
(6) Cooling
Water
( 1)
(13) Process Chamber
I
(17) TC2*
(15) Low
Pressure Gauge
(18) Manual
Throttle Valve
i
(22) Gate Valve
1 1
1
1
(23) Throttle
Valve
(19) Automatic
Throttle Valve
(12) TCi*r
3
(16) Isolation
Valves
(24) Baffle
T
(14) High
Pressure Gauge
r\* 2
(21) Foreline
Valve
(20) Diffusion Pump
V®'
(26) System
fn>feni
Vent
Flow Meter
Figure 3.2
(11) Mechanica
Roughing
Pump
* TCj and TC2 are Thermal
Conductivity Vacuum Gauges
(27) Exhaust
Flow Control an d V acuum System for MCPRs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
1159A m ass flow controllers along w ith a MKS type 247C power su p p ly /
digital re a d o u t/se t point source (4). The flow control ranges of th e m a s s
flow controllers are 0 - 7.2 seem CH4, 0 - 7.4 seem CO2, 0 - 500 an d 0 10,000 seem H 2' The gases are mixed before they reach the base-plate
(5). Cooling w ater (6) was also supplied to th e base-plate (5). The
microwave cavity applicator (7) sits on top of the base-plate (5).
Discharges (8) are confined in th e lower section of the cavity applicator
(7) w hich is usually cooled by b o th air (9) an d w ater (10)
The m echanical roughing pum p (11) is the ALCATEL 2033 pum p
w hich h a s an air pum ping speed of 10.9 1/s (23 cfin). It has a base
pressu re of 7.5 x 10*2 mTorr (1 x 10~4 mbar) w hen used with ALCATEL
200 pum p oil w hich is a type of m ineral oil distilled under vacuum . This
type of oil can be u sed to pum p corrosive products and provides reduced
back-stream ing. The pum p base p ressure is m easured by a MKS type
286 therm al conductivity vacuum gauge TCj (12) which h as a
m eaningful p ressure m easurem ent range of 5 - 200 mTorr. The pressure
in the process cham ber (13) is m easured by a MKS Baratron type 122A
1000 T orr full scale high pressure gauge (14) for process pressure
m easurem ent and a MKS B aratron typel27A 0.1 Torr hill scale low
pressu re gauge (15) for the cham ber b ase p ressu re m easurem ent. These
two pressu re gauges are connected to th e cham ber (13) through isolation
valves (16) so th a t they can be isolated from the cham ber whenever not
in use. The isolation valves (16) en su re th a t accurate pressure readings
can be obtained by stabilizing the zero point of the pressure gauges. MKS
type PDR-C-1C and type PDR-D-1 power supplies/digital readouts are
connected to these two pressure gauges an d display the pressure
readings. A MKS type 286 therm al conductivity vacuum gauge TC2 (17)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
with m eaningful p ressu re m easurem ent range of 5 - 200 mTorr is used
for general indication of the pressure in the process chamber. It is quite
useful during th e cham ber initial pum p down after the cham ber is
vented to the atm ospheric pressure for cleaning, sam ple loading a n d /o r
unloading, etc.
The pressure in th e process cham ber (13) is controlled by throttle
valves between the cham ber (13) and th e m echanical pum p (11). A
Varian m anual throttle valve (18) was used to control the cham ber
pressure in the experim ents conducted in th e first and second generation
MCPRs an d some of th e earlier experiments conducted in the third
generation MCPR. A MKS type 253A 20 m m sealing butterfly throttling
valve w ith MKS type 252 exhaust valve controller (19) w as used to
control the cham ber p ressure in the later experim ents conducted with
the third generation MCPR.
A lower pressure pum ping system is also p a rt of th e vacuum
system. It was u sed to calibrate the zero point of th e low pressure gauge
(15) and check vacuum tightness of th e vacuum system in the
experiments described in this dissertation. A V arian VHS-6 (Model No.
0184) 6 inch diffusion pum p (20) provides the lower pressure pumping
capacity to the vacuum system. This diffusion pum p h a s an optimum
operating pressure range of 1 x 10"3 to 1 x 10 '9 Torr w ith an air pumping
speed of 2400 1/s. The fore-line valve (21) connects or isolates the
diffusion pum p (20) from the mechanical pum p (11). A Varian 8 inch
viton-sealed swing gate valve (22) provides the isolation or connection of
the process cham ber (13) to the diffusion pum p (20). The process
cham ber pressure is controlled by a MKS 6 inch throttle valve (23). A
Varian low profile w ater cooled baffle (24) is u sed to prevent the diffusion
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
pum p oil from back stream ing into the process cham ber (13). By closing
the m anual valve (18) an d /o r the 20 m m throttle valve (19) on the
cham ber direct roughing line, and opening th e gate valve (22) and fore­
line valve (21), the lower pressure pum ping system can be used. By
closing the gate valve (22) and th e fore-line valve (21), the lower pressure
pum ping system is isolated from the rest of th e vacuum system.
Nitrogen (25) is used for system vent (26) and exhaust (27) purge,
because flammable gases H2 (1) and CH 4 (2) are u sed in the experiments
described in th is dissertation. Nitrogen w ith flow rate of twenty (or more)
tim es of th e total flow rates of flammable gases is used to dilute the
exhaust gas m ixture in the exhaust (27) pipe so th a t th e resultant
exhaust gas m ixture is no longer flammable. Nitrogen is also used to vent
the process cham ber (13) during cleaning, sam ple loading and
unloading, etc. It helped to m in im ize the out-gassing from the cham ber
walls during cham ber pum p down.
3.2.4
Com puter Monitor System
A com puter safety monitor system<41> was used to monitor the
system operating conditions and control th e sh u t down procedure in the
experim ents described in Chapter 6 . A block diagram of this com puter
m onitor system is shown in Figure 3.3. The prim ary task of this
com puter m onitor system is to control experim ent time and sh u t down
sequence. It is also u sed to ensure the safe operation of the system. As is
shown in Figure 3.3, th e system operating pressure, reflected microwave
power and in p u t microwave power readings are u sed as input signals for
the com puter m onitor system. The operating sta tu s (ON/OFF or OPEN/
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
c
o
^
M
P
^
u
^
T
E
R
O perating P ressure
Reflected Microwave Power
E
In p u t Microwave Power
X
P
E
R
I
M
E
N
T
M
O
N
I
T
A
L
CH 4 Gas Flow ON/OFF
O
R
CO 2 Gas Flow O N /O FF
S
_
s
H2 Gas Flow O N/OFF
Y
S
T
E
M
Microwave Power ON/OFF
A utom atic Valve OPEN/CLOSE
Figure 3.3
S
vX
_
_
Com puter Monitor System Block Diagram
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T
E
M
55
CLOSE) of th e flow controllers, microwave power source and the
autom atic throttle valve is controlled by the com puter monitor system
through a series of relays.
This com puter m onitor system w as designed such th a t the default
state of the experim ental system is in a disabled state where the gas
flows an d microwave power are OFF and the autom atic valve is CLOSED.
A com puter program is needed to enable th e experimental system,
m onitor the experimental conditions and control the shut-down
sequence. The flow ch art of a com puter m onitor program<41> is shown in
Figure 3.4. As shown, in th is program, the operating pressure and
reflected microwave power upper limit and th e experiment running time
are first set. The experimental system is th e n enabled so th a t the gas
flows an d microwave power can be tu rn ed on and the autom atic throttle
valve can operate in a n autom atic mode to control the system pressure.
An experimental start-u p procedure (see section 6.3.3.2) follows to sta rt
th e experiment after which the tim er in the com puter monitor system is
started. During the experiment, a checking loop is used to compare the
operating pressure, reflected microwave power an d tim er with the pre-set
values to determine th e state of the experiment. If, a t any time during the
experiment, th e operating pressure a n d /o r th e reflected microwave power
go over the pre-set u p p er limit, the program directs the experiment into
an emergency shut-dow n sequence. In th is sequence, the microwave
power is turned off followed by the shut-off of all gas flows. The
autom atic throttle valve is th en closed to isolate the vacuum pum p from
th e process cham ber. These three steps follow each other very closely
w ith the speed of the com puter. Otherwise th e program directs the
system into a norm al shut-dow n procedure (see section 6.3.3.2) w hen the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
1. Set operating pressu re and
reflected MW power u p p er limit
2. Set experiment running tim e
3. Enable the experim ental system
4. Experimental sta rt-u p procedure
(see section 6.3.2.2)
5. S tart tim er
1. Check operating pressure
and reflected MW power
2. Check timer
Timer
expires
Normal shut-dow n
procedure
(see section 6.3.2.2)
Operating pressure
or reflected MW
power goes over
th e u p p er limit
Emergency shut-down
sequence
1. T urn off microwave
power
2. T urn off all gas flows
3. Close th e automatic
throttle valve
Figure 3.4 Monitor Program Flow C hart
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
tim er expires.
3.3
The First G eneration Seven Inch MCPR
3.3.1
Introduction
The first generation seven inch MCPR (MCPR7-1) was the first
reactor concept investigated for diam ond film growth. It was developed in
th e Fall of 1986 and b uilt by J . A sm ussen an d placed into operation in
the Fall of 1987 a t Norton Company in Salt Lake city, Utah. It w as based
on a microwave plasm a reactor concept invented a t Michigan State
University and patented by A sm ussen et al.<42"50> The technology is
licensed to Wavemat, Inc. by Michigan State University. Norton Company
h a s a n exclusive sub-license from Wavemat for the application of this
technology to diamond film deposition. The reactor u sed in our
investigation is a later version which w as commercialized by Wavemat,
Inc.
3.3.2
Reactor Geometry
The cross-sectional views of the MCPR7-1 are shown in Figure 3.5.
As shown, the reactor consists of the cylindrical cavity walls (1) which
form th e outer conducting shell of the cavity applicator. The cavity walls
are formed from a 7 inch inside diameter, open ended metallic cylinder. A
w ater cooled (41) sliding sh o rt (2), which is electrically connected to the
cavity walls (1) via the finger stock (3), forms the top end of the cavity. It
can be moved back and forth along th e longitudinal axis of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To Vacuum Pump
LEGEND
1) Cavity Walls
2) Sliding Short
(3) Finger Stock
(4) Cavity bottom surface
(5) Quartz Dome
(6) Discharge Zone
(7) Gas Input Tunnel
(8) Gas Distribution Ring
(9 ; Metal Screen
(10) Cooling Water Tunnel
(11) Graphite Base
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(40)
(41)
(42)
Quortz Tube
Substrate Holder
Substrate
Top Window
Optical Pyrom eter
Power Coupling Probe
Outer Conductor
Base Plate
Water Cooling
O -ring
Figure 3.5 Schem atic Drawing of MCPR7-1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
cylindrical cavity walls ( 1) to change the electrical and physical height of
the cavity applicator. The cavity bottom surface (4) and the water-cooled
(10) base-plate (40) form th e lower end of the cavity applicator.
Microwave power is coupled into th e cavity applicator through an
adjustable coaxial power in p u t port which is comprised of th e power
coupling probe (17) and its outer conductor (18).
Reactive gases flow into th e discharge zone (6) via the gas in p u t
tu n n el (7) inside the base-plate (40). The 9.25 cm i.d. quartz dome (5)
confines the working gas to the lower section of th e cavity applicator
where th e microwave fields produce a plasm a discharge adjacent to the
su b strate (14). A m etal screen (9), which is attached to the bottom of the
base-plate (40), allows gases to flow into the vacuum pum p system b u t
prevents microwave energy from radiating out of th e applicator. The
su b strate (14) and su b strate holder (13) are placed on top of a quartz
tube (12). The quartz tube (12) stan d s on a graphite base (11) which in
tu rn sits on the m etal screen (9). A plasm a discharge is ignited in the
dom e-shaped zone (6) by exciting th e cavity in a single discharge loaded
reso n an t mode. The plasm a discharge can be viewed through the top
screened window (15), through which the substrate tem perature can also
be m easured using a optical pyrometer (16).
3.3.3
Reactor Operation
The general experimental operation of this type MCPR h as been
described in detail elsewhere <51-55>. Thus, only a brief description is
given here. Differences from earlier work <51' 55> are operation a t high
pressure (30 - 80 Torr) and w ithout a static magnetic field. The ignition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
m atching, and internal cavity tuning are sim ilar to th a t reported earlier
for microwave electrotherm al th ru sters <56> and microwave broad beam
ion sources<51>.
This MCPR can create a microwave discharge w hen excited in a
single cavity electromagnetic mode. The mode diagram of an ideal 7”
cavity is show n in Figure 3.6.<57> Each mode w as experimentally
evaluated for its potential to deposit diam ond films a t discharge pressure
of 30 - 80 Torr. A discharge was started by reducing the H2/C H 4 gas
pressure to 5 -10 Torr, and applying microwave power which would then
ignite a discharge th a t entirely filled the quartz cham ber. The discharge
pressure w as th en increased to an operating condition of 30 - 80 Torr
while length and probe tuning th e cavity to a m atched condition.
At ignition the discharge completely filled th e quartz tube, b u t as
pressure increased to 30 - 80 Torr, the discharge contracted and
separated from the surrounding quartz walls an d assum ed a shape
related to th e field p attern of the exciting electromagnetic mode. T hat is,
th e discharge became “arc like” and its shape and position varied with
cavity mode excitation. This arc like behavior becam e particularly evident
a t the 40 - 80 Torr pressures.
Experim ents dem onstrated th a t one particular mode was superior
to all others for diamond film deposition. This particular mode w as
identified a s the TM0i i mode by electromagnetic field pattern
m easurem ents. A detailed description of field p attern m easurem ents is
provided in chapter 5. A schem atic drawing of the field pattern of the
TM0u mode in a discharge loaded cavity is shown in Figure 3.7. As
shown in Figure 3.5, w hen excited in this mode, the discharge hovered
over and w as in direct contact with the su b strate w hich was placed along
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.2
>>
o
c
<u
D
O'
Q
)
i_
2 .8
-
>111
>*n
li—
Constant Frequency--Variable Length
C
o 2.4c
o
w>
(U
0)
N
5
2 .0
-
O
Constant Length— Variable Frequency
Lit c m ( r e s o n a n t len g th )
Figure 3.6 Mode Diagram of an Ideal 7” Cavity
62
TM011
M ode
Discharge
- £ field
line
field
line
Figure 3.7 Field P attern of Discharge Loaded TMq h Mode
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
the cavity axis. Under discharge loaded deposition conditions, the cavity
length, Lg, w as set a t approximately 7.2 cm and th e probe position, Lp,
w as adjusted to about 1 .8 cm. Deposition experiments were performed
w ith th is electromagnetic mode.
3.3.4
S ubstrate Holder and Q uartz Domes
A variety of substrate (14) m aterials have been used for diamond
film deposition in this reactor. These m aterials include silicon wafers,
silicon nitride, carbide drill bits, tu n g sten wires, etc. The m ost frequently
used su b strate m aterials are silicon wafers since they are readily
available, common su b strate m aterials in integrated circuits and bench
m ark m aterials for reactor performance determ ination.
The su b strate holders (13) were u su ally m ade of graphite,
am orphous carbon or boron-nitride. In m ost holders, recess areas are
m ade on the top surface to accommodate th e substrate (14) m aterials to
prevent them from moving during the deposition processes. The typical
su b strate holder (13) size had a diam eter of 5 cm or smaller.
The cylindrical quartz tubes (12) were cu t from a quartz tube with
an outside diam eter of about 1". Typical quartz tube (12) length ranges
from 2.5 to 4 cm, depending on the experimental conditions, substrates
and su b stra te holders.
Besides holding the quartz tu b e (12), the graphite base (11) also
helps to attract the plasm a discharge (6) to the su b strate (14) so th a t
reactive species in the plasm a discharge (6) can be effectively utilized for
diam ond film growth.
The optim um su b strate (14) an d su b strate holder (13) position was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
found empirically by num erous experimental ru n s. The location varied
with su b strate size an d type, gas pressure an d flow, in p u t power, etc.
This reactor is designed to accept several height, 9.25 cm i.d.
quartz domes (5). There is a range for the optim um quartz dome (5)
height an d this optim um range depends slightly on su b strate size and
location. For quartz dome height below this optim um range, the hot
plasm a u sed for diam ond film growth m akes contact w ith the quartz
dome top surface. Film deposition on the quartz dome and in extreme
cases quartz dome overheating and melting m ay result. For quartz dome
height above this optim um range, a second plasm a appears in direct
contact w ith the quartz dome top surface, w eakening the intensity of the
plasm a in direct contact w ith th e substrate. Again, film deposition on the
quartz dome and in extrem e cases quartz dome overheating an d melting
m ay result. The heights of the quartz domes u sed in m ost diamond film
deposition experim ents described in this dissertation are 4.35 cm and 6
cm. The O-ring groove (42) of th e quartz domes is 2 cm below the cavity
bottom surface (4). T hus, the top of the domes is located 2.35 cm and 4
cm above the cavity bottom surface (4).
3.3.5
Reactor Performance
Good quality diam ond films have been deposited using th is reactor
on a variety of su b strate m aterials. The actual th in film deposition
performance of th is reactor have been described in detail elsewhere<58>.
Thus, only a brief description is given here.
Typical experim ental pressures vary from 30 Torr to 80 Torr using
m ixtures of hydrogen (50 seem - 400 seem) an d m ethane (0.5 seem to 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
seem). Under these conditions, absorbed 2.45 GHz power varied from
250 W to 800 W. The su b strate area is generally 4 cm by 4 cm or
smaller. The linear growth rate is in the range of 0.4 to 0.8 |im /hour.
A typical Ram an spectrum of a diamond film (WDF08) deposited in
this reactor under the following experimental conditions is shown in
Figure 3.8.
CH4 flow rate = 2.16 seem
H2 flow rate = 175 seem
Pressure = 72.5 Torr
Microwave Power absorbed = 440 W
Though good quality diamond films can be deposited in this
reactor, it h as the following limitations. First, the vacuum o-ring seal is
located close to the hot p lasm a/su b strate region. The su b strate h as an
operating tem perature of the order of 700 - 1100 °C. Hence for any
specific operating pressure, there is a limit on the absorbed power level
and su b strate area to prevent th e o-ring failure. Second, the power input
comes from the side of the cavity walls an d produces a n inherent nonuniform electromagnetic “near" field close to the excitation probe. For a
small area substrate (5 cm in diam eter or smaller), the effect is not
significant. But as the su b strate area a n d /o r in p u t power level is
increased, this “near” field effect gets stronger. The discharge is attracted
by the probe’s strong “n ear” field onto the quartz dome walls, creating a
non-uniform plasma, causing non-uniform film deposition on the
sub strate and heating u p the quartz walls.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I
!
1 0 8 0 0 .0
8 8 0 0 .0
CH4 = 2.16 seem
H2 = 175 seem
pressure = 72.5 Torr
MW power = 440 W
6 8 0 0 .0
2 8 0 0 .0
8 1 5 .2 0
4 2 0 .0 0
6 9 0 .0 0
9 7 0 .0 0
1 2 5 0 .0
1 5 3 0 .0
1 8 1 0 .0
RECIPROCAL CM
Figure 3.8 Raman Spectrum of a Diamond Film Deposited in MCPR7-1
67
3.4
The Second Generation Seven Inch MCPR
3.4.1
Introduction
In order to solve the near field problem faced in the MCPR7-1,
several versions of second generation MCPR were investigated for
diamond film growth.
3.4.2
Reactor Geometry
The second generation MCPR, called th e MCPR7-2 in short, is
schematically shown in Figure 3.9. As shown, the difference between this
MCPR and the MCPR7-1 is th a t the power probe coupling (19) is now
located at the top end of the cavity, instead of on th e side of the cavity.
The sliding sh o rt (2) and coupling probe (19) adjustm ents, Lg and
Lp. provide the internal cavity impedance tuning m echanism to minimize
the reflected power.
New 7" cylindrical cavity walls (1) were designed and built which
did n o t have the side feed port th a t existed in the conventional 7 inch
cavities. This cavity w as designed to be longer th a n th e conventional
cavity so th a t TM012 an d TM013 modes can be excited. There are two
w indow s/ports (22a,22b) on th e new cavity. The cooling air inlet (22a) is
for air in p u t to cool th e cavity and quartz dome (5). The w indow /air
outlet (22b) serves as th e exhaust port for th e cooling air and also as a
viewing port through w hich the substrate tem perature is m easured by an
optical pyrometer.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
LEGEND
22b-»-
Cavity Walls
Sliding Short
Finger Stock
Cavity Bottom Surface
Quartz Dome
Plasm a Discharge
Gas Input Tunnel
Gas Distribution Ring
Metal Screen
Cooling Water Tunnel
Graphite Base
Quartz Tube
S ubstrate Holder
S ubstrate
Input Coupling Probe
Outer Conductor
Cooling Water
Cooling Air Inlet
Window/Air Outlet
vwjf/j
Figure 3.9
Schematic Drawing of MCPR7-2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
3.4.3
TM01n Modes Excitation
The experimental operation of this reactor is similar to th e MCPR71. It u se s the same mode diagram of an ideal 7” cavity shown in Fig. 3.6.
and follows a similar starting procedure as th a t of MCPR7-1. The
difference between the th is reactor and MCPR7-1 is the mode th a t is
u sed in diamond film deposition.
TMqi i mode can be easily excited in this reactor. The plasm a
discharge (6) created is sym m etric with respect to th e axis of the cavity
walls (1). But the probe n ear field effect still exists in this reactor with
T M on mode excitation. U nder high input power conditions, th e plasm a
discharge (6) is attracted to the top surface of th e quartz dome (5) since
th e coupling probe (19) is located close to the top surface of th e quartz
dome w ith TMqi i mode excitation.
To eliminate the n ea r field effect, th is reactor is excited using either
th e discharge loaded TM012 or TM013 mode. The field patterns of TM012
and TM013 modes are shown in Figure 3.10. With TM012 or TM013 mode
excitation, the cavity length Lg is either doubled or tripled from th a t of
TMqi 1 mode excitation. Hence the near field of th e coupling probe (19) is
moved away from the quartz dome region and now h a s little effect on the
geometry of the plasm a discharge (6) th a t is created. U nder discharge
loaded TM012 mode excitation, th e cavity length, Lg, was set a t
approximately 14.4 cm an d the probe position, Lp, w as adjusted to about
3.2 cm. Deposition experim ents with this reactor (MCPR7-2) were
performed with this electromagnetic mode. Under discharge loaded
TMqi3 mode excitation, the cavity length, Lg, w as set a t approximately
21.6 cm and the probe position, Lp, was adjusted to about 3.2 cm. TMq13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
r\
C
7>
L.
s:
r\
Field Patterns of Discharge Loaded TMq12 and TM0 i 3 Modes
70
Figure 3.10
IN
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
mode w as used in the later p a rt of the experim ents conducted in the
third generation MCPR (MCPR7-3) which is described in section 3.6.
Another advantage of th is reactor over the MCPR7-1 is its reliable
operation. In the side feed configuration u sed in MCPR7-1, since
competing degenerate and n ear degenerate TM and TE modes can all be
excited, there exists interference from degenerate an d near degenerate
dipole modes to the symmetric mode excitation th a t is desirable for
• diamond film deposition. B ut in th e end feed configuration used in this
reactor, only symmetric TM modes are excited. Hence, the mode
excitation zone in th e cavity are better separated an d the interference
from degenerating TE m odes is eliminated. This m akes the operation of
this reactor easier and more reliable u n d er all the substrate operating
conditions th a n MCPR7-1.
3.4.4
End Feed Assembly
Initially, end feed assem bly with 7 /8 " inside diam eter probe sleeve
(20) an d 0.35" diam eter coupling probe (19)<59> w as tested for diamond
film growth in February, 1991. Good quality diam ond films were
deposited on silicon wafers. After a few experim ents, it was noted th a t
the finger stock (3) on the sliding short (2) becam e fragile. It was over
heated since the sliding short (2) to which th e 7 /8 " end feed assembly is
attached and the cavity walls ( 1) do not have w ater cooling to cool the
finger stock (3) in the high in p u t power conditions required for diamond
film deposition. Also, th e 7 /8 ” coaxial cable w as too small to handle the
high power (over 1 kW) required for diamond film deposition on larger
surface areas (5 cm in diam eter or larger).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
To solve the over heating and power handling problems, a new end
feed w ith 1 5 /8 " inside diameter probe sleeve (20) and 5/8" diameter
coupling probe (19) w as designed and built. It w as designed to be
attached to a w ater cooled (21) sliding sh o rt (2). This new end power feed
assem bly h a s been u sed to operate a t 4 kW over 20 hours in continuous
experiments a n d over 2000 hours in total ru n n in g time w ithout any
noticeable damage.
Air cooling is u sed to cool the cavity walls (1) and quartz dome (5)
u nd er high power in p u t conditions. The cooling air inlet (22a) was
initially designed and built to be 0.5” in diam eter an d compressed air
w as u sed as th e cooling air supply. It was a noisy an d unreliable air
cooling system since it sometimes overloaded the a ir compressor in the
Engineering Research Complex. T hat air cooling system was replaced by
a new design. In the new design, the cooling air inlet (22a) was enlarged
to 2 ” in diam eter and double screened window w as soldered onto the
cavity walls (1) to ensure good electrical continuity of the cavity walls ( 1).
A Dayton (model No. 4C443A) air blower w ith free air delivery speed of
100 cfm is u sed to supply the cooling air through th e enlarged cooling air
inlet (22a). This new system is a m uch quieter an d more reliable air
cooling system.
3.4.5
Reactor Performance
This reactor h a s been used to deposit diam ond films over larger
areas th a n th a t in MCPR7-1. The largest area covered was 5.7 cm in
diameter. The linear growth rates are in th e sam e range as in MCPR7-1,
namely, 0.4 - 0.8 |im /h o u r.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
The 7 /8 inch coaxial power end feed system w as u sed for diamond
him growth on 3 cm x 3 cm square silicon su b strates with th e following
typical experimental conditions:
CH4 flow rate = 2 seem
H2 flow rate = 300 seem
absorbed microwave power = 620 W atts
Pressure = 45 Torr
The 1 5 /8 inch coaxial power end feed system w as u sed for
diam ond film growth on 5.7 cm diam eter silicon su b strates w ith the
following typical experimental conditions:
CH4 flow rate = 3 seem
H2 flow rate = 300 seem
absorbed microwave power = 1600 W atts
pressu re = 40 Torr
A typical R am an spectrum of a diamond film (EDF-2) deposited in
th is reactor u n d er the following experimental conditions is shown in
Figure 3.11.
CH4 flow rate = 2 seem
H2 flow rate = 300 seem
absorbed microwave power = 560 W atts
p ressure = 50 Torr
With the end feed, this reactor design solves th e non-uniformity
problem caused by th e probe “near" field, b u t the o-ring over-heating
problem th a t existed in MCPR7-1 w as still a problem and limited the
performance of th is reactor.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
with permission of the copyright owner. Further reproduction prohibited without permission.
7000
CH4 = 2 seem
H2 = 300 seem
pressure = 50 Torr
MW power = 560 W
6000
CO 5000
c
CD
c
4000
3000
1100
1200
1300
1400
1500
1600
1700
frequency (cm*1)
Figure 3.11 Ram an Spectrum of a Diamond Film Deposited in MCPR7-2
75
3.5
3.5.1
Microwave Coupling Methods
Introduction
A series of experiments were performed to find the optimum
configuration for the excitation of microwave discharges for diamond film
deposition over large surface areas. The following excitation variations
were compared: (I) loop an d probe coupling, (II) cavity end feed and side
feed, and (III) TM0 n , TM0 i 2. TM0 i 3 and T E onm odes excitation.
3.5.2
Loop Coupling and Probe Coupling
In an attem pt to excite TE qi i mode, whose field pattern is shown
in Figure 3.12, a loop coupling system w as designed. It is schematically
shown in Figure 3.13. The loop coupling system w as realized by first
designing and constructing a series of open loops (arc) (35) from small
b rass rods. The probe in the side feed was taken o u t and replaced by
these open loops. The open loop was attached at one end to the end of
the center conductor of th e power input p o rt and pressed against the
outer conductor (18) wall at th e other end. The loop can be moved
horizontally to change th e area of the loop an d rotated to change the
orientation of the loop inside the cavity. A new outer conductor tube for
the power input port w as designed and built. To eliminate microwave
radiation leakage during the linear movement and rotation of th e loop,
this outer conductor tube did not have the narrow opening machined on
the original outer conductor tube for the indication of probe length in the
probe coupling.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
TE 011 M ode
- S field
line
- H field
line
Figure 3.12 Field P attern of TEqi i Mode
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Ezzzzzzza
VZZZZZZZZZZZZ1
v.'/,v.v
To Vacuum Pump
LEGEND
1
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Cavity Walls
Sliding Short
Finger Stock
Cavity bottom surface
Quartz Dome
Plasm a Discharge
Gas Input Tunnel
Gas Distribution Ring
Metal Screen
( 10!
(11
(12;
(13
14
(15
(16'
(18'
(35
Cooling Water Tunnel
Graphite Base
Q uartz Tube
S u b stra te Holder
S u b stra te
Top Window
P yrom eter Sensor
O uter Conductor
Power Coupling Loop
Figure 3.13 Schem atic Drawing of a Loop Coupling System
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
The cavity sh o rt length Lg, loop area an d loop orientation were the
three param eters varied to excite loaded cavity modes and m atch
microwave power into th e loaded cavity. It w as found th a t the TMqi i
mode w as the easiest mode to be excited and the associated plasm a was
produced. The TE0n mode w as n ot excited using th is loop. W hen the
TMqi i mode was excited w ith the loop, the coupling w as not efficient,
possibly due to the loose contact between th e o u ter conductor (18) wall
and th e open loop (35). This loose contact problem can not be easily
solved since the loop is rotated and also moved linearly in order to m atch
the in p u t power into th e cavity. Arcing takes place a t th e loose contacts
a t high in p u t power conditions. Hence, loop coupling can not be easily
used a t high in p u t power conditions w hich are required for diamond film
growth over larger surface areas.
Compared w ith loop coupling, probe coupling does not have the
loose contact and arcing problems. T hus it becam e the favorable choice
for the excitation of plasm a for diamond film deposition.
3.5.3
End Coupling and Side Coupling
An end feed can excite symmetric TM modes and keep the near
field away from the p lasm a/su b strate region by exciting higher order
modes in diamond film growth process. These are advantages over the
side feed where the probe is located on the cavity side wall and a nonsymmetric field is always produced. At high in p u t power conditions, a
reactor using side feed faces the problem caused by n ear field if it is to be
operated at lower order m odes. It also faces the competing excitation
problem caused by degenerating or closely degenerating dipole TE and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
TM m odes if it is to be operated a t higher order modes. By contrast, the
end feed excitation m akes the operation of the diamond film deposition
reactor easy and reliable by selective excitation of symmetric TM modes.
It is a favorable choice for the excitation of microwave plasm a for uniform
diam ond him deposition on larger surface areas.
3.5.4
Mode Excitation
The mode patterns of TM qh, TM012. ™ o i 3*
TE0n are shown
in Figures 3.7, 3.10, and 3.12. In side feed probe coupling and loop
coupling, TMqi i mode is th e easiest mode to be excited b u t the n ear
electromagnetic fields attract th e plasm a to the side walls in the case of
■*r
high power input. In end feed probe coupling, TE0i i mode can n o t be
easily excited and TM qh, TMqi2 an d TM013 modes are easily excited. But
in th e case of TMqi i mode excitation, the n ea r fields attract th e plasm a
to th e quartz disk top wall and limit high power input. The TM012 or
TMqi3 mode is the ideal mode to be u sed for plasm a excitation for
diam ond film deposition in th e end feed probe coupling configuration,
since th e n ea r field effect can be reduced by keeping the probe away from
the quartz disk, and high power can be in p u t to produce a large
cylindrically symmetric plasm a.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
3.6
The Third Generation Seven Inch MCPR
3.6.1
Introduction
In order to be able to deposit diam ond film uniformly on large
surface areas, a new 7 inch microwave plasm a disk reactor was designed
and built. It w as first tested for diamond film growth in May, 1991.
3.6.2
Reactor Configuration
3.6.2.1
Introduction
This reactor w as designed to operate in various configurations.
These configurations include the basic configuration, forced flow
configuration, je t configuration, lower cavity resonance configuration and
down stream configuration, etc. A detailed description of each
configuration is provided in following sections. Preliminary versions of
these configurations have been investigated for diamond film growth.
3.6.2.2
The Basic Configuration
The principle components of the basic configuration reactor are
displayed in the cross-sectional view of Figure 3.14. This reactor u ses the
optimized end feed probe coupling. The power feed is the 1 5 /8 inch
coaxial power in p u t assem bly (19, 20). The end feed assembly (19,20)
along w ith th e sliding sh o rt (2) form top end of th e cavity. The lower
section of the cavity consists of the bottom surface (4), the base-plate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
m -n
23 n
L3?SNQ.
(1) Cavity Walls
(2) Sliding Short
(3; Finger Stock
(4 i Cavity Bottom Surface
i Quartz Dome
(6) Plasma Discharge
(13) Substrate Holder
(1 4 ) Substrate
(19) Power Coupling Probe
(20) Probe Sleeve
(21) Cooling Woter
(22a) Cooling Air Inlet
(22b) Window/Air Outlet
Gas Input Tunnel
Distribution Plate
Air Cooling Tunnel
Cooling Water Tunnel
Laser Port
Quartz Tube
Metal Tube
Metol Choke
Metal Plate
Base Plate
Quartz Flange
0 -r in g Seal
Figure 3.14 The Basic Experimental Configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
(43), a m etal tube (resonance breaker) (29), a metal plate (31) and a metal
choke sleeve (30).
The reactive gases, w hich are supplied through the source gas
input tu n n el (23) and an n u lar source gas ring on th e gas distribution
plate (24), is confined a t the lower section of the cavity by th e quartz
dome (5). The quartz dome (5) is enlarged in both diam eter and height.
W ater cooling is incorporated in the sliding sh o rt (2) an d th e base plate
(43) by using annular w ater cooling tunnels (21, 26). There is also a air
cooling tunnel (25) b u ilt in th e base plate (43) for the cooling of the
quartz dome walls u n d er high in p u t power operations. Windows (27) are
incorporated in the b ase plate for laser induced fluorescence diagnosis of
the microwave plasm a u n d er diam ond film deposition conditions. A to p /
cross-sectional view of the base plate (43) is shown in Figure 3.15.
The substrate (14) to be coated w ith diam ond film lays on top of a
sub strate holder (13) w hich is supported by a quartz tube (28). Different
height quartz tubes (28) were u sed to change to th e position of the
su b strate w ith respect to the plasm a discharge (6). If available, a
m echanical moving stage can be used to change th e su b strate position.
The m etal tube (29) w as a 3 inch outside diam eter stainless steel tube. It
functioned a s a resonance breaker. It ensured th a t th e plasm a discharge
(6) stayed on top of the su b strate by breaking the cavity resonance and
the discharge excitation condition in the sm aller cavity und ern eath the
substrate. By ensuring th a t th e plasm a discharge (6) is only created
above th e substrate, the resonance breaker (29) improves th e efficiency
of the microwave in p u t power. It also improves the reliability and
repeatability of the experimental results and the area on a su b strate th a t
can be uniformly coated with diamond film. The m etal choke (30) sleeve
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
27
26
26
7 /fl
2 7 -------
26
43
23
LEGEND
(5) Quartz Dome
(6) Plasm a Discharge
(23) Gas Input Tunnel
(25) Air Cooling Tunnel
(26) Cooling Water Tunnel
(27) Laser Port
(43) Base Plate
Figure 3.15 The Base-plate of MCPR7-3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
provides a floating end of the cavity an d a choke of the microwave
radiation.
This design minimizes th e plasm a volume by creating a
hem isphere shaped plasm a adjacent to the substrate. The cylindrical
sym m etry of the tuning m echanism and system configuration ensures
th a t th e plasm a generated h as a n inherent cylindrical symmetry. This
reactor is m ounted on a vacuum cham ber with cham ber walls and a
cham ber outlet leading to a vacuum pum p.
3.6.2.3
The Forced Flow Configuration
Figure 3.16 is a schem atic drawing of a forced flow configuration.
This configuration is different from the basic configuration by the way
the gas flow passage is controlled. In the basic configuration, the gas flow
is n o t controlled, it flows n aturally inside the quartz dome (5). W hereas
in this forced flow configuration, a metal plate (31) an d a quartz tube (28)
are p u t together to force the gas to flow through a flow p attern regulator
(32). The flow pattern regulator (32) is a plate with a series of holes,
directing th e way th a t the gas flows through the plasm a and th e
sub strate. This configuration increases n o t only th e efficiency of the
source gas b u t also the uniform ity of th e coating by influencing the
shape of the plasm a discharge (6) through changing th e flow pattern into
the plasm a discharge (6).
Figure 3.17 (a) show s one example of flow regulator, where a
num b er of sm aller substrates are being coated. Variation of in p u t power,
p ressu re and flow rates along w ith the flow pattern regulator produces
uniform plasm a over th e sm aller su b strates and ensu res uniform
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
p u iuMg g u u .
LEGEND
;i Cavity Walls
2'
Sliding Short
Finger Stock
Cavity Bottom Surface
Quartz Dome
6 Plasma Discharge
’3: Substrate Holder
14 Substrate
19 Power Coupling Probe
20 Probe Sleeve
21 Cooling Water
22a) Cooling Air Inlet
#
22b) Window/Air Outlet
(23 Gas Input Tunnel
(24 Distribution Plate
(25 Air Cooling Tunnel
(26 Cooling Water Tunnel
(27 Laser Port
(28 Quartz Tube
(29 Metal Tube
(31 Metal Plote
(32! Flow Regulator
( 43 ; B ase Plate
(80 Optical Pyrometer
Figure 3.16 The Forced Flow Experim ental Configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3.17
Flow Pattern Regulators
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
coatings on them. Figure 3.17 (b) shows another example of flow pattern
regulator, where a large piece of su b strate is being coated. Variation of
in p u t power, pressure and flow rates along with th e flow p attern
regulator produces a uniform plasm a over the su b strate so th a t a
uniform diamond coating can be obtained.
3.6.2.4
The J e t Configuration
Figure 3.18 is a schem atic drawing of a je t configuration where the
substrate (14) and su b strate holder (13) are located in a region separated
from th e plasm a discharge (6). The m etal plate (31) and quartz tu b e (28)
are p u t together to force the gas flow through the je t grid (33) which is an
electrically conducting plate w ith a series of holes. The size of the holes,
the flow rate and the vacuum pum p’s pumping speed determ ine the
pressure difference between th e plasm a discharge (6) region and
cham ber region w herein the su b strate (14) is located. In th is
configuration, the reactive species are forced to flow at a higher speed
over selected areas so th a t th e deposition rates at the selected areas are
increased.
3.6.2.5
The Lower Cavity Resonance Configuration
Figure 3.19 shows a lower cavity resonance configuration where
the substrate (14) is located a t th e lower section of the cavity w ithout the
u se of metal resonance breaker (29) shown in Figure 3.14. This was the
first configuration used for diam ond film growth in the th ird generation
MCPR.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
legend
(11 Cavity Walls
(2 { Sliding Short
(3 (i Finger Stock
(4 ( Cavity Bottom Surface
(5 i Quartz Dome
(6 ) Plasma Discharge
(131 Substrate Holder
(14, Substrate
(1 9 ) Power Coupling Probe
(2 0 Probe Sleeve
(2 1 ) Cooling Water
(2 2 o ) Cooling Air Inlet
[22b) Window/Air Outlet
Gas Input Tunnel
Distribution Plate
Air Cooling Tunnel
Cooling Water Tunnel
Laser Port
Quartz Tube
Metal Tube
Metal Plate
Jet Grid
B ase Plote
Figure 3.18 Je t Experim ental Configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
LECEND
(1)
(21
(3 )
(41
(51
(6 )
Cavity Walls
Sliding Short
Finger Stock
Cavity Bottom Surfoce
Quartz Oome
Plasma Discharge
13 Substrate Holder
(K Substrate
[19j Power Coupling Probe
(20 Probe Sleeve
(21 Cooling Water
(2 2 o ) Cooling Air Inlet
i 22b) Window/Air Outlet
Gas Input Tunnel
Distribution Plate
Air Cooling Tunnel
Cooling Water Tunnel
Loser Port
Quartz Tube
Metal Choke
Metal Plate
B ase Plate
Figure 3.19 Lower Cavity Resonance Configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
3.6.2.6
The Down-stream Configuration
Figure 3.20 shows a down-stream configuration where the
su b strate (14) is located outside the cavity and plasm a discharge (6)
region.
The substrate can be biased with a voltage during the deposition.
The tem perature of the substrate can be adjusted by a heater or cooling
tunnels. Perm anent magnets can be placed in the base-plate (43) around
the plasm a discharge (6) region to enhance plasm a formation at low
deposition pressures.
3.6.3
Reactor Operation
The reactor operation of the third generation 7” MCPR (MCPR7-3)
is sim ilar to th a t of MCPR7-2. The m ain difference is th a t TM013 mode is
used in later p art of the experiments conducted in MCPR7-3 where the
cavity length, Ls, was set at approximately 21.6 cm and the probe length,
Lp, w as adjusted to about 3.2 cm.
3.6.4
Reactor Performance
3.6.4.1
Introduction
Preliminary versions of the five operating configurations have been
investigated for diamond film growth. A sum m ary of the reactor
perform ance in the preliminary versions of is given below. Detailed
description of reactor performance in the force flow configuration is given
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
A
22b— !
ISSENR
Cavity Walls
Sliding Short
Finger Stock
Cavity Bottom Surface
Quartz Dome
Plasm o Discharge
Substrate Holder
Substrate
Power Coupling Probe
Probe Sleeve
Cooling Woter
’2 2 a ) Cooling Air Inlet
|22b) Window/Air Outlet
2 3 ) Gos input Tunnel
,24, i Distribution Plate
,25, Air Cooling Tunnel
26ji Cooling Water Tunnel
(271 Laser Port
2 8 ) Quartz Tube
31 Metal Plate
.34, i Heater
,43) B ase Plate
Figure 3.20 Down-stream Experim ental Configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
in chapter 6 .
3.6.4.2
The Basic Configuration
Diamond film w as deposited on 4" silicon w afer un d er the following
experimental conditions:
H 2 gas flow = 300 seem
CH 4 gas flow = 3 seem
Pressure = 20 Torr
MW power absorbed = 2 kW
Seeding: scratched w ith 1 pm diam ond powder.
The resu ltan t diamond film deposition rate w as slow (~ 3 m g/hr),
possibly due to low operating pressure and power. Both graphite and
am orphous carbon su b strate holders were used in diamond film
deposition processes in th is configuration. It w as found th a t higher film
growth rates were obtained w ith the am orphous carbon substrate
holders, possibly due to easier carbon dissolution from the amorphous
carbon holder into the reactive gas mixture. The am orphous carbon
holders were distorted in the high tem perature deposition processes.
3.6.4.3
The Forced Flow Configuration
Using the forced flow configuration, a 12 cm diam eter discharge
was created with 200 seem hydrogen and 2 seem m ethane gas flow at 40
Torr w ith 4 kW 2.45 GHz absorbed power. The discharge area produced
is larger th a n any discharge area reported in the open literature. The
discharge created h ad a symmetric, hem isphere shape over the substrate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
surface. Diamond films have been deposited uniformly on 4 inch
diam eter silicon wafers. The uniform coated area is larger th a n any
coated area reported in the open literature. A detailed description of the
reactor performance in this configuration is provided in chapter 6 .
3.6.4.4
The J e t Configuration
A prelim inary version of th is configuration was u sed for diamond
film deposition. It is schem atically shown in Figure 3.21. In this
prelim inary version, gas was forced to flow through the center hole on
the graphite grid. The center hole is 0.5 inch in diam eter an d the
substrate was located approximately 1 cm below the graphite grid.
W ithout using any cooling m echanism , increased deposition rate (~ 1
pm /hr) over 1.5 cm diam eter area had been achieved. Typical
experimental conditions are as follows:
H2 flow rate = 400 seem
CH4 flow rate = 6 seem
CO2 flow rate = 2 seem
pressure = 95 Torr
absorbed microwave power = 1.67 kW
The deposition rate w as limited because of su b strate melting at
high pressure and tem perature conditions. It is expected th a t by adding
cooling tunnels undern eath the su b strate to control the su b strate
tem perature, higher p ressure and power can be used to further increase
the deposition rates.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
22b —
—4 2 5 S 2
29—
LEGEND
(1) Cavity Walls
(2) Sliding Short
(3) Finger Stock
(4) Cavity Bottom Surface
(5) Quartz Dome
(6) Plasma Discharge
(1 3 ) Substrate Holder
(1 4 ) Substrate
(1 9 ) Power Coupling Probe
(2 0 ) Probe Sleeve
(2 1 ) Cooling Water
(22a) Cooling Air Inlet
Figure 3.21
(22b) Window/Air Outlet
(2 3 ) Gas Input Tunnel
12 4 ) Distribution Plate
(2 5 ) Air Cooling Tunnel
(2 6 ) Cooling Water Tunnel
(2 7 ) Laser Port
(2 8 ) Quartz Tube
(2 9 ) Metal Tube
(3 1 ) Metol Plate
(3 3 ) Jet Grid
(4 3 ) Base Plate
Preliminary J e t Experimental Configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
3.6.4.5
The Lower Cavity Resonance Configuration
Diamond film w as deposited on a 3 inch silicon wafer u nder the
following experimental conditions:
H2 gas flow = 200 seem
CH4 gas flow = 2 seem
Pressure = 30 Torr
MW power absorbed = 2 kW
The resultant diamond film was fairly uniform. The deposition rate
was ~ 3.56 m g/hr. It w as found th a t a plasm a discharge (6) generated in
this configuration h ad to compete with a discharge generation process
u ndern eath the su b strate holder (13). The com peting process made the
starting procedure complicated and unreliable.
3.6.4.6
The Down-stream Configuration
A preliminary experim ent with this configuration was conducted in
w hich a 3 inch su b strate h eater (34)<60> w as used to control the
su b stra te tem perature.
A schematic drawing of the heater is shown in Figure 3.22.<60> As
shown, it is a resistan t h eater an d consists of three layers, a top boronnitride (BN) (46), a pyrolytic graphite (47) and a lower boron-nitride (BN)
(48) layer where the pyrolytic graphite layer is the middle layer. The
exposed pyrolytic graphite areas (49) a t the two ends are electrical
contacts. This heater w as specified to m eet the following heating
requirem ents, namely, it (a) h a s a 3" heating zone, (b) can achieve
tem peratures up to 1500 °C, (c) can achieve tem perature uniformity of ±
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
34
46
S
*
♦
47
48
LEGEND
34 ) Heater
46) Top BN layer
£47) Pyrolytic Graphite
(48) Lower BN layer
(4 9; Contacts
Figure 3.22 Schem atic Drawing of a 3” S ubstrate Heater
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
10 °C over the heating zone, (d) operates w ith 60 Hz, 0 - 200 Volts AC
power, (e) draws a m axim um current of 15 A an d (f) can be used at
pressures from 1 mTorr to 100 Torr in H2/C H 4 gas m ixture environment.
The experimental conditions were:
H2 gas flow = 400 seem
CH4 gas flow = 4 seem
Pressure = 15 Torr
MW power absorbed = 1 . 5 kW
H eater voltage = 100 Volts, 60 Hz
The resu ltan t film growth rate w as low an d the film uniformity was
poor.
3.7
Comparison of Seven Inch MCPRs
This third generation MCPR offer advantages over the reactors
described in chapter two and the first two generations of MCPRs in th a t
it creates a large symmetric, disk to hem isphere shaped discharge for
uniform diamond film deposition over large surface areas. The symmetry
of the plasm a discharge is ensured by the symmetric, end feed power
in p u t configuration. The power input system , excitation probe and probe
sleeve, have large dim ensions so th at high microwave power (2 to 5
kilowatts) can be delivered into the cavity. High in p u t power, low near
field effects, a large quartz disk and the design configuration of keeping
the vacuum seal away from the heated areas ensure th a t a large
diam eter plasm a can be created. The equator plane of the hemispherical
shaped discharge is created and defined by th e cavity bottom surface
w ith the help of a metallic step near the outside surface of the quartz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
disk an d the lower cavity resonance breaker. The quartz disk ensures
only th e plasm a adjacent to th e su b strate is produced an d th e efficiency
of th e absorbed power is maximized. The su b strate is located n ear the
equator of th e hemisphere, hence the spacial uniformity of the
distribution of reactive species an d th e thickness of deposited diamond
film is ensured.
This reactor employs some of the basic features of th e conventional
microwave-cavity plasm a reactors (MCPR). They are: (1) cylindrical
cavity, (2) internal matching, i.e., sliding short and variable probe; (3)
excitation of TMqih modes for deposition of films. However, there are
some im portant differences between the MCPR7-3 and the conventional
microwave reactors. They are: (1) th e probe is m ounted on the sliding
short, n o t on the cavity side walls, th u s all the internal cavity
adjustm ents take place on the sliding short; (2) th e probe is located a t
th e center of the short and th u s only TM modes are easily excited; (3) the
probe is located far away from th e discharge region so th a t the probe
n ea r field does not interfere w ith the TMoin mode electromagnetic fields
w ithin the discharge region; (4) larger diam eter quartz disk can be used
and (5) m uch higher microwave power can be delivered to th e reactor to
create a larger diameter plasm a; (6) w ith the help of the lower cavity
resonance breaker, the cavity bottom surface and th e metallic step
defines the equator of the symmetric, disk-like or hem isphere shaped
plasm a w hen the cavity is excited in TMqih modes; (7) the su b strate
holder is adjustable in th a t it can be moved up and down independently,
this feature together w ith the independent sliding sh o rt and excitation
probe movement allow the movement of th e relative position of the
su b strate w ith respect to the plasm a so th a t the optim um deposition
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
conditions can be reached; (8) th e forced reactive gas flow through the
plasm a n ear the su b strate improves the efficiency of the reactive gases,
and the uniformity of the deposited film; (9) film deposition down stream
from th e plasm a; ( 10) independent and selected a rea substrate heating or
cooling increases the deposition areas and film uniformity.
3.8
Fourteen an d Eighteen Inch MCPRs
3.8.1
Introduction
The feasibility of diam ond film growth over larger surface areas
using larger diameter cavity MCPRs was investigated using 14" and 18"
side feed MCPRs with 2.45 GHz and 915 MHz power. The te st results are
described in sections 3.8.2 and 3.8.3.
3.8.2
Fourteen Inch MCPR
The feasibility of diamond film growth using a side power feed 14”
MCPR<61> and 2.45 GHz power was tested using th e Chuang 1.2 kW
2.45 GHz power source. The schem atic drawing of the experimental set­
u p s are shown in Figure 3.23 for su b strate supported by a quartz tube
and Figure 3.24 for su b strate supported by a m etal tube. As shown,
these schem atic drawings are sim ilar to th a t show n in Figure 3.5. The
m ain differences are th a t (a) the cavity inside diam eter is now 14", (b) the
vacuum seal is placed away from the discharge zone and (c) the quartz
dome is 10” in diam eter an d 3.5" in height.
It w as possible to couple microwave power into th e cavity and a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
Sliding Short
y/ / / / / / s/ y/ s/ s/ / / y/ y/ / / s/ / / y/ ss/ s/ / / ss/ / / / / / / s/ / / / / / s/ / / / / s/ s/ s/ / / / / s/ / / / / / / / / / sss/ s / / s/ / / / / / / .
14.0 Microwave Cavity
Quartz Dome
Probe
Plasma
9999994
Substrate
& Holder
Quartz Tube
Perforated
Metal Screen
\ /
To Vacuum Pump
Figure 3.23 S u b strate Supported by a Q uartz Tube in 14" Cavity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
Sliding Short
♦
7//syjw?ss/s>vrssy/ssyy/Vf/y?///y-/ys//y/sss//rs/j^^^
s!
14.0" Microwave Cavity
Probe
Quartz Dome
< ^s/sss* /s//S /M V s3
N .
Plasma
\
Substrate
& holder
■
Metal
Support
^
\
/
Perforated
Metal Screen
To Vacuum Pump
Figure 3.24
S u b strate Supported by a M etal T ube in 14" Cavity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
plasm a disk can be produced. The plasm a disk was approximately 3 inch
in diam eter at 20 Torr w ith 700 W atts power input. The plasm a disk was
suspended above the su b strate surface.
From the resonant cavity lengths for the TM modes displayed in
Figure 3.25 and Table 3.1, it is expected th a t TM0 i 2 mode can be excited
in a 14 inch cavity w ithout m uch interference from th e other TM modes.
With an end feed and a properly designed base plate, it is expected th a t
discharge loaded TM012 mode can be excited in a 14” MCPR. Diamond
film m ay be deposited in th is reactor w ith a su b strate moving stage and
sufficient microwave power ( 3 - 1 5 kW).
3.8.3
Eighteen Inch MCPR
The concept of plasm a discharge generation over larger surface
area using 915 MHz power in a larger cavity w as proposed by J .
A smussen.
To te st this concept, a n 18” MCPR<62> with side feed coupling was
used w ith a 500 W, 915 MHz power supply. The schem atic drawing of
the 18" MCPR is sim ilar to th a t shown in Figure 3.5. The cavity inside
diam eter is 18". The side feed probe sleeve has an inside diam eter of 1 5 /
8 " and the coupling probe h a s a n outside diam eter of 5/8".
After a series of experiments, it w as found th a t th e plasm a
discharge loaded TMq h mode can be easily excited. B ut the available
power level was too low to be u sed for actual diamond film growth over a
significant area.
The feasibility of diam ond film growth using the 18” MCPR and
2.45 GHz power was tested using the Chuang 2.45 GHz 1.2 kW power
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
n
o 8*
w
CM
CN
ON—
O<
CM
CM « s
OW
co
2 ««
-S'
Resonant Cavity Lengths of TM Modes in 14* Cavity
103
Figure 3.25
,o
CM
o
O
CO
o
CO
o
cs
o
(d'u)x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
Table 3.1
R eso n an t Cavity Lengths of TM M odes in 14" Cavity
X(n,p)
Cavity Mode
Cavity Length (cm)
2 .405000
(TM011)
6.346941
2.405000
(TM012)
12.69388
2.405000
CTM013)
19.04082
3.832000
CTMlll)
6.746388
3.832000
(TM112)
13.49278
3.8 3 2 0 0 0
CIM113)
20.23916
5.136000
(TM211)
7.407754
5.136000
(TM212)
14.81551
5.136000
(TM213)
22.22326
5.520000
CTM021)
7.689620
5.520000
(TM022)
15.37924
5.520000
(TM023)
23.06886
6.380000
(TM311)
8.564938
6.380000
(TM312)
17.12988
6.380000
(TM313)
25.69481
7.016000
(TM121)
9.577787
7 .016000
(I'M 122)
19.15557
7.016000
(TM123)
28.73336
7,588000
(TM411)
11.02720
7.588000
CTM412)
22.05441
7.588000
(TM413)
33.08161
8.41 7 0 0 0
(TM221)
15.86860
8.41 7 0 0 0
(TM222)
31.73721
8.771000
(TM511)
22.24383
8.771000
(TM512)
44.48766
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
source. It was found th a t it was difficult to couple the 2.45 GHz power
into th e cavity efficiently to create a centered plasm a. This could be due
to the large num ber of degenerate and nearly degenerate modes present.
The instability of th e power source may have also contributed to this
inefficient power coupling problem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER FOUR
THE MICROWAVE-CAVTIY JET REACTORS (MCJR)
4.1
Introduction
Diamond film deposition w ith high growth rates is desirable in
m any applications. One m ethod to increase the growth rate is to increase
th e reactive species concentrations over the growth surface. The
microwave cavity je t reactor (MCJR) is one type of plasm a reactor th a t
m ay be able to deposit diamond films with high growth rates. In the
microwave cavity je t reactor, the excited species from the microwave
discharge are directed through a nozzle onto a substrate. Since the
microwave discharge can be m aintained a t high pressure (100 Torr to 1
atmosphere), the n u m b er of active species available for diamond film
growth is high. Thus higher growth rates are expected. One advantage
th a t th e MCJRs have over some other types of je t reactors, su ch a s DC
arc je t reactors, is th a t MCJRs do not have electrodes, th u s eliminating
th e possible contam ination from th e electrodes. The MCJR w as originally
developed a t MSU as microwave electrotherm al th ru ste rs<10,11>.
In this chapter, three microwave cavity je t reactors, th e first,
second an d third generations of microwave cavity je t reactors, MCJR-1,
MCJR-2 and MCJR-3, are described. The MCJR-1 and MCJR-2 have
been developed and investigated for diamond film growth. The MCJR-3
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
was designed b u t h as not been built due to lack of funding.
The experiments described in th is chapter are preliminary
experim ents in investigating the potential of microwave cavity je t reactors
in diam ond film deposition. Diamond films have been deposited w ith a
different type of microwave je t reactor<37,38> which was reviewed in
chapter 2. Linear growth rates in th e order of 12 jim /hour have been
reported w ith th a t reactor.<37,38>
4.2
4.2.1
Experimental Systems
Introduction
All th e microwave cavity je t reactors th a t are described in this
dissertation use microwave power as th eir energy source and vacuum
pum ping system is used for reactive gas supply an d operating pressure
control. The microwave power supply, gas flow and vacuum pum p
system u sed in the MCJRs are described below in Sections 4.2.2 and
4.2.3.
4.2.2
Microwave Power Delivery System
The microwave circuit used to deliver microwave power to the je t
reactors is similar to th a t described in Section 3.2.2. The only difference
was th e microwave power source th a t w as used. In the experiments
described in this chapter, a Thermex (Model No. 4074) 2.45 GHz, 2.5 kW
power source was used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
4.2.3
Flow Control and Vacuum Pum p System s
A schem atic drawing of the gas flow control and vacuum pum p
system is show n in Figure 4.1. As shown, th e reactive gases, high purity
(99.99% or better) H2 and CH4, are supplied by cylinder gas tan k s (1)
and (2). The gas flow rates are m easured by m ass flow meters (3) and
controlled by needle valves (4). The flow m eters m easure flow rate s with
full scale flow ranges of 10 seem (CH4) an d 500 seem (H2). The reactive
gases are combined and mixed and as they flow into the quartz tube (5)
which extends into the microwave cavity applicator (6). They flow
through th e cavity applicator (6) and the b ase plate (7) into the process
cham ber (8). A m echanical roughing pum p (10) p um ps the reactive gases
through a m anual valve (9) and o u t into th e ex h au st (11). The flow rate of
purge gas supplied by nitrogen gas cylinders ( 12) is m easured by a tube
flow m eter (13) an d controlled by needle valves (14). It is added to the
exhaust ( 11) to dilute the flammable reactive gases.
The p ressu re in the process cham ber region is m easured by a
Hastings therm al conductivity vacuum gauge (15) w ith meaningful
pressure m easurem ent range of 5 - 200 mTorr an d a MKS type 220B
10,000 Torr full scale baratron pressure gauge (16). The therm al
conductivity vacuum gauge (15) w as used to m onitor cham ber base
pressure an d th e MKS pressure gauge w as u sed to m easure the process
pressure. The cham ber pressure is controlled by the gas flow rates and
the m anual throttle valve.
The instrum entation feed through p o rt (17) on top of the process
cham ber is used for heater power supply an d therm al couple electrical
signal transm ission. With this instrum entation feed through, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
(17) Instrumentation
Feed
Through
Port
(18) Vacuum
Feed
Through
Port
(19) View
^ Port
(8) Process
Chamber
(9) Manual Valve
t
E
(21) Suspension
Tube
(16) Baratron
Pressure
Gauge
(10) Mechanical
Roughing
Pump
(11) Exhaust
(20)
Substrate
(7) Base
Plate
(15) Thermal
Conductivity
Vacuum Gauge
(14) Needle
Valve
(13) F lo w _ * j; 1
Meter
^
T
(6) Microwave
Cavity
Applicator
(5) Quartz Tube
(4)Needle
Valves
Figure 4.1
(3) Flow
Meters
Flow Control and Vacuum System for MCJRs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
tem perature of the substrate can be m easured and an independent
substrate h eater can be used to vary th e su b strate temperature.The
vacuum feed through p o rt (18) is used to suspend the substrate (20) and
change its position via th e suspension tu b e (21). The downward facing
substrate (20) is m ounted on the lower end of the suspension tube (21).
The upper end of the suspension tube (21) is a h alf inch stainless steel
tube which can be sealed by a half inch u ltra-to rr compressed o-ring
vacuum fitting (18) welded to the center of th e process cham ber (8) top
surface. When cooling liquid or gas flows inside a hollow suspension tube
(21), the su b strate (20) can be cooled. P art of th e process cham ber can be
viewed through the view port (19).
4.3
The First Generation MCJR
4.3.1
Introduction
The first generation MCJR (MCJR-1) w as the first je t reactor
investigated for diamond film growth. It w as developed in 1988 and was
built from a n existing microwave cavity an d an existing stainless steel
chamber.
4.3.2
Reactor Geometry
A schem atic drawing of this reactor is shown in Figure 4.2. As
shown. This reactor h as a cylindrical side-wall (1) which forms the outer
conducting shell of the cavity applicator. A water-cooled (21) sliding short
(2) forms the lower end of the cavity applicator. The sliding short (2) can
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I ll
First Generation MCJR (M CJR-1)
28
t r
26— :
10— (
LEGEND
Cavity Walls
Sliding Short
Finger Stock
Cavity Top Surface
Quartz Tube
Plasm a Discharge
Gas Input
Outer Conductor
Power Input Probe
Substrate
Substrate Holder
(13)
(14)
(21)
(22)
(25)
(26)
(27)
(28)
(29)
(30)
Linear Feed Through
O -rin g Seal
Cooling Water
B rass Top Plate
Cooling Water
Window/Air Outlet
Cooing Air Inlet
P ro c ess Cham ber
Ultra—Torr Fitting
B rass Collar
Figure 4.2 Schematic Drawing of MCJR-1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
be moved along the cylindrical axis of the cavity wall (1) to change the
position of the lower end of the cavity applicator. A w ater cooled (25)
b rass top plate (22) w ith the cavity top surface (4) form th e top end of the
cavity applicator. The b rass top plate (22) and th e cavity wall (1) are
soldered together. Microwave energy is coupled into th e applicator
through the power in p u t probe (10) and its coaxial outer conductor (9).
Reactive gas m ixture enters the cavity applicator from the gas
in p u t p ort (7) where it is contained inside th e 1" outside diam eter quartz
tube (5). The gas m ixture flows upward, it is heated, ionized and
dissociated inside th e cavity applicator. The microwave discharge then
flows over and around the su b strate (11) and su b stra te holder (12) into
the stainless steel process cham ber (28) after w hich it is pum ped o ut by
a vacuum pump.
The “ultra-torr" vacuum fitting (29) n ear th e bottom of the stainless
steel process cham ber (28) serves as a vacuum seal p o rt to the quartz
tube through the com pressed o-ring seal (14). The quartz tube inner
space n ear the “ultra-torr” vacuum fitting (29) functions as the nozzle for
this je t reactor. Two inch diam eter openings exist a t th e center of the
sliding sh o rt (2) and th e b rass top plate (22) so th a t quartz tu b es of less
th a n 2 inch in outside diam eter can extend through th e center of the
microwave applicator. B rass collars (30) are u sed to reduce the size of
the opening to fit sm aller quartz tubes and to reduce th e microwave
radiation leakage from the cavity applicator. The “ultra-to rr” vacuum
fitting (29) on the stainless steel cham ber is an 1 inch ultra-torr fitting,
hence quartz tubes (5) of 1 inch in outside diam eter an d b rass collars
(30) of 1 inch in inside diam eter were used w hen experimenting w ith this
reactor concept.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
Besides being cooled by th e water cooling tubing (21) on the sliding
short (2) and the w ater cooling tubing (25) on the b rass top plate (22),
this reactor is also cooled by cooling air. The cooling air entered the
cavity applicator through th e cooling air inlet (27) an d exited through the
screened w indow /air outlet (26) which also served as a viewing port.
4.3.3
Reactor Operation
The experimental operation of this MCJR h a s been described in
detail else where <10*11*63,64> Thus> onjy a brief description is given
here. Differences from earlier work <10JL63.64>
operation a t lower
pressure (30 - 80 Torr) and lower flow rates (100 - 400 seem). The
ignition, matching, and internal cavity tuning are simila r to th a t reported
earlier for microwave electrotherm al thrusters <n>.
The mode diagram of a n ideal 7” cavity is shown in Figure 3.6. This
MCJR can create a microwave discharge w hen excited in a single cavity
electromagnetic mode. A discharge was started by reducing th e H2/CH 4
gas pressure to 5 -10 Torr, and applying microwave power which would
th en ignite a discharge th a t filled the quartz tube. The discharge
pressure was th en increased to an operating condition of 30 - 80 Torr
while adjusting the length, Lg, and the probe, Lp, tu ning the cavity to a
m atched condition.
When the sliding sh o rt length Lg is adjusted to about 7.45 cm and
the probe length Lp to 8.1 mm, the discharge loaded TM qh mode is
excited. A schematic drawing of the field p attern of th e TMqi 1 mode in a
discharge loaded cavity is shown in Figure 3.7. As show n in Figure 4.2,
with the help of the b rass collar (30), the su b strate (11) and substrate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
holder (12), only the plasm a n ear the top of the cavity applicator is
excited. W ith th is excitation configuration, the plasm a is brought d o se to
the su b strate so th a t the reactive sp ed es can effidently flow over the
substrate. Deposition experiments were performed w ith this excitation
configuration.
Comparing this microwave plasm a reactor w ith the other
microwave cavity plasm a reactors (MCPR) described in chapter 3, the
stru ctu re of th is reactor h as interesting features. Referring to Figure 4.2,
the MCJR is m ounted upside down com pared w ith MCPR. This idea was
brought u p by J . A sm ussen in 1986. In th is MCJR, the reactive gases
flow upw ard an d are heated and dissodated by absorbing microwave
energy in the cavity applicator. The reactive sp ed es generated in a h o t
discharge (6) naturally flows upw ard tow ards the substrate (11). This is
due to th e fact th a t under the sam e p ressu re and for the sam e gas,
higher tem perature gas h a s a lower density and ten d s to flow naturally
upw ard in a cooler, higher density gas environment. Therefore, reactive
sp ed es are effectively utilized for diamond film growth. Another
interesting feature about this configuration is th a t th e input reactive
gases are draw n toward the hot discharge (6) w hen the reactive sp ed es
generated in th e hot discharge (6) flows up and around the su b strate ( 11)
toward the pum p. Hence, the in p u t reactive gases are effidently utilized
for diam ond film growth in this m ounting and gas flow configuration.
4.3.4
Reactor Performance
The first generation MCJR h as been tested in diamond film growth
experim ents from September, 1988 to February, 1989 in more th a n 17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
experim ents. W hen th e substrates (11) w ere lowered until they were in
contact w ith the discharge (6). diamond films were grown on the
substrates. The su b strates used were typically 1 cm x 1 cm silicon wafers
and silicon nitride.
A typical experiment (DTF04) w as conducted under the following
experim ental conditions:
CH4 flow rate = 1 seem
H2 flow rate = 120 seem
pressure = 50 Torr
cavity short length Lg = 7.45 cm
probe length Lp = 8.1 m m
absorbed microwave power = 300 W atts
There are a few limitations in the first generation MCJR. First, it is
not a “true" je t reactor since the “nozzle” diam eter is the inside diam eter
of th e q u artz tu b e (5), which is close to 1 inch. Second, because of the
long “neck” in the ultra-torr fitting (29), th e discharge (6) is usually
located far below the open area of the stainless steel cham ber and the
su b strate ( 11) is usually lowered into th e “neck” area for diamond film
deposition. This configuration p u ts a lim it on the size of the su b strate to
be sm aller th a n the inner area of the q u artz tube, which in the present
reactor is less th a n 1” in diameter. Third, there are severe lim itations on
the in p u t power level and operating p ressu re in th e operation of this
reactor. Since the discharge is in close contact w ith the quartz walls
which h a s a melting tem perature in th e sam e range as the tem perature
of reactive gases in a discharge (6) suitable for diamond film growth, the
deposition p ressure w as limited to less th a n 100 Torr and the power
input w as limited to less th an 500 W atts so th a t th e quartz tube walls
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
would n ot melt. And finally, the experim ental reproducibility in this
reactor w as poor because m aterials were deposited on the quartz tube
walls.
4.4
4.4.1
The Second Generation MCJR
Introduction
To reduce th e nozzle size and scale u p th e su b strate area, the
second generation je t reactor (MCJR-2) w as developed in 1989. New top
end plate of th e cavity applicator and lower end plate of the stainless
cham ber were designed and built. Diamond film deposition experiments
were conducted w ith this reactor.
4.4.2
Reactor Geometry
A schem atic drawing of this reactor is show n in Figure 4.3. As
shown, the new end plate (23) of the cavity applicator is machined from a
b rass plate. It h as a n 1.1 inch diameter opening a t the center. A w ater
cooling tu n n el (16) is incorporated in the plate w hich is located close to
the discharge (6) zone so th a t efficient cooling ca n be achieved. The Oring spacer (18) functions primarily as a shock resist mechanical support
to the quartz plate (32). A new cavity shell w hich does n ot have an end
plate soldered onto it is u sed as the cavity w alls (1). Finger stock (8) is
used to electrically connect the brass end plate (23) to the cavity walls
( 1).
The lower end of the stainless steel cham ber w hich w as used in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
Second Generation MCJR (M CJR-2)
f
26— .
1— 23
II
10— <
LEGEND
'1) Cavity Walls
(16)
(2 j Sliding Short
(17)
[3) Finger Stock
(18)
(41 Cavity Top Surface
M9*
(5) Quartz Tube
(21
(6) Plasm a Discharge
(23)
(7) Gas Input
(26)
(81 Finger Stock
(27)
(9) Outer C onductor
(28)
[10) Power Input Probe
(31
11) Substrate
(32)
^12) Substrate Holder
(33)
13) Linear Feed Through
(34)
l15) O -ring Seal____________ (35 j
Figure 4.3
Cooling Water
Cooling Water
O—ring S pacer
J e t Nozzle
Cooling Water
B rass End Plate
Window/Air Outlet
Cooling Air Inlet
P ro cess Cham ber
S.S. End Plate
Quartz Flange
U ltra-T orr Fitting
Outer Quartz Tube
Cooling Gas Inlet
Schematic Drawing of MCJR-2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
MCJR-1 w as cu t off an d a Varian 8 inch diam eter stainless steel Conflat
flange w ith 6 inch opening n ear the center w as welded onto the stainless
steel cham ber. This welded plate serves a s a vacuum bolt joint between
th e stainless steel cham ber (28) and the stainless steel end plate (31).
The stainless steel end plate (31) is m ade from a V arian 8 inch stainless
steel Conflat blank flange. There is a 3.8 inch opening n ear the center so
th a t su b strates up to 3.8 inch in diam eter can be u sed in the stainless
steel cham ber. A w ater cooling tunnel (17) is incorporated in the plate for
the cham ber cooling. The O-ring seal (15) on th is plate serves as the
vacuum seal between the stainless steel end plate (31) and the quartz
plate (32).
The je t assem bly is m ade of an one inch q uartz tube (5) welded to
the center of a quartz plate (32), which is 5.5 inches in diameter and
0.25 inch thick. A 1 /3 2 inch diameter hole w as drilled a t the center of
the quartz plate (32). It serves as the je t nozzle (19). The one inch quartz
tube (5) passes through the cavity applicator while the quartz plate (32)
is sandwiched between the b rass end plate (23) an d stainless steel end
plate (31). Ultra-torr fitting (33) seals the lower end of the quartz tube
and m akes the gas flow transition from stainless steel tubing to the
quartz tu b e (5). On th e outside of the 1 inch quartz tube (5), another
quartz tube (34) of 1.5 inch in inside diam eter is u sed to transport the
safety and cooling N2 gas to the discharge (6) region. This outer quartz
tube (34) is sealed to th e inner quartz tube (5) n e a r the lower end of the
sliding sh o rt (2) and left open n ear the discharge region (6). N2 gas enters
this outer quartz tube (34) from the gas inlet (35) an d exits into the cavity
applicator n ear the discharge zone (6). This double tu b e cooling
m echanism is efficient since it forces the cooling gas to flow close to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
hot discharge zone (6).
4.4.3
Reactor Performance
This second generation MCJR h a s been investigated for diamond
film growth from November, 1989 to Ja n u ary , 1990. More th a n 17
experim ents were performed with th is reactor. Diamond-like films have
been grown on silicon substrates. The operation of this reactor is sim ilar
to th a t of th e MCJR-1.
The R am an spectrum of a typical film (JDF-10) deposited u nder
the following experimental conditions is show n in Figure 4.4:
CH4 flow rate = 3.5 seem
H2 flow rate = 504 seem
Ar flow rate - 252 seem
pressure = 84 Torr
cavity short length Lg = 7.12 cm
probe length Lp = 3 cm
absorbed microwave power = 350 W
The discharge generated in th is reactor w as alm ost in direct
contact w ith the quartz tube walls. The p ressu re w as limited to 100 Torr
and the power in p u t w as limited to 500 W atts so th a t the quartz tu b e
walls would n o t melt. It w as not possible to generate a high tem perature
discharge to p ass through the nozzle (19) to achieve the su b strate (11)
tem perature required for diamond film growth. The reproducibility was
poor because m aterials were deposited on th e quartz tube walls.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 4 0 .0 0
8 3 0 .0 0
m
CH4 = 3.5 seem
H2 = 504 seem
At = 252 seem
p = 84 Torr
Ls = 7.12 cm
Lp = 3 cm
P( = 350 W
W 7 2 0 .0 0
r—
2
6 1 0 .0 0
5 0 8 .0 7
|— '— i— i— i— |— i
42 0 .0 0
6 9 0.00
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 5 0 .0
|
i— i— •— |— i— i— i— i— |— i— i— i— i— |- i— i— i— i—
9 70 .0 0
1250.0
1530.0
RECIPROCAL CM
Figure 4.4 Raman Spectrum of a Film Deposited in MCJR-2
1810.0
121
4.5
4.5.1
The Third Generation MCJR
Introduction
In order to keep th e quartz walls away from th e hot plasm a
discharge zone, the third generation MCJR (MCJR-3) was designed. With
this design, th e quartz wall erosion and reproducibility problems faced by
the previous two generations of MCJR may be solved. Much higher power
in p u t and higher operating p ressure experiments m ay be conducted so
th a t higher diamond film growth rates may be achieved.
4.5.2
Reactor Geometry
A schem atic drawing of th is reactor is show n in Figure 4.5. As
shown, new b rass end plate (24) of the cavity applicator and new end
plate (31) of th e stainless steel cham ber are designed.
The b rass end plate (24) h a s a 4.62 inch opening near the center.
W ater cooling tu n n els (16) are incorporated in th is plate. O-ring spacer
(18) is used to support th e quartz flange (32). Finger stocks (8) are used
to electrically connect the b rass end plate (24) to th e cavity walls (1).
The stainless steel end plate (31) is designed to be machined from
a V arian 8 inch Conflat blank flange. It has a 0.5 in ch opening n ear the
center. A series of stainless steel nozzle inserts w ith different nozzle sizes
are designed to be fit into the 0.5 inch opening. An enlarged drawing of
the nozzle insert is show n in Figure 4.6. W ater cooling (17) is
incorporated in this plate and O-ring seal (15) is u se d to seal the quartz
flange (32) to th e stainless steel end plate (31).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
Third Generotion MCJR (M CJR-3)
HT
12—*
|— 13
28
— 24
tt
10 — <
LEGEND
1) Cavity Walls
(2 1 Sliding Short
(3) Rnger Stock
(4) Cavity Top Surface
(5) Quartz Tube
16 1 Plasm a Discharge
[7) Gas Input
f8 ) Rnger Stock
(9) Outer Conductor
10) Power Input Probe
.11) S ubstrate
^ 2 ) S ubstrate Holder
l13) Linear Feed Through
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(241
(261
(27)
(28)
(311
(32)
O -ring Seal
Cooling Water
Cooling Water
O -ring Spacer
J e t Nozzle
Quartz Dome
Cooling Water
Brass End Plate
Window/Air Outlet
Cooling Air Inlet
P rocess Chamber
S.S. End Plate
Q uartz Flange
Figure 4.5 Schem atic Drawing of MCJR-3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
Q.
O
I—
C
o
0
-*->
0
(0
1
(0
in
o
o
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Enlarged Drawing of the Nozzle Insert
>
Figure 4.6
•
£
<u
124
A quartz assem bly is used to tran sp o rt and confine the reactive gas
mixture. The quartz assem bly is made of a one inch quartz tube (5)
welded to a hem ispherical quartz dome (20) w hich h as an inside
diam eter of 4.25 inches and an outside diam eter of 4.5 inches. This
hem ispherical quartz dome (20) is welded n ear th e equator to a quartz
flange (32) of 6 inch in diam eter and 0.25 inch thick. The one inch quartz
tube (5) passes through the sliding short (2) of th e cavity applicator and
the quartz dome is housed inside the cavity applicator. The quartz flange
(32) is sandwiched between the b rass (24) and stainless steel (31) end
plates.
4.5.3
Remarks
This third generation MCJR has been designed b u t h a s not been
built. It is expected th a t the discharge an d quartz wall direct contact
problem an d the poor reproducibility problem encountered in first and
second generation MCJR will be solved w ith this je t reactor. Higher
operating pressure an d higher in p u t power which are desirable for
diam ond film growth a t higher growth rate s are expected to be achievable
w ith this je t reactor.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER FIVE
MICROWAVE ELECTRIC FIELDS IN THE MCPR
5.1
Introduction
In order to develop microwave reactors th a t can deposit diam ond
films over large surface areas an d a t high growth rates, it is im portant to
improve the knowledge of the fundam ental plasm a/chem ical reactions of
film deposition and to develop a n un d erstan d in g of the electromagnetic
field /p la sm a interactions.
W hen an ideal cylindrical cavity of radius a is excited w ith a
reso n an t mode (TEnpq or TMnpq modes) a t frequency f0, the cavity
lengths Lf are determ ined by the following equations,<65>
(Lr)™ = qica(4ic2a2f^ e n - x ^p)-I/2
(eq. 5.1)
for TMnpq modes, and
—
1/2
(Lr)™q = q n a ( 47t2a 2f ^ e n - x ’2p)
(eq. 5.2)
for TEnpq m o d es, where n , p and q are non-negative integers, x^p are
zeros of th e Bessel function, J n(x), an d x ’np are zeros of the derivative of
th e Bessel function, J 'n(x).
125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
Each of these resonant inodes, TEnpq or TM npq, h as well defined
electric an d magnetic field patterns which are expressed as the analytical
solutions to the Maxwell's equations with perfectly conducting cylindrical
walls as the boundaries.
When the cavity is excited with a plasm a discharge, the electric
and magnetic field p attern s of these resonant m odes are disturbed. The
determ ination of electromagnetic field p attern s in a general plasm a
discharge loaded cylindrical cavity from analytically solving the Maxwell’s
equations h a s not been achieved yet. This is due to th e fact th a t plasm a
discharge can be created a t various conditions, su c h as from sub-mTorr
to over 1 atm osphere pressures, with various geometries an d various
electromagnetic properties. The electromagnetic properties of these
plasm a discharges a s of yet have not been well established. Also, it is still
a challenge to obtain solutions to Maxwell’s equations in a cavity loaded
w ith electromagnetically lossy material, su c h as th e plasm a discharge.
This chapter describes the techniques and resu lts of the
m easurem ent of electric field strengths along the cavity inside walls in a
plasm a discharge loaded cylindrical cavity (MCPR7-1).<10,13,14,70> The
electromagnetic (microwave) /p lasm a conditions required to grow
diamond th in film are discussed. Specifically, th e (relative) spatial
variations of the exciting electromagnetic field p a tte rn s are m easured.
These spatial electric field m easurem ents an d intensities are th en related
to other im portant experimental param eters such a s gas mixture, flow
rate, in p u t power, su b strate tem perature, and discharge pressure, etc.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
5.2
Experimental M easurem ent System
The electromagnetic field p attern m easurem ent system consists of
a vertical b a r shown in Figure 5.1 an d two horizontal cylindrical b ars
enclosing th e cavity as show n in th e cross sectional view of Figure 5.2.
They are soldered onto th e outside surface of the cavity side walls. The
top an d bottom horizontal cylindrical b a rs are located a t (A) 50 m m and
(B) 20 m m from the cavity bottom surface, respectively. A series of holes,
about 2.2 mm in diameter, are drilled into the vertical and th e horizontal
b ars an d completely through the inside surface of th e cavity’walls. The
spacing between the holes on th e vertical b ar is 10 m m an d th e holes on
th e horizontal circles have a n an gular spacing of 7.5°.
A semirigid micro-coaxial diagnostic probe is inserted into the
holes u n til the tip of the probe is flush w ith the inside surface of the
cavity side walls (see insert on Figure 5.2). Power readings, w hich are
proportional to the square of th e rm s electrical field strength norm al to
the inside cavity walls, are obtained by connecting a power m eter to the
other end of the micro-coaxial probe during the experim ents.<10>
The relationship between th e diagnostic probe power reading Pp
and th e actual rm s electrical field strength E, w here E is defined by
2 ^ ^
E = E • E, was determined by u sin g a 6 ” diam eter em pty cylindrical
b ra ss cavity.<56*10,70>
The calibration procedure w as performed as follows. The em pty
cylindrical b rass cavity w as excited w ith th e TM qh mode a n d several
2.45 GHz in p u t power levels were used. The corresponding probe power
readings were taken w ith a power m eter attached to the other end of the
probe whose tip was set flush w ith th e inside surface of 6 " cavity walls.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
36
Lh"
I
(A)
i JO H
r
(HD L
— 5
|
V
JJ.W
T ?IA 1 1
V JJ.V
SfjftYY
Y YTStY
r
1J
3 7 ^ t» _ i8
u j i j j j j L — ^7
•’/sY'S.'/S/
To Vacuum Pump
LEGEND
1)
f2)
l3 j
(4)
(5 1
(6J
(7)
(8)
[9}
,10)
Cavity Walls
Sliding Short
R nger Stock
Cavity bottom surface
Quartz Dome
Plasm a Discharge
Gas Input Tunnel
Gas Distribution Ring
Metal Screen
Cooling Water Tunnel
Figure 5.1
(11
(12
(13
(1 4
(15
(16
(17
(18
(36
Graphite Bose
Q uartz Tube
S u b stra te Holder
S u b strate
Top Window
Pyrom eter Sensor
Input Coupling Probe
Outer Conductor
Vertical Bar
(A) Sc (B) Horizontal Rings
Electric Field M easurem ent System
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C irc u m fe re n tia l
a n g le
M ic ro c o a x ia l
d ia g n o stic
P ro b e
0 "(3 6 0 * ) ---------------
<C>
(A)
Pow er
in p u t
sid e
1
\
\
/
Q u a rtz
dom e
WV^
S id e w in d o w
Figure 5.2 Horizontal Cross-sectional View
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
W hen th e probe w as inserted into th e probe hole a t z = 3 .7 cm above the
cavity bottom surface an d th e cavity length d is adjusted to 7.795 cm for
TMq i i mode excitation, the following table of diagnostic probe power
reading vs. in p u t pow er w as obtained,
Table 5.1 Diagnostic Probe Power Reading versus In p u t Power
Incident Power
Pi (W)
0.8526
0.5858
0.3608
0.1093
Reflected Power
Pr (W)
0.0229
0.0134
0.0046
0.0011
Probe Power
PD 0iW)
48
33
20.5
6.3
an d th e following table of absorbed power P j = Pj - Pr, probe power Pp
an d th e ir ratio Pd/P p is generated,
Table 5.2 Absorbed Power an d Probe Power Ratio
A bsorbed Power
Pd (W)
0.8297
0.5724
0.3560
0.1082
Probe Power
PpQiW)
48
33
20.5
6.3
Power Ratio
Pd/Pp
1.73 x 10 4
1.73 xO 4
1.74 x 10 4
1.72 x 10 4
w here we see th e power ratio stay s co n stan t w hen th e absorbed power is
changed, i.e., Pd/Pp = 1.73 x 104.
In order to establish th e relationship betw een the electric field
stren g th E n ea r the cavity inside walls an d the probe power reading Pp.
The relationship betw een the electric field stren g th E near th e cavity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
inside walls and th e absorbed power Pd. is first derived. Thereafter, the
relationship between E an d Pp can be established from Pd/Pp = 1.73 x
104.
Using the perturbation theory for lossy conducting surfaces and
assum ing th a t the power absorbed by th e em pty cavity is dissipated on
the cavity walls, the following power balance equation is established,<65>
Pd = SR<£|H|2d s
(eq. 5.3)
w hereas for b rass m aterial a t f0 = 2.45 GHz, 91 is given by,<65>
91 = 5.01 x 10~7JT0 = 0.0248
(eq. 5.4)
For TMq ii mode, H field h as m ainly th e <(>component, w hich is
given by,<65>
H* = C ( 5 i ) J i ( ^ r ) cos ( ^ )
(eq. 5.5)
where Xqj = 2.405, a = 3" = 7.62 cm and C is a constant.
Using equations 5.4 and 5.5, we get, from equation 5,3,
Pd = 91(j!|H|2d s =
m
C2 (^oi
a
27cajJ (x01) Jc o s 2 ( ^ ) dz + 2 j 2 jcrJj (-^“ •r) d r
(eq. 5.6)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
Since Xq, = 2.405. J^x^,) = 0.5191. d / a = 7.795/7.62 = 1.02, 91 =
0.0248, we have, from equation 5.6,
Pd = 0.367C2
(eq. 5.7)
The electric field E near th e cylindrical cavity inside walls h as
mainly the com ponent normal to the walls, i.e., E = IEj-1, where Er is
given by the following equation,<65>
Cxq.jc
nz
Erlr = a = Jl S J >(X<») S ln ( T )
which leads to.
l^ lr-a =
Sta2<^ )
( e q ’ 5 ' 8 )
where w = 2 tc x 2.45 x 109, z /d = 3 .7 /7 .7 9 5 = 0.43. When the other
param eters used to obtain equation 5.7 are also used, we obtain, from
equation 5.8,
lEi f |r _a = 2-34x 1q7c2
fcq- 5.9)
From equations 5.7 and 5.9, we obtain the following relationship
between th e electric field strength E near th e cylindrical cavity inside
walls, E = I Er l , an d the absorbed power, Pd,
|Er|2| r =a = 6.37 x 10 7Pd
(eq. 5. 10)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
Using the resu lt from table 5.2, i.e., Pd/Pp = 1*73 x 104, we obtain
the following relationship between the electric field strength E near the
cavity inside walls and the probe power reading, Pp,
E 2 = |E J2|
= 1*1 x 10 12Pp
(eq. 5.11)
where the u n it of E is volt/m and the u n it of Pp is W.
5.3
5.3.1
M easurem ent Results
Electromagnetic Field Patterns
The axial an d circumferential electric fields were m easured while
the MPDR w as operating under diamond thin film deposition conditions.
An example of a field p attern m easurem ent is displayed in Figure 5.3 for
the case of a discharge th a t is symmetrically centered over th e substrate.
The experimental conditions were: pressure = 65 Torr, H2 flow rate = 300
seem, CH4 flow rate = 1.5 seem, absorbed microwave power = 590 W and
substrate tem perature - 975 °C. As shown in Figure 5.3, the electric field
at a constant height is nearly independent of circumferential angle <>. The
slight variation in intensity versus $ can be attributed to the near field of
the exiting probe. Also shown in Figure 5.3, the axial field w as m easured
and is compared to an em pty waveguide TMqi standing wave (dashed
line).
These field m easurem ents clearly show th a t th e mode is <j>
independent, and indicate th a t the plasm a loaded TMqi 1 mode was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
2.5
2
Top horizontal circle
1.5
At******'*’
+ ♦■*■-►+
*»««
1
*
Bottom horizontal circle
§
0.5
0
£
0
1
CO
•a
£
£
13
1.6
§o
1.2
0
90
-
180
270
Circumferential Angle (degree)
£
0.8
0.4
Vertical direction
Height Along Cavity Wall (cm)
Figure 5.3 M easured Electric Field Strength D istribution
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
360
135
excited during deposition.
The shape of the electric field p attern s shown in Figure 5.3 are
typical of the electric field m easured for a <|>symmetric, axially centered
discharge operating u n d er a wide range of different experimental
conditions. As experimental conditions change, i.e., pressure from 30 70 Torr, gas flow and in p u t power, only a slight variation of Lg and Lp
(less th a n a few mm) are required to m aintain a well m atched and
centered discharge. However, the discharge can also be m aintained in a n
off-centered and non-sym m etrical position if Lg an d Lp are changed from
the optim al discharge centered conditions a n d /o r if th e substrate itself is
positioned off the center of the cavity axis. W hen th is occurs, the
electromagnetic field p attern s are altered. The vertical field intensities
retain the standing wave field pattern shown in Figure 5.3, however the
circumferential electric field displays a skewed p attern where the electric
field intensities are the highest n ear the coupling probe (cj) - 180°) and
decrease as <j>varies form 180° to 0 or 360 degrees. Film growth is
usually non-uniform an d non-symmetric u n d er th ese experimental
conditions.
Figure 5.4 sum m arizes the m easured, plasm a loaded TMq h mode
electric and magnetic field p attern s and th eir relationship to the
discharge and substrate. The discharge is formed a t the open end of the
cavity inside the quartz disk and its length can be as long as a quarter
wavelength ~3.6 cm. The m ost intense electric field com ponents are
tangential to the discharge boundary. As shown, th e discharge is
completely separated firom the quartz walls and is in direct contact with
the su b strate and holder. The m etal base-plate w alls adjacent to the
discharge an d the su b strate help stabilize th e arc. In this configuration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
TM 011 M ode
field
line
field
line
Figure 5.4 M easured Field P attern in MCPR
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
ions an d free radicals derived from hydrogen an d m ethane impinge
directly on the su b strate heating th e wafer by conduction an d convection
and provide the necessary species for the appropriate su b strate surface
chemistry.
To further illustrate th e relationship between electric field and
discharge boundary, a com parison of plasm a discharge excited in a
rectangular cavity and th a t in o u r cylindrical cavity is shown in Figure
5.5. Since the electric field h a s different distribution in th e two different
cavities, the plasm a discharge takes on a different shape. W hat the two
different configurations have in common is th a t th e intense electric field
adjacent to the discharge are tangential to the discharge boundary. This
relationship can serve u s as a guide in designing microwave cavities for
a specific applications.
5.3.2
Power Balance an d Discharge Power Density
The microwave power coupled into the p lasm a loaded applicator is
given by P<j = P i-P r- where Pj is th e incident power and Pr is the reflected
power. The power coupled into th e applicator Pd divides itself between
th e power absorbed in th e conducting applicator walls P^. and th e power
delivered to the discharge load Pa. The exact division of th e power Pd
between the walls and th e discharge load depends on th e relative losses
in the discharge versus the losses in the applicator walls.
The experimentally m easured electric field intensities show n in
Figure 5.3 are typical order of m agnitude electric field stren g th s inside
the cavity under diam ond film thin film deposition conditions. These
m easured electric field stren g th s can be used to estim ate the power loss
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o
o
"o
o
"v_
TD
C
O
c
2
Q)
C
2
<D a>
M
—
CO a ;
♦
>
D
O
O
D
o>
c
o
•u
o
o
a:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5.5
>
Discharge Boundaries and Electric Fields
138
139
on the cavity walls and th e cavity loaded Q u n d er diamond film
deposition conditions.
To estim ate th e power loss on the cavity walls, the m easured
electric field strength along the cavity inside walls is first expressed
empirically by the following equation, from Figure 5.3,
TtZ.
\Ej\\r _a = 1.45X 104sin(-g-)
(eq. 5.12)
From equations 5.8 and 5.12, we obtain th e following equation to
determ ine th e constant C,
a
coead
C = 1.45 x 104----------
r
x0iJcJi(x0i)
(eq. 5.13)
where for the MCPR used in the field p attern m easurem ents, a = 3.5" =
8.89xl0"2 m an d d = 7 .2 x l0 "2 m. Hence, th e co n stan t C is given by,
C = 3.22
(eq. 5.14)
From equations 5.6 and 5.14, th e power loss on the cavity walls is
given by
Pb = C2XojSRJ2 (x01)
+ 2nj = 3.54 (W)
(eq. 5.15)
Therefore, the d ata shown in Figure 5.3 yields an estimated
absorbed wall power of less th an 4 W indicating m ost of the 590 W of
absorbed power is coupled to the d isch arg e/su b strate holder. The
absorbed microwave power efficiency is greater th a n 99%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
The Q factor of th e plasm a discharge loaded cavity applicator is
estim ated by first calculating th e energy stored in th e cavity applicator.
Assuming th a t th e electromagnetic fields were only slightly altered
by the presence of th e discharge and th a t th e electromagnetic fields in
the discharge were n o t m uch different from those w hen the discharge
was not present, the electric field components, Ej-, E<j, and Ez can be
expressed by th e following equations,<65>
E - CX° 1,r
r joaead J i (“T r) s in ( T )
(eq. 5.16)
E^ = 0
(eq. 5.17)
. 7CZ.
C X01 _ ,x01
2J o (-^ " r ) c o s (-x
T )
jcoea'
(eq. 5.18)
Ez =
The energy stored in the cavity applicator is given by,<65>
W = 2We = £jlE |2dV = e J lE ^ d V + e J lE /d V =
= e(^ )
toe
£oi)
a
2n
J ? (^ ir)rd r +
^ L g l j j cos2 (^ z ) d z j j 2 (^ i r ) r d r =
2
C7t dx 2 J 2 (x l
( ad } + < ~a r )
2
(eq. 5.19)
where C is given by equation 5.14. Hence th e energy stored in the cavity
applicator is.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
W = 2.3 x 10-6 (J )
(eq. 5.20)
Since the absorbed microwave power by the plasm a discharge
loaded cavity applicator is
= 590 W, the loaded cavity Q factor is given
by><65>
o)W
Q = — = 60
d
(eq. 5.21)
Thus, the absorbed microwave power is very efficient, i.e., P(j ~ Pa.
A visual estim ation of th e discharge volume is 100 cm 3 a t th e low
pressure of 20 Torr. As p ressure increases to 70 Torr, the volume
decreases (for a constant power of 590 W) to 30 cm3. Thus, th e discharge
absorbed power density varies from approximately 6 W /cm 3 to 20 W /
cm3.
5.3.3
Electric Field Strength
Diagnostic probe power readings were taken when cham ber
pressure, hydrogen and m ethane flow rates, an d microwave power input
were varied. Each set of m easurem ents were taken for a <j>symmetric,
well-centered discharge. The resu lts are summ arized in Figures 5.6 and
5.7 in w hich diagnostic probe power is plotted vs. in p u t power, flow rate,
and pressure. The m easured diagnostic probe power is proportional to
the square of electric field, i.e., it is proportional to the energy stored in
the loaded cavity fields. The power absorbed, P^, is equal to th e total
power dissipated within the cavity. Thus, the ratio of the probe power to
the absorbed power is proportional to the discharge loaded cavity Q.
Figure 5.6 shows u n d er a constant pressure of 70 Torr an d three
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
7.5
w
n■
O
100/0.5 seem
200/1.0 seem H—
300/1.5 seem -a—
Pressure = 70 Torr
X,
u
i
65
1
l
o
s
U
5.5
400
450
500
550
600
Absorbed Microwave Power (W)
Figure 5.6 Coaxial Probe Power versus Microwave Power and Flow Rate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
7.5
£
ini
o
u
7
I
I 65
cu
0)
55 Torr
1
o
o
-i— 60 Torr
■o— 65 Torr
* - 70 Torr
*
H2 = 200 seem
CHa. = 1 seem
5.5
450
500
550
600
Absorbed Microwave Power (W)
Figure 5.7 Coaxial Probe Power versus Microwave Power and Pressure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
different flow rates, th e diagnostic probe power, an d hence the square of
th e cavity electric field strength, is approximately proportional to the
power in p u t and only slightly influenced by th e flow rates.
Figure 5.7 shows for a constant flow of 200 seem H2 and 1.0 seem
CH4, the probe power is approximately proportional to th e in p u t power
and is only slightly influenced by the variation in pressure.
These m easurem ents dem onstrate th a t the discharge loaded cavity
Q varies little (~ 60) over the different flow rates, pressu res and input
powers. The absorbed microwave power efficiency is high over the wide
variations of experim ental conditions shown in Figures 5.6 and 5.7.
Thus, the power absorbed by the applicator is essentially equal to the
power dissipated in th e p lasm a/su b strate holder.
5.3.4
S ubstrate Tem perature
S ubstrate tem perature plays im portant role in diamond film
growth process. It affects diamond film growth rate an d morphology, two
of the leading figures of m erit in diamond film growth process. Figures
5.8 and 5.9 show su b strate tem perature versus absorbed microwave
power, flow rates an d pressure.
Figure 5.8 shows u n d er a constant pressure of 60 Torr and three
different flow rates, th e su b strate tem perature increases a s absorbed
power is increased and is only slightly influenced by th e flow rates in our
reactor.
Figure 5.9 shows for a constant flow of 200 seem H2 and 1.0 seem
CH4, th e su b strate tem perature increases w hen the absorbed power and
pressure are increased.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
960
100/0.5 seem
200/1.0 seem
-fl— 300/1.5 seem
O
Pressure = 60 Torr
940
2
0)
a
920
900
450
500
550
600
Absorbed Microwave Power (W)
Figure 5.8 Substrate Tem perature vs. Microwave Power and Flow Rate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
1000
O
<u
§
960
aJ
0)
I
1
co
*3
CO
70 Torr
65 Torr - t —
60 Torr -B—
55 Torr -X—
H2 = 200 seem
CHa = 1 seem
920
880
450
500
550
600
650
Absorbed Microwave Power (W)
Figure 5.9 S ubstrate Tem perature vs. Microwave Power and Pressure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
5.4
D iscussion
These experiments indicate th a t u n d er the diam ond thin film
deposition conditions of 30 to 80 Torr the microwave discharge is in a
transition between a n am -bipolar diffusion controlled discharge and a
freely floating microwave arc.<56,63,66' 68> The discharge is excited by the
lossy plasm a loaded TMq h cylindrical cavity mode where, as shown in
Figure 5.4, the m ost intense electric field adjacent to the discharge is
tangent to the discharge boundary. This mode couples to the discharge
very efficiently and h a s been used to excite high p ressu re “arc like”
microwave discharges for other applications.<56,63,66“68> It also h a s the
sam e relationship between electromagnetic field p attern s and the
discharge boundaries th a t was present in the earliest microwave plasm a
diam ond thin film deposition experiments,<30> a s show n in Figure 5.5.
Since the microwave arc appears to be fundam ental to the
observed diamond film deposition, several of its features are briefly
discussed here to provide a better understanding of th e discharge. The
microwave arc, like lower-frequency and DC arcs, is a therm ally inhomogeneous discharge. It h as a hot central core, an d therm al gradients
exist between the discharge center and surrounding walls. Microwave
energy is readily coupled into th e electron gas in the h o t discharge center
because of its reduced gas density, and because n eu tral gas species are
also readily ionized an d excited in the hot discharge region.
Major energy losses form the discharge occur by heat conduction,
convection, and radiation. W hen large tem perature gradients are present,
h eat conduction losses become an im portant loss process. This loss
m echanism includes the contribution to the h eat conductivity by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
molecules, atoms, electrons, and ions, and also chem ical reactions such
as the transport of dissociation energy and ionization energy to the arc
hinges by free radicals and ions. Owing to the high-pressure
environment, electron, ion an d free radicals recom bine quickly outside
the h o t central core or on a solid surface in contact w ith the arc core and
th u s convert their dissociation, ionization, and excitation energy into
therm al energy. The resu lt is a discharge with radially varying gas
tem perature, ionization rate, and volume recom bination rate. Gas
tem perature, ionization, dissociation, etc., are h ighest in the center of the
discharge, while volume recom bination and de-excitation of the different
species increase away from th e discharge center as th e cooler, denser gas
regions near the walls are approached. The central discharge core gas
tem peratures vaiy w ith gas type and pressure an d can be in excess of
2000 °k <66-68> while tem peratures external to the discharge are
controlled by wall tem peratures and the tem peratures of the gas flowing
around the discharge. If gas tem peratures are 2000 °K or more in
hydrogen gas, then atomic hydrogen can be created by therm al
dissociation as well as free electron impact.<69>
5.5
Sum m ary
The following conclusions were reached from th is study, under
diamond film deposition conditions in MCPR7-1:
(1)
The excitation mode w as identified to be discharge loaded
TM011 mode.
(2)
Maximum electric field intensity in the cavity was ~ 150 V /
cm.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
(2)
Discharge loaded cavity quality factor w as - 60.
(3)
The tangential com ponent of th e E field is the m ain
discharge excitation field.
(4)
Average power density in th e discharge w as ~ 6 W /cm 3 at 20
Torr an d 20 W /cm 3 a t 70 Torr.
(5)
U nder a constant pressure, th e square of the cavity electric
field strength is approximately proportional to th e power input an d only
slightly influenced by th e variation in flow rates.
(6)
At a co n stan t gas flow, the square of the cavity electric field
strength is approximately proportional to th e power in p u t and only
slightly influenced by th e variation in pressure.
(7)
U nder a co n stan t p ressure, the su b strate tem perature
increases w ith the power inp u t (~ 0.4 °C/W) and is only slightly
influenced by the variation in flow rates.
(8)
At a co n stan t gas flow, the su b strate tem perature increases
with th e power in p u t (~ 0.4 °C/W) u n d er each working pressure. Also, at
the con stan t gas flow an d input power, the su b strate tem perature
increases (~ 3.5 °C/Torr) with increasing pressure.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER SIX
DIAMOND FILM DEPOSITION IN THE MCPR
6 .1
Introduction
In various practical applications, diamond films with various
thicknesses and morphologies are desired. In order to determine the
growth rates and morphologies of diamond films deposited under various
experimental conditions and hence characterize th e performance of the
third generation microwave cavity plasm a reactor for diamond film
growth, diamond films are grown on three and four inch silicon wafers
u nd er various experimental conditions. In this chapter, the diamond film
growth rates, growth efficiencies and morphologies are presented with
respect to variations in substrate tem perature, gas composition and flow
rate. Typical uniformity profile an d Raman spectrum are also presented.
The objective of depositing diamond films on 3” and 4” silicon
wafers is to achieve the highest possible growth rate with good uniformity
(judged by visual inspection).
The sections in this chapter are arranged as follows. Section 6.2
describes the experimental methodology by which th e reactor is
characterized. Reactor block diagram and experimental param eters are
used to describe the experimental methodology. Section 6.3 describes the
experimental configuration and the common experimental procedures
150
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
used in the param etric study. The characteristic behaviors of this
experimental configuration are described. Section 6 .4 describes the
results of a n experim ental param etric stu d y of diam ond film deposition
on 3 ” silicon wafers. The goal of this study is to achieve the highest
growth rate possible since achieving good uniform ity (judged by visual
inspection) over 3" silicon wafer is possible u n d er a wide range of
experimental conditions in this MCPR. Film growth rates, growth
efficiencies an d morphologies of the deposited diam ond films under
various deposition conditions are described. Typical uniformity profile
and R am an spectrum are also presented. Section 6.5 describes a study
of experimental set-up for diamond film deposition on 4 Nsilicon wafers.
Here film uniform ity an d growth rate are th e judging criteria. Section 6.6
describes a param etric study of diamond film deposition on 4” silicon
wafers. The goal of th is study is to achieve the highest growth rate
possible while m aintaining good uniformity (judged by visual inspection).
There exists a compromise between the uniform ity and growth rate in
diamond film deposition on 4” silicon wafers in the p resen t reactor. When
reasonable uniform ity is required, the m ain limiting factor on the growth
rate comes from the safe operating tem perature limit and th e coating
speed of the quartz dome. Film growth rates, growth efficiencies and
morphologies of diam ond films deposited u n d er various experimental
conditions are described. Typical uniformity profile and Ram an spectrum
are also presented.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
6.2
Experimental Methodology
6.2.1
Introduction
In order to orderly investigate th e performance of th e MCPR, it is
essential to identify th e experimental param eters an d establish a method
of evaluating the operational performance of th e MCPR. This section
describes th e various experim ental param eters, m ethods and “figures of
m erit” u sed in the evaluation.
The deposition experim ents conducted in MCPR have m any
experim ental param eters. The reactor I/O block diagram shown in Figure
6.1 displays the relationship between these experim ental param eters.
These experimental param eters can be divided into three groups: (1)
independent input experim ental param eters, (2) internal, dependent
experim ental param eters, and (3) external, o u tp u t param eters. Each of
them is sum m arized in m ore detail below.
I.
Independent Input Experimental Param eters
(i)
Substrate size, shape an d other physical properties
(ii)
Reactor geometry param eters
(a)
cavity applicator construction
(b)
quartz dome geometry
(c)
base-plate design
(d)
gas flow configuration
(e)
su b strate holder geometry and
(f)
su b strate location
(g)
end feed vs. side feed
(h)
electromagnetic mode excitation an d cavity tuning
m aterial
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
153
Reactor geometry parameters
Seeding procedures
Pressure
Microwave
power
Growth rate
Morphology
Substrate temperature
Discharge volume
^
Uniformity
^
Raman spectra
Power density
Gas
composition
Residence time
Growth
Efficiency
etc.
Total
flow rate
^
etc.
Start-up and shut-down procedures
Deposition time
etc.
Figure 6 .1 Reactor I/O Block Diagram
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
(j)
(iii)
vacuum system - pum ping speed, volume, etc.
Deposition process param eters
(a)
su b strate seeding procedure
Cb)
start-u p and shut-dow n procedures
(c)
in p u t power variation vs. time (CW vs. pulsed power,
(d)
deposition time, etc.
etc.)
(iv)
Macroscopic controllable in p u t param eters
(a)
operating pressure, p
(b)
absorbed microwave power, Pt
(c)
gas composition expressed in % of H 2 gas flow, i.e. H 2,
CH 4/H 2, CO2/H 2, etc.
(d)
II.
total in p u t gas flow rate, ft-
Internal, D ependent Experim ental Param eters
(i)
S u b strate tem perature, Ts
(ii)
Discharge volume, Vd, an d deposition area, Aj
(iii)
Discharge power density, w
(iv)
Gas residence time, tr
(v)
Electric field strengths an d electrom agnetic field
distributions w ithin th e reactor
(vi)
D ensities of discharge species
(a)
densities of charged species
(b)
densities of excited species
(c)
densities of atomic an d m olecular species
(d)
variation of these densities w ithin the reactor
(especially above the substrate)
(vii)
Energy distributions an d tem p eratu res of discharge species
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
(viii) Flow patterns of discharge species w ithin the reactor
(ix)
III.
6.2.2
Fluxes of discharge species onto th e su b strate
External, O utput Param eters
(i)
Film linear growth rate (pm /hr) and weight gain, u (mg/hr)
(ii)
Film morphology
(iii)
Film uniformity
(iv)
Film Ram an spectra
(v)
Film growth efficiency
(a)
growth rate vs. power, Pt
(b)
growth rate vs. power density, w
(c)
growth rate vs. power flux, P t/A j
(d)
growth rate vs. total flow rate, ft
(e)
carbon atom conversion efficiency
Independent, Input Experim ental Param eters
It is obvious th a t the experiment h ere h a s m any independent input
param eters. T h u s in order to develop a n u n d erstanding of the
operational performance of the reactor, it is b est to empirically optimize
and fix som e of the independent experim ental param eters.
T he su b stra te size and physical properties are usually fixed. Three
and four in ch silicon wafers are used in th e experiments described in
this chapter. The cavity applicator construction, quartz dome geometry,
base-plate design, gas flow configuration, su b stra te holder geometry and
m aterial, su b strate location, power feed configuration, and cavity
electrom agnetic mode excitation are empirically optimized to create the
optim um discharge geometry for uniform su b stra te heating and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
156
m inim um quiartz dome heating and coating.
Among the deposition process param eters, th e seeding an d sta rt­
up procedures are empirically optimized to generate reproducible and
high diamond nucleation density. The s h u t down procedure is chosen
su ch th a t the formation of non-diam ond carbon on th e deposited film
surface is minimized. CW power is used in all the experim ents described
in th is chapter. The deposition time is fixed for each set of experiments.
More detailed description of these independent param eters is given in
relevant sections.
The four macroscopic controllable in p u t experimental param eters,
i.e., pressure (p), absorbed microwave power (Pt). gas composition and
total gas flow rate (ft), are varied during the experim ents in order to
u n d erstan d the characteristic behavior of MCPR u n d er diamond film
deposition conditions an d to optimize the diamond film deposition
process in MCPR.
6.2.3
Internal, D ependent Experimental Param eters
The internal, dependent experimental param eters vary w ith
changes in the independent experimental param eters. S ubstrate
tem perature Ts and discharge volume
are two im portant internal
experimental param eters in th e diamond film deposition process. The
su b strate tem perature Ts greatly affects the growth rate, morphology and
Ram an spectra of the deposited film. The discharge volume V j
determ ines the deposition a rea and film uniformity.
These internal experim ental param eters are determ ined a n d /o r
defined a s follows:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157
su b strate tem perature Ts is m easured by a n optical pyrometer,
discharge volume Vd is estim ated by visual estimation,
the discharge power density w is defined by
w = Pt/V d
where Pt is the absorbed microwave power, and
th e gas residence tim e tr is defined by
tr = pVq/f t (sec)
where p (expressed in Torr) is th e pressure, Vq (expressed in liters) is the
quartz dome volume above the su b strate and ft (expressed in Torr-liter/
sec, 1 T orr-liter/sec = 79.05 seem) is the total flow rate.
These dependent in tern al experimental param eters vary in a non­
linear fashion w hen th e in p u t experimental param eters vary. In order to
control the deposition process, these param eters m u st be controlled. It is
also often desirable to m easu re some of these internal experimental
param eters in-situ during th e deposition process. If these m easurem ents
are performed v ersus tim e during the deposition process, the
understanding of the deposition process can be greatly improved.
The other internal, dependent experimental param eters will n o t be
discussed in th is dissertation.
6.2.4
External, O u tp u t Experimental Param eters
The reactor perform ance is evaluated in term s of th e external,
output experimental param eters. These o u tp u t p aram eters can be
m easured a n d /o r calculated after the film deposition is completed. Film
growth rate is expressed in term s of both linear growth rate (pm/hour)
and weight gain, u (m g/hour). Weight gain is obtained from the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
deposition tim e and weighing th e wafer before an d after th e deposition
process using a weight balance. The linear growth rate is calculated from
th e weight gain, the area of the w afer an d the density of diamond. 3.51
g /c m 3.
The film morphology is obtained from observation under the optical
microscope and photographs tak en w ith Scanning Electron Microscope
(SEM). The film uniformity is obtained by both visual estimation and
laser interference reflection technique.<71> The film quality is
characterized by R am an spectroscopy.
The following “figures of merit" are u sed to quantify th e film growth
efficiency of the MCPR:
(1)
k j (mg/kW-hr) = u /P t: u versus Pt. growth rate versus
absorbed microwave power,
(2)
k 2 (mg-cm3/kW -hr) = u /w : u versus w, growth rate versus
power density.
(3)
k 3 (mg-cm 2/kW -hr) = u /S : u versus S, growth rate versus
power flux S. S is defined by S = Pt/Ajj. where
is the area of the
substrate.
(4)
gas flow efficiency, k 4 (mg/liter) = u / f t: u versus ft, growth
rate versus total flow rate.
(5)
carbon conversion efficiency, k s (%): it is defined as the
percentage of carbon atom s in the in p u t gases th a t are converted into the
diam ond film.
These output param eters allow th e MCPR to be compared w ith
other reactors and allow the com parison of the MCPR as experimental
in p u t conditions change. It is also noted here th a t the optimization of one
performance criteria, su ch as uniformity, m ay n o t optimize another
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
performance criteria, su ch as growth rate.
6.3
6.3.1
Experimental Operational Characteristics
Introduction
In th is section, the basic experimental configuration for diamond
film deposition on 3” and 4" silicon wafers is described. The common
experim ental techniques, i.e., seeding, start-u p an d shut-dow n
procedures, are described in detail. The operational characteristics of
this experimental configuration, i.e., internal experim ental param eters
vs. in p u t experimental param eters, are presented.
6.3.2
Experimental Configuration
The forced flow experimental configuration w hich w as described in
detail in chapter 3 is the experim ental configuration u sed for diamond
film deposition on 3Mand 4 ” silicon wafers. A schem atic drawing of this
experim ental configuration is displayed in Figure 3.16.
Figures 6.2 an d 6.3 display th e schem atic drawings of two cavity
shells u sed in the experim ents. The m ain difference between the two
cavity shells is th a t there exist four small windows (optical access
windows for emission an d laser induced fluorescence spectroscopy
experiments) in the cavity shell (II). The microwave power coupling
efficiency of cavity shell (II) is better th a n th a t of cavity shell (I). B ut the
discharge generated w ith cavity shell (I) is confined to the silicon wafer
better th a n th a t with cavity shell (II). The differences are especially
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
---------------------------------------------
Viewing
window
Viewing
window
Figure 6.2 Schem atic Drawing of Cavity Shell (I)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
V ie w in g
window
O p t i c a l window
Figure 6.3 Schem atic Drawing of Cavity Shell (II)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
obvious w hen diamond films are deposited on 4” silicon wafers. For
diam ond film deposition on silicon wafers, cavity shell (I) is a more
favorable choice.
The quartz domes m ade by different m anufacturers an d those
constructed at different tim es by the sam e m anufacturer result in small
differences in geometry. These small differences in geometry do not have
m uch effect on diamond film deposition on su b strate surfaces which are
3.25” in diam eter or sm aller. B ut when diamond films are deposited on
4 ” silicon wafers, these sm all differences in geometry influence the
optim um deposition condition, cavity tuning and th e resulting film
morphology, uniformity, and growth rate, etc.
In th e present experimental configuration, the substrate holder
(13) an d flow regulator (32) are combined into one holder (13:32). The
m etal tu b e (29) used here h as an outside diam eter of 3”, inside diameter
of 2.875” and a length of 1.5”.
6.3.3
Common Experimental Procedures
In th is section, th e common experimental procedures, i.e., seeding,
start-u p and sh u t down procedures, which are used in all the
experim ents discussed in th is chapter, are described in detail.
6.3.3.1
Seeding Procedures
S ubstrate seeding h a s been used prior to deposition experim ent to
enhance diamond nucleation density. Three types of seeding procedures
have been used in the experiments described in th is dissertation. They
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163
include (1) seeding by h an d scratching su b stra te w ith 0.25 pm diamond
paste, (2) seeding by h an d scratching su b strate w ith 1 pm diamond
powder and (3) seeding by photo resist seeding m ethod<22,72>.
The first two m ethods are self-explanatory. They are simple to
perform b u t the repeatability is poor. The third m ethod is the best in
term s of both repeatability and nucleation density and it is the seeding
method used in all the experiments described in th is chapter.
The photo resist seeding method w as derived from th e standard
photolithography procedure used in integrated circuit fabrication. In this
method, diamond powder an d photo resist are first mixed ultrasonically.
The resulting m ixture is then used to cover the entire wafer. The wafer is
in tu rn spined in a spinner which results in a uniform coating on the
wafer. Diamond seeds are embedded in th is uniform coating.
A standard recipe<73> of the photo resist seeding procedure is as
follows:
(a) mix 142 mg Amplex <74> 0.1 pm diam ond powder with 16
ml of Shipley<75> type A photoresist thinner,
(b) ultrasonically mix the m ixture for 15 m inutes,
(c) add 42 ml of Shipley 1470 photo resist,
(d) ultrasonically mix the m ixture for 15 m inutes,
(e) apply the m ixture onto th e wafer,
(f) spin th e wafer with the following sp in speed and time: 0 4000 rpm in 10 seconds, stay a t 4000 rpm for 30 seconds an d slow down
to 500 rpm in 5 seconds,
(g) bake the wafer a t 135 °C for 30 m inutes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
6.3.3.2
S tart-up an d Shut-down Procedures
The com puter m onitor system described in section 3.2.4 is used in
the experiments described in this chapter. The following start-u p and
shut-dow n procedures are followed.
I.
Start-up procedure:
(a) evacuate th e system to below 5 mTorr,
(b) tu rn on H 2, CH4 and CO2 gas flow,
(c) tu rn on 1.5 kW microwave power w hen pressure reaches
10 Torr,
(d) increase microwave power as pressure increases,
(e) start tim ing w hen pressure and microwave power reach
set levels.
In the present experim ental system, th e p ressu re rise time is about
10 T orr/m in with 400 seem H 2 gas flow.
II.
Shut-down procedure:
(a) tu rn off CH4 and CO2 gas flow together,
(b) wait 3 or 5 m inutes for process self-cleaning,
(c) tu rn off microwave power,
(e) tu rn off H 2 gas flow,
(f) evacuate the experimental system.
6.3.4
Experimental Operational Characteristics
It was shown in section 6.2 th a t the microwave cavity plasm a
reactor can be characterized by a set of input, in tern al an d output
experimental param eters. As displayed in Figure 6.1, th e substrate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
tem perature, Ts> and discharge volume, Vj, are internal experimental
param eters which depend on all independent in p u t experimental
param eters. In this section, the characteristic behavior of th e internal
experim ental param eters is investigated as th e in p u t experimental
param eters, i.e., pressure, microwave power, gas composition and total
gas flow rate, are varied.
The forced flow experim ental configuration show n in Figure 3.16 is
u sed in all the experim ents described in this section. Here, th e
supporting quartz tu b e (28) h a s a n inside diam eter of 95 mm, an outside
diam eter of 100 mm an d a length of 50 mm.
Three inch silicon wafers are used as su b strates in these
experim ents. They have th e following physical properties: prim e grade, ptype, Boron doped, <100> orientation, 1 - 1 0 ohm-cm in resistivity, 76
m m in diameter, and 356 - 406 pm in thickness.
To help describing th e characteristic behavior of the experimental
reactor, the following nom enclatures are defined:
p <--> Pressure,
Pt <—> Absorbed Microwave Power,
ft <--> Total Gas Flow Rate,
Cg <--> G as Composition,
Ts <—> S u b strate Tem perature,
V j <—> Discharge Volume.
The experimental procedures to obtain the characteristic behavior
of the internal experim ental param eters are a s follows. A diam ond film
w as deposited on a 3" silicon w afer u n d er the following experimental
conditions: H2 flow rate = 400 seem, CH4 flow rate = 6 seem, CO2 flow
rate = 2 seem, pressure = 51 Torr, absorbed microwave power = 2.34 kW,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
su b stra te tem perature ~ 900 °C an d deposition tim e = 5 hours. This
diam ond film was estim ated to be more th a n 2.5 pm thick, judging from
previous experiments conducted u n d er sim ilar experimental conditions.
All th e m easurem ents were conducted after this diamond film h ad been
formed. The substrate tem perature. Ts, w as m easured by a n optical
pyrom eter which w as m ounted on a tripod and fixed at the sam e location
throu g h o u t the entire experiment. The discharge volume, Vd, w as
estim ated by visual estim ation. The neighboring su b strate tem perature
readings an d discharge volume estim ations were recorded by varying a
m inim um num ber of in p u t experimental param eters and waiting at least
5 m in u tes for the experiment to reach steady state.
Figure 6.4 shows the variation of substrate tem perature Ts vs.
p ressu re p and absorbed microwave power P^. As shown, th e substrate
tem perature Ts is very sensitive to variations in pressure p an d only
varies slowly as the absorbed microwave power Pt is varied.
The discharge volume Vd is sensitive to both the variation in
p ressu re p and absorbed microwave power Pt. Keeping pressure p
constant, the discharge volume Vd increases with increasing absorbed
microwave power Pt- Keeping th e absorbed microwave power Pt constant,
th e discharge volume Vd decreases w ith increasing pressure p.
In Figure 6.4, the lower power limit is defined by the m inim um
power needed to generate a discharge volume (Vd)min th a t covers a 3"
diam eter substrate area. The u p p er power limit is defined by the
m axim um power th a t can be used to operate the reactor safely w ithout
over-heating th e quartz dome. The upper discharge volume lim it (Vri)may
is the volume of the quartz dome above th e substrate.
Figure 6.5 shows the variation of substrate tem perature Ts vs. CH4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
980
p = 60 T o n
Substrate Temperature, Ts (°C)
950
p = 55 Ton920
p = 50 Ton890
860
p = 45 Ton-
830
H2 = 400 seem
CH4 = 2 seem
3" silicon wafer
p = 40 T o n
800
1.9
2.2
2.5
Absorbed Microwave Power, Pt (kW)
Figure 6.4 S ubstrate Temperature versus Pressure and
Absorbed Microwave Power
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.8
168
896
k L 894
P*
e
B
892
2
<D
Q-
.1 890
H
S
-l
•4t—
»
p = 50 Torr
Pt = 2.32 kW
H 2 = 400 seem
3" silicon wafer
886
0
1
2
3
4
5
CH4 Flow Rate (seem)
Figure 6.5 S ubstrate Tem perature versus CH 4 Concentration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
concentration. As shown, the su b strate tem perature Ts is n ot sensitive to
variation in CH4 concentration.
Figure 6.6 shows the variation of substrate tem perature Ts vs. CO2
concentration. As shown, the su b strate tem perature Ts varies slowly with
low CO2 concentration and decreases w hen CO2 concentration is further
increased. Low CO2 concentrations are used in th e diamond film
deposition experiments described in th is chapter. Figures 6.5 and 6.6
show th a t the su b strate tem perature Ts is not sensitive to variation in
gas composition Cg. It is also found by visual estim ation th a t the
discharge volume Vd is not sensitive to variation in gas composition Cg.
Figure 6.7 shows the variation of substrate tem perature Ts vs.
absorbed microwave power Pt an d total flow rate ft. As shown, the
su b strate tem perature Ts increases gradually w hen the absorbed
microwave power Pt is increased and it is not sensitive to variations in
total gas flow rate ft- It is also found by visual estim ation th a t the
discharge volume Vd is n ot sensitive to variations in total gas flow rate fj.
6.3.5
Sum m ary
The following conclusions can be drawn from the experiments
described in the previous section:
(1)
Substrate tem perature Ts increases sharply w ith increasing
pressu re (~ 6 °C/Torr) an d a t constant pressure, Ts increases gradually
(~ 0.02 °C/W) with increasing absorbed microwave power. It is not
sensitive to variations in gas composition and total flow rate.
(2)
Discharge volume Vd decreases with increasing p ressu re and
increases w ith increasing absorbed microwave power. It is n o t sensitive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
Substrate Temperature, Ts (°C)
892
890
888
886
884
Ho = 400 seem
CH4 = 4 seem
p = 50 Torr
Pt = 2.32 kW
3" silicon wafer
882
880
0
1
2
3
4
5
CO2 Flow Rate (seem)
Figure 6.6 S u b strate Temperature versus CO 2 Concentration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
896
seem Ho
seem H 2
seem Ho
seem Ho
seem Ho
* 894
Vh
B
892
Ui
a,
S
890
p = 50 Torr
CH4/H 2 = 1%
3" silicon wafer
888
^
886
2.1
2.2
2.3
2.4
2.5
2.6
Absorbed Microwave Power, Pt (kW)
Figure 6.7 S ubstrate Tem perature versus Total Flow Rate and
Absorbed Microwave Power
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172
to variations in gas composition and total flow rate.
Using the definitions of discharge power density w and power flux
S,
w = Pt/V d
S = Pt/Ad
where A^ is the su b strate area, the following additional conclusions can
be drawn.
(3)
Power density, w, increases with increasing pressure and at
constant pressure, it is not sensitive to variation in absorbed microwave
power. That is, u nder a constant pressure, as th e absorbed microwave
power increases, the discharge volume also increases, resulting in small
variation in absorbed power density. Power density is n ot sensitive to
variation in gas composition and total flow rate.
(4)
Power flux S increases with increasing absorbed microwave
power. It is not sensitive to variation in gas composition and total flow
rate.
These results on the characteristic behavior of the MCPR under
diamond film deposition conditions dem onstrate th a t the reactor behaves
in an unique repeatable fashion. A set of experimental curves can be
used to describe the su b strate tem perature variations as pressure,
absorbed microwave power, gas composition and total flow rate vary.
This set of curves can be used to understand and describe the
experimental behavior of the reactor. The allowable experimental
operating space is defined by these curves. If certain substrate
tem perature conditions are desired, these curves can be utilized to
determine the pressure and absorbed microwave power required. These
curves serve as the experimental reactor operating energy “road map".
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173
6.4
6.4.1
Diam ond Film Deposition on 3 ” Silicon Wafers
Introduction
Diamond films w ith various physical properties are deposited on 3"
silicon wafers u n d er various experim ental conditions. In order to
optimize th e diam ond film deposition process, it is necessary to study the
co-relation between th e physical properties of th e deposited diam ond
films an d th e experim ental conditions. Hence a param etric stu d y of
diam ond film deposition on 3 ” silicon wafers w as conducted. In this
study, the following experimental param eters are considered:
I
II.
III.
The independent in p u t experim ental param eters:
(1)
absorbed microwave power, Pt,
(2)
pressure, p,
(3)
gas (H2, CH4, CO2) composition, and
(4)
total flow rate, ft-
The internal experim ental param eters:
(1)
su b strate tem perature Ts, an d
(2)
gas residence time, tr.
The o u tp u t experim ental param eters:
(1)
film linear growth rate (jim /hour),
(2)
film weight gain u (m g/hour),
(3)
film growth efficiencies:
(4)
(i)
growth rate vs. power flux, i.e., u vs. S,
(ii)
gas flow efficiency: IC4 (mg/liter) = u /f t,
(iii)
carbon conversion efficiency, ks (%).
film morphology,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174
(5)
typical film uniformity, and
(6)
typical film R am an spectrum .
In th is study, diam ond films are deposited o n 3” silicon wafers
using the forced flow configuration shown in Figure 3.16. The physical
properties of the 3" silicon wafers and quartz tu b e (28) used in these
experim ents are the sam e as th a t described in section 6.3.4.
Descriptions of relevant experimental operational energy and
chem istry space are also included whenever possible.
6.4.2
Film Growth Rate an d Growth Efficiency
6.4.2.1
Effects of S u b strate Tem perature
S ubstrate tem perature is a n im portant internal experimental
param eter in diamond film deposition process. In th is study, the
su b strate tem peratures are achieved by choosing th e appropriate
pressures. The absorbed microwave powers are adjusted such th at the
discharge volumes are large enough to cover th e silicon substrates.
Figure 6.8 displays diam ond film linear growth rate and weight
gain versus substrate tem perature and microwave power flux with gas
flow rates o f 200 seem H 2 and 4 seem CH4. The pressu res an d absorbed
microwave powers u sed to achieve the three su b strate tem peratures are
show n in Table 6 .1 located above Figure 6 .8 .
Figure 6.9 (a) show s th e locations of these su b strate tem perature
conditions on the characteristic energy “map". Figure 6.9 (b) displays the
relevant carbon conversion efficiency and gas flow efficiency vs. substrate
tem perature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
Table 6.1
Substrate Temperature Conditions (I)
Temperature, Ts (°C)
Pressure, p (Torr)
MW power, Pt(kW)
750
33
1.47
850
45
2.05
950
57
2.51
Power Flux, S (W/cm2)
[15.80
12.64
0.6
9.48
0.4
6.32
0.2
3.16
2
o
750
850
950
Substrate Temperature, Ts (°C)
Figure 6.8 Linear Growth Rate and Weight Gain versus
Substrate Tem perature and Power Flux (I)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
0.8
H2 = 200 seem
CH4 = 4 seem
3" silicon wafer
176
H2 = 200 seem
CH4 = 4 seem
3" silicon wafer
750% Highest growth rate point
-L .-------------- 1--------------- -I— _________ L
9
0.95
7
0.74
5
0.53
(5s.
&
l
I
c
1
I
O
o
d
0
1
o
3
750
850
950
Gas Flow Efficiency (mg/liter)
1.6
1.8
2
2.2
2.4
2.6
2X
Absorbed Microwave Power, Pt (kW)
(a)
0.32
Substrate Temperature, Ts (°C)
Cb)
Figure 6.9
(a) Location of Substrate Temperature Conditions on Energy
Map and Cb) Carbon Conversion Efficiency an d Gas
Flow Efficiency versus Substrate Tem perature (I)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
Figure 6 .10 displays diamond film linear growth rate and weight
gain versus su b strate tem perature and microwave power flux with gas
flow rates of 200 seem H2. 2 seem CH4 and 2 seem CO2. The pressures
and absorbed microwave powers to achieve the five tem peratures are
shown in Table 6.2 located above Figure 6.10.
Figure 6 .11 (a) shows the locations of these su b strate tem perature
conditions on the characteristic energy “m ap”. Figure 6 .11 (b) displays
the relevant carbon conversion efficiency an d gas flow efficiency vs.
su b strate tem perature.
Figure 6 .12 shows diamond film linear growth rate an d weight gain
versus su b strate tem perature and microwave power flux w ith gas flow
rates of 400 seem H2. 4 seem CH4 and 2 seem CO2. The pressures and
absorbed microwave powers used to achieve the three tem peratures are
shown in Table 6.3 located above Figure 6.12.
Figure 6.13 (a) shows the locations of these su b strate tem perature
conditions on the characteristic energy “m ap”. Figure 6.13 (b) displays
the relevant carbon conversion efficiency and gas flow efficiency vs.
su b strate tem perature.
As displayed in Figures 6.8 through 6.13, th e linear growth rate,
weight gain, carbon conversion efficiency and gas flow efficiency generally
experience peak values w hen the su b strate tem perature is varied
between 700 °C and 1000 °C. It is also noted from these figures th a t the
peak values are gas composition dependent. The locations of these gas
compositions on the B achm ann C-H-O phase diagram are show n in
Figure 6.14. It is also show n in Figures 6.9(a), 6.11(a) and 6.13(a) th a t
the com binations of p ressures an d absorbed microwave powers to
achieve these peak values fall w ithin the boundaries of th e characteristic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
Table 6.2
Substrate Temperature Conditions (II)
T em perature, Ts (°C)
750
800
850
900
950
Pressure, p (Torr)
33
39
45
51
57
MW power, Pt (kW)
1.48
1.76
2.05
2.34
2.64
Power Flux, S (W /cm2)
0.3
4.74
H2 = 200 seem
CH4= 2 seem
C02 = 2 seem
3" silicon wafer
3.95
2
tUD
3.16
0.2
2
o
0.15
2.37
750
800
850
900
950
Substrate Temperature, Ts (°C)
Figure 6 .10 Linear Growth Rate a n d Weight Gain versus
Substrate Temperature a n d Power Flux (II)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
179
_
1000
O
o_^
r? 95d
1
90C
2v
a
S
85C
s
80C
H2 = 200 seem
CH4 = 2 seem
C 0 2 = 2 seem
3” silicon wafer
1
*
cn
*§
CO
75C
7<Y
% Highest growth rate point
1
1
1.6
■
1.8
2
■
■
*
*
2.2
2.4
2.6
2.8
Gas Flow Efficiency (mg/liter)
Absorbed Microwave Power, Pt (kW)
(a)
m
750
800
850
900
950"
Susbstrate Temperature, Ts (°C)
(b)
Figure 6.11
(a) Location of Substrate Tem perature Conditions on Energy
Map and (b) Carbon Conversion Efficiency and Gas
Flow Efficiency versus S ubstrate Temperature (II)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
180
Substrate Temperature Conditions (III)
Temperature, Ts (°C)
Pressure, p (Torr)
MW power, Pt (kW)
0.55
§
°-5
927
50
2.61
964
55
2.67
Power Flux, S (W/cm2)
59.3
60.9
H2 = 400 seem
CH4 = 4 seem
C 0 2 = 2 seem
. 3" silicon wafer
1001
60
2.74
8.69
7.90
I
V. 0.45
7.11
o
O
0.4
6.32
0.35
5.53
927
964
1001
Substrate Temperature, Ts (°C)
Figure 6 .12 Linear Growth Rate and Weight Gain versus
S ubstrate Tem perature and Power Flux (III)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
Table 6.3
181
O
p_r
,
100(
(0 950-
1
900-
2<u
I
s
H2 = 400 seem
CH4 = 4 seem
C02 = 2 seem
3" silicon wafer
850-
800-
I
£
CO
75C-
# Highest growth rate point
CO
70I 4
4.5
Gas Flow Efficiency (mg/liter)
sp
o'
^
1.6
1.8
2
2.2
2.4
2.6
2.8
Absorbed Microwave Power, Pt (kW)
(a)
I
I
hi
§
1
a;
£
O
O
G
O
t
O
927
964
1001
Substrate Temperature, Ts (°C)
(b)
Figure 6.13 (a) Location of S ubstrate Tem perature Conditions on Energy
Map and (b) Carbon Conversion Efficiency and Gas
Flow Efficiency versus S ubstrate Tem perature (III)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
&?
Figure 6.14
<NI
Location of Gas Compositions on the C-H-O Phase Diagram
(I)
182
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183
energy “m ap”.
In the growth rate and growth efficiency vs. gas composition and
total flow rate experiments th a t follow, th e su b stra te tem perature is
chosen to be fixed a t either 900 °C or 950 °C. The CH4 concentration is
varied first in the growth rate vs. gas composition experiments.
6.4.2.2
Effects of CH4 Concentration
Figure 6.15 displays the diamond film linear growth rate, weight
gain and carbon conversion efficiency vs. CH4 gas concentration u nder
the following experimental conditions: H2 flow rate = 400 seem, pressure
= 51 Torr, absorbed microwave power = 2.34 kW, substrate tem perature
~ 900 °C an d deposition time = 6 hours.
As shown, the film linear growth rate an d weight gain experience
peak values with CH4 to H2 volume ratio of ab o u t 6 /4 0 0 = 1.5%. The
carbon conversion efficiency decreases as CH4 concentration is
increased.
Figure 6.16 displays th e diamond film linear growth rate, weight
gain an d carbon conversion efficiency vs. CH 4 gas concentration u nder
the following experimental conditions: H2 flow rate = 400 seem, pressure
= 57 Torr, absorbed microwave power = 2.5 kW, su b strate tem perature ~
950 °C and deposition time = 6 hours.
As shown, the film linear growth rate an d weight gain experience
peak values w ith CH4 to H2 volume ratio of ab o u t 5 /4 0 0 = 1.25%. The
carbon conversion efficiency decreases w hen CH4 concentration is
increased beyond th is optim um CH4 concentration.
Figure 6.17 displays th e locations of the gas compositions used in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
v°
7
2 ,
I
a
s
6
u
8
5
1O
1o
4
3
0.7
i
0.6
I
111.06
H? = 400 seem
Ts ~ 900 °C
p = 51 Torr
Pt = 2.34 kW
3 silicon wafer
9.48
H 0.5
7.90
0.4
6.32
0.3
4.74
8
O
CH4 Flow Rate (seem)
Figure 6.15 Linear Growth Rate, Weight Gain and Carbon
Conversion Efficiency versus CH4 Concentration (I)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
O
e0
185
8
6
4
2
5.5
0.7
11.06
0.6
9.48
|
0.5
I
7.90
«
0.4
6.32
O 0.3
4.74
f
;§
0.2
4.5
5.5
CH4 Flow Rate (seem)
3.16
Figure 6.16 Linear Growth Rate, Weight Gain and Carbon
Conversion Efficiency versus CH4 Concentration (II)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
4.5
o
ii n
uu
o<
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.17
&?
CM
Location of Gas Compositions on the C-H-O Phase Diagram
(II)
186
187
Figures 6.15 and 6.16 on the Bachm ann diagram .
The experim ental param eters used to obtain the peak growth rate
values are used as the starting experimental param eters in the following
sections.
6.4.2.3
Effects of CO 2 Concentration
Figure 6.18 displays the diamond film linear growth rate, weight
gain an d carbon conversion efficiency vs. CO2 gas concentration under
th e following experimental conditions: H2 flow rate = 400 seem, CH4 flow
rate = 6 seem, pressure = 51 Torr, microwave power absorbed = 2.34 kW,
su b strate tem perature - 900 °C and deposition tim e = 6 hours.
As shown, the film linear growth rate, w eight gain and carbon
conversion efficiency experience gradual decline a s the CO2 gas
concentration is increased. Ciystalline diam ond films are deposited with
these four gas compositions.
Figure 6.19 displays the film linear growth rate, weight gain and
carbon conversion efficiency vs. CO2 gas concentration u n d er the
following experimental conditions: H2 flow rate = 200 seem, CH 4 flow rate
= 6 seem, pressure = 51 Torr, microwave power absorbed = 2.46 kW, and
deposition time = 6 hours.
As shown, th e film linear growth rate, weight gain and carbon
conversion efficiency experience sharp increases a s CO2 flow rate is
increased from 3 to 3.5 seem, where a d ark film is deposited with 3 seem
CO2 gas flow and a crystalline diamond film is deposited w ith 3.5 seem
CO2 gas flow. The film linear growth rate, weight gain and carbon
conversion efficiency are decreased as CO2 gas flow rate is further
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Carbon Conversion Efficiency (%)
188
T
I
T
11.06
Ht = 400 seem
CEL = 6 seem
0.6
Pt = 2.34 kW
3" silicon wafer
9.48
0.5
7.90
0.4
6.32
C 0 2 Flow Rate (seem)
Figure 6.18 Linear Growth Rate, Weight Gain and Carbon
Conversion Efficiency versus C 0 2 Concentration (I)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
Growth Rate (^m/hour)
0.7
189
<g».
|
3S
3.5
5
o
3
Cd
§
0
<J
g 2.5
1
3.5
4.5
0.8
12.64
0.7
11.06
0.6
9.48
I
2
I
<•>
S
0-5
0.4
Hi = 200 seem
CHa = 6 seem
p = 51 Torr
Pt = 2.46 kW
3" silicon wafer
3.5
4
4.5
C 0 2 Flow Rate (seem)
7.90
6.32
Figure 6.19 Linear Growth Rate. Weight G ain and Carbon
Conversion Efficiency versus C 0 2 Concentration (II)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
O
190
increased from 3.5 seem to 4.5 seem. Crystalline diamond films are
deposited in the later three cases.
Figure 6.20 shows th e locations of the gas compositions used in
Figures 6.18 and 6.19 on th e B achm ann C-H-O phase diagram.
From Figures 6.15 an d 6.18, we see the optim um gas composition
for high rate diamond film growth a t 900 °C is the following: H 2 flow rate
= 400 seem and CH4 flow rate = 6 seem. This gas composition is used in
the following growth rate versus total flow rate experiments.
6.4.2.4
Effects of Total Flow Rate
Figure 6.21 displays the diamond film linear growth rate, weight
gain an d carbon conversion efficiency vs. total gas flow rate an d gas
residence time under the following experimental conditions: CH4 to H2
volume ratio = 1.5%, p ressu re = 51 Torr, absorbed microwave power =
2.34 kW, substrate tem perature - 900 °C and deposition time = 6 hours.
Figure 6.22 displays the diam ond film linear growth rate, weight
gain an d carbon conversion efficiency vs. total gas flow rate an d gas
residence time under the following experimental conditions: CH4 to H2
volume ratio = 1.25%, p ressu re = 57 Torr, absorbed microwave power =
2.5 kW, substrate tem perature ~ 950 °C and deposition time = 6 hours.
As shown in Figures 6.21 and 6.22, the film linear growth rate and
weight gain increase with increasing total flow rate w hen the total flow
rate is low and gas residence tim e is high. As the total flow rate is further
increased and the gas residence tim e decreases, th e linear growth rate
and weight gain sta rt to satu rate. The carbon conversion efficiency
decreases with increasing total flow rate an d decreasing gas residence
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CM
• *
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.20
Location of Gas Compositions on the C-H-O Phase Diagram
\
(III)
191
192
Residence Time, tr (seconds)
f
6.5
12.64
11.06
0.7
9.48
1
0.5
7.90
% 0,4
2
6.32
O
C H 4 AH7 =
0.3
0.2
1 .5
%
Pt = 2.34 kW
3" silicon wafer
50
100
200
(a)
300
400
50
100
200
300
400
4.74
Weight Gain (mg/hour)
8.7
0.8
3.16
I<u
I
a
c
0
£
1
o
O
a
o
t
o
H2 Flow Rate (seem)
(b)
Figure 6.21 (a) Linear Growth Rate, Weight G ain and (b) Carbon
Conversion Efficiency versus Total Flow Rate and Gas Residence Time (I)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
Residence Time, tr (seconds)
14.5
0.7
9.7
1.06
*
I
9.48 J:
0.6
WD
•v.
't
a
7.90
0.5
1
0
6.32
0.4
£
3 silicon wafer
0.3
100
200
300
4.74
400
300
400
(a)
a
2
1
U
G
2
2
o
O
c3
a
100
200
H2 Flow Rate (seem)
(b)
Figure 6.22 (a) Linear Growth Rate, Weight Gain and (b) Carbon
Conversion Efficiency versus Total Flow Rate an d Gas Residence Time (II)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194
time.
6.4.3
Film Morphology
6.4.3.1
Introduction
Scanning Electron Microscopy (SEM) is u sed to identify the
morphologies of films deposited u n d er various experimental conditions.
The films deposited u n d er the various experim ental conditions described
in section 6.4.2 are characterized.
6.4.3.2
Effects of S ubstrate Tem perature
Figure 6.23 displays th e SEM photographs of three films deposited
w ith su b strate tem peratures: (a) 750 °C, (b) 850 °C and (c) 950 °C. The
pressu res and absorbed microwave powers used to achieve these
tem peratures are listed in Table 6 .1 located above Figure 6.8 . The other
deposition conditions are: H2 flow rate = 200 seem. CH4 flow rate = 4
seem, an d deposition time = 4 hours.
As shown, the deposited films evolve from “cauliflower” film a t 750
°C, to crystalline film a t 850 °C an d sm all grain film a t 950 °C.
Figure 6.24 displays th e SEM photographs of five films deposited
w ith su b strate tem peratures: (a) 750 °C, Cb) 800 °C, (c) 850 °C, (d) 900
°C an d (e) 950 °C. The pressures and absorbed microwave powers u sed
to achieve these tem peratures are listed in Table 6.2 located above Figure
6 .10. The other deposition conditions are: H 2 flow rate = 200 seem, CH4
flow rate = 2 seem, CO 2 flow rate = 2 seem , an d deposition time = 4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
Figure 6.23 Effects of S ubstrate Tem perature on Film Morphology (I)
H2 = 200 seem, CH4 = 4 seem and
Ts = (a) 750, (b) 850 and (c) 950 °C
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
196
Figure 6.23 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
197
■
-y
(a)
f *
1 3 K U
■
X 1 5 , 0 0 0
Figure 6.24 Effects of S ubstrate Tem perature on Film Morphology (II)
H2 = 200 seem, CH4 = 2 seem, C 0 2 = 2 seem and
Ts = (a) 750, (b) 800, (c) 850, (d) 900 and (e) 950 °C
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
198
1 3 K O
Figure 6.24 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.24 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
hours.
As shown, the films deposited u n d e r these conditions are
crystalline films with slightly different average grain sizes. The average
grain size increases gradually as the su b strate tem perature is increased
from 750 °C to 950 °C.
6.4.3.3
Effects of CH4 Concentration
Film morphology vs. CH4 concentration is shown in Figure 6.25
where the SEM photographs of six films are shown. These films are
deposited u nder the following experimental conditions: H2 flow rate =
400 seem, pressure = 51 Torr, microwave power absorbed = 2.34 kW,
substrate tem perature ~ 900 °C, deposition time = 6 hours and CH4 flow
rate (a) 3, (b) 4, (c) 5, (d) 6 , (e) 6.5 and (f) 7 seem.
As shown, the average grain size of the films decreases gradually
as the CH4 flow rate is increased from 3 to 7 seem.
Figure 6.26 displays the SEM photographs of five films deposited
under the following experimental conditions: H2 gas flow = 400 seem,
pressure = 57 Torr, absorbed microwave power = 2.5 kW, substrate
tem perature ~ 950 °C, deposition time = 6 hours an d CH4 flow rate (a) 4,
(b) 4.5, (c) 5, (d) 5.5 and (e) 6 seem.
As shown, the average grain size of th e films decreases gradually
also a s the CH4 flow rate is increased from 4 to 6 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
201
Figure 6.25 Effects of CH4 Concentration on Film Morphology (I)
H2 = 400 seem, p = 51 Torr, P* = 2.34 kW, Ts - 900 °C, and
CH4 * (a) 3, (b) 4, (c) 5, (d) 6 , (e) 6.5 and (f) 7 seem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
Figure 6.25 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203
(e)
Figure 6.25 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.26 Effects of CH4 Concentration on Film Morphology (II)
H2 = 400 seem, p = 57 Torr. Pt = 2.5 kW, Ts - 950 °C, and
CH4 = (a) 4, (b) 4.5, (c) 5, (d) 5.5 and (e) 6 seem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.26 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
206
Figure 6.26 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
207
6.4.3.4
Effects of CO2 Concentration
Film morphology vs. CO2 concentration is shown in Figure 6.27
where th e SEM photographs of four films are shown. These films are
deposited u n d er the following experimental conditions: H2 flow rate =
400 seem, CH4 flow rate = 6 seem, p ressure = 51 Torr, absorbed
microwave power = 2.34 kW, substrate tem perature ~ 900 °C, deposition
tim e = 6 h o u rs and CO 2 flow rate (a) 0, (b) 1, (c) 2, and (d) 3 seem.
As shown, the average grain size of the films increases gradually as
the CO2 flow rate is increased from 0 to 3 seem.
Figure 6.28 displays the SEM photographs of four films deposited
u n d er the following experim ental conditions: H2 flow rate = 200 seem,
CH4 flow rate = 6 seem, pressure = 51 Torr, absorbed microwave power =
2.46 kW, deposition tim e = 6 hours and CO2 flow rate (a) 3, (b) 3.5, (c) 4,
an d (d) 4.5 seem.
As shown, the average grain size of th e films increases gradually
also as th e CO2 flow rate is increased from 3 to 4.5 seem.
6.4.3.5
Effects of Total Flow Rate
Film morphology vs. total flow rate is shown in Figure 6.29 where
the SEM photographs of five films are shown. These films are deposited
u n d er the following experimental conditions: CH4 to H2 volume ratio =
1.5%, pressure = 51 Torr, absorbed microwave power = 2.34 kW,
substrate tem perature ~ 900 °C, deposition time = 6 hours and H2 flow
rate (a) 50, (b) 100, (c) 200, (d) 300 and (e) 400 seem.
As shown, th e average grain size experiences first a slight increase
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
208
(a)
(b)
Figure 6.27 Effects of CO2 Concentration on Film Morphology (I)
H2 = 400 seem, CH4 = 6 seem, p = 51 Torr, Pt = 2.34 kW, Ts ~ 900 oC,
and CO2 = (a) 0, (b) 1, (c) 2 an d (d) 3 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 5 K U
X 1 0
, 0 0 0
Figure 6.27 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
210
(a)
Figure 6.28 Effects of CO 2 Concentration on Film Morphology (II)
H2 = 200 seem, CH4 = 6 seem, p = 51 Torr, Pt = 2.46 kW,
and CO2 = (a) 3, (b) 3.5, (c) 4 and (d) 4.5 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 3 K U
Figure 6.28 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
212
Figure 6.29 Effects of Total Flow Rate on Film Morphology (I)
CH4/H 2 = 1.5%, p = 51 Torr, Pt = 2.34 kW, Ts - 900 °C, and
H2 = (a) 50, fb) 100, (c) 200. (d) 300 and (e) 400 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213
Figure 6.29 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
214
1 SKU
Figure 6.29 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
215
when the H2 flow rate is increased from 50 to 100 seem and then a slight
gradual decrease as th e H2 flow rate is fu rth er increased toward 400
seem.
Figure 6.30 displays the SEM photographs of five films deposited
under th e following experimental conditions: CH4 to H2 volume ratio =
1.25%, p ressu re = 57 Torr, absorbed microwave power = 2.5 kW,
substrate tem perature ~ 950 °C, deposition time = 6 hours and H2 flow
rate (a) 50, (b) 100, (c) 200, (d) 300 and (e) 400 seem.
As shown, the average grain size experiences first an increase
when the H 2 flow rate is increased from 50 seem to 100 seem and th en a
slight gradual decrease as the H2 flow rate is further increased toward
400 seem.
6.4.4
Typical Ram an Spectrum
Figure 6.31 displays the Raman spectrum <76> of a diamond film
deposited on a 3 ” silicon wafer under the following experimental
conditions: H2 flow rate = 400 seem, CH4 flow rate = 4 seem, CO2 flow
rate = 2 seem, pressure = 50 Torr, absorbed microwave power = 2.61 kW
and su b strate tem perature ~ 927 °C. The sh arp peak near 1332 cm -1
indicates the excellent quality of the deposited diamond film.
6.4.5
Typical Uniformity Profile
The thickness profiles of a diamond film are displayed in Figures
6.32 and 6.33.<71> They were obtained by laser interference/reflection
technique. The diamond film was deposited on a 3" silicon wafer under
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
216
Figure 6.30 Effects of Total Flow Rate on Film Morphology (II)
CH 4/H 2 = 1.25%, p = 57 Torr, Pt = 2.5 kW, Ts - 950 °C, and
H2 = (a) 50, (b) 100, (c) 200 , (d) 300 and (e) 400 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
217
1 3 K U
X 1 0 i 0 0 0
Figure 6.30 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
218
(e)
Figure 6.30 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
i
12000.0
9 8 0 0 .0
7 6 0 0 .0
H2 = 400 seem
CH4 = 4 seem
CO2 = 2 seem
p = 50 Torr
Pt = 2.61 kW
Tc ~ 927 °C.
w 5 4 0 0 .0
219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iI
H
3 2 0 0 .0
1 0 6 4 .4
4 3 0 .0 0
7 0 0 .0 0
9 8 0 .0 0
1 2 6 0 .0
1 5 4 0 .0
1 8 1 0 .0
RECIPROCAL CM
Figure 6.31 Raman Spectrum of a 3" Diamond Film Deposited in MCPR7-3
■
3
2.5
in
c
o
o
2
2
-
.S
u
220
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
i
3.5
E
= 400 seem
CH<j = 4 seem
c o 2 = 2 seem
h2
I
p
pt
T
0.5
0
i
-30
-20
-10
0
10
Wafer Surface ( Millimeters)
5= 5 5 T o rr
= 2.67 kW
- 964 °C.
i
20
i
30
Figure 6.32 Cross-sectional Thickness Profile of a 3" Diamond Film
—
Figure 6.33
Three Dimensional Thickness Profile of a 3" Diamond Film
221
( suojorui) iujij
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
222
the following experimental conditions: H2 flow rate = 400 seem, CH4 flow
rate = 4 seem, CO2 flow rate = 2 seem, p ressu re = 55 Torr, absorbed
microwave power = 2.67 kW and substrate tem perature ~ 964 °C.
6.4.6
Sum m ary
From the param etric study of diamond film deposition on 3” silicon
wafers, the results can be summarized as follows:
(1)
There exists an optimum su b strate tem perature range for
high rate diamond film growth with a suitable gas composition. The
growth rate is gas composition dependent.
(2)
There exists a n optimum CH4 concentration range for high
rate diamond film growth at a suitable su b strate temperature.
(3)
Addition of CO2 to a H 2/CH 4 discharge tends to dilute some
of the reactive species in the discharge.
(4)
The growth rate increases w ith increasing total flow rate in
the region of 50 -200 seem and saturates in th e region of 300 - 400
seem, keeping gas composition and su b strate tem perature constant. This
suggests:
(i) a t the low flow rates of 50 - 200 seem, the deposition rate
is limited by chemically active species hitting the substrate,
(ii)
at the higher flow rates of 300 - 400 seem, some of the
input gas by-passes th e deposition process.
(5)
Carbon conversion efficiency decreases with increasing total
flow rate and decreasing gas residence time.
(6)
Film morphologies depend on su b strate temperature, gas
composition and total gas flow rate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
223
(7)
Within all th e CH4 volume concentrations used, the average
grain size of deposited films decreases as CH4 volume concentration is
increased.
(8)
Within all th e CO2 volume concentrations used, the average
grain size of deposited films increases as CO2 volume concentration is
increased.
(9)
Ram an spectrum exhibits the excellent characteristics of a
diamond film deposited.
(10)
Diamond film w ith excellent uniformity (better th a n 2%) h as
been deposited.
(11)
Higher deposition rates m ay require changing the gas flow
configuration.
(12)
The m axim um diamond film growth rate obtained on a 3"
silicon wafer is ~ 0.67 pm /hour. The experimental conditions are: H2 flow
rate = 400 seem, CH4 flow rate = 6 seem, absorbed microwave power =
2.34 kW, su b strate tem perature ~ 900 °C and pressure = 51 Torr.
6.5
6.5.1
Study of Experim ental Set-ups with 4" Silicon Wafers
Introduction
W hen diamond films are deposited on 4" silicon wafers in the
present reactor, the experim ental set-ups affect th e film growth rate,
morphology and uniformity. Using growth rate and uniformity as judging
criteria, a series of experim ents are conducted to obtain optimum
experimental set-ups for uniform diamond film deposition on 4” silicon
wafers. The following independent in p u t experimental param eters are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
224
considered: (1) cavity shell geometry, (2) quartz dome geometry, (3) wafer
holder material. (4) wafer location, (5) cavity excitation mode and (6)
seeding mixtures.
In th is study, diam ond films are deposited on 4" silicon wafers
using the forced flow configuration shown in Figure 3.16. Here the
supporting quartz tube (28) h as an inside diameter of 110 mm, an
outside diameter of 115 mm.
The 4 ” silicon wafers used in these experiments have the following
physical properties: prime grade, p-type, boron doped, <100> orientation,
24 - 36 ohm-cm in resistivity, 99.5 -100.5 mm in diameter, and 500 550 |im in thickness.
6.5.2
Cavity Shell Geometry
The two cavity shells shown in Figures 6.2 and 6.3 are used for
diamond film deposition on 4” silicon wafers. Diamond films were
deposited in the reactors employing the two cavity shells under the
following experimental conditions: H2 flow rate = 400 seem, CH 4 flow rate
= 4 seem, CO 2 flow rate = 2 seem, pressure = 47.5 Torr, absorbed
microwave power = 2.58 kW, quartz tube length = 5.09 cm, and
substrate holder m aterial = graphite. The two reactors exhibit the
following differences as displayed in Table 6.4.
Table 6.4
Comparison of Two Cavity Shells
Coupling Efficiency
Growth Rate (mg/hour)
Cavity Shell (I)
74.5%
10.6
Cavity Shell (II)
82.8%
9.1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
225
Hence, under otherwise the sam e experim ental conditions, the
microwave power coupling efficiency is higher b u t th e diamond film
growth rate is slower in th e reactor employing cavity shell (II). It was also
found th a t w hen diam ond films are deposited on 4 ” diam eter silicon
wafers, the discharge created in th e reactor employing cavity shell (II)
tends to float toward the top surface of the quartz dome, w hich caused
the quartz dome to be coated faster.
The reactor employing cavity shell (I) h as been found to be a
preferable choice for diam ond film deposition on 4 ” silicon wafers with
th e presen t experimental configuration. It is used in all the experiments
th a t follow.
6.5.3
Q uartz Dome Geometry
With the same drawing, quartz domes m ade by different
m anufacturers and those constructed at different tim es by th e same
m anufacturer present sm all differences in geometry. W hen diamond
films are deposited on 4” silicon wafers, these sm all differences in
geometry affect the optim um deposition conditions a n d the resulting film
morphology, uniformity, and growth rate, etc.
Two quartz domes (Qd#l and Qd#2) made by the sam e
m anufacture were used for diam ond film deposition on 4” silicon wafers
u n d er th e following experimental conditions: quartz tube length = 5.2
cm, su b strate holder m aterial = graphite, H 2 flow rate = 400 seem, CO2
flow rate = 2 seem, CH4 flow rate = 6 seem, pressure = 35 Torr, absorbed
microwave power = 2.15 kW and su b strate tem perature ~ 780 °C. The
deposited films show differences in both growth rate and morphology as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
226
displayed in Table 6.5 and Figure 6.34:
Table 6.5
Comparison of Two Q uartz Domes
Morphology
Growth Rate (mg/hour)
6.5.4
Q d#l
Figure 6.34 (a)
9.42
Qd#2
Figure 6.34 (b)
8.6
S ubstrate Location
S u b strate location affects the su b strate tem perature and the
resulting film properties, su ch as growth rate, morphology and
uniformity, etc. In the present experimental configuration, the substrate
location is varied by changing th e length of quartz tube (28).
I
Two experiments were conducted w ith quartz dome # 1(Qd# 1)
and quartz tubes of lengths of 5.09 cm and 5.2 cm. The experimental
conditions are: H2 flow rate = 400 seem, CH 4 flow rate = 6 seem, CO2
flow rate = 2 seem, pressure = 35 Torr, su b stra te holder material =
graphite, and absorbed microwave power = 2.12 kW
The effects of substrate location on su b strate tem perature Ts,
growth rate and uniformity are displayed in Table 6 .6 , where the
uniform ity is judged by visual inspection.
Table 6.6
S u b strate Location
5.09 cm
5.20 cm
Effects of S ubstrate Location (I)
Ts (°C)
790
776
Growth Rate
10.9 m g /h o u r
9.42 m g /h o u r
Uniformity
good
fair
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
227
Figure 6.34 Effects of Q uartz Dome Geometry on Film Morphology
H 2 = 400 seem, CO 2 = 2 seem, CH4 = 6 seem, p = 35 Torr
Pt - 2.15 kW, Ts - 780 °C. w ith (a) Q d#l and (b) Qd#2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
228
As shown, th e b etter choice of quartz tu b e lengths is 5.09 cm,
judging from both growth rate and uniformity.
II.
Three experim ents were conducted w ith quartz dome #2
(Qd#2) and three q uartz tu b es of lengths of 5.05, 5.09, and 5.2 cm. The
experimental conditions are as follows: H2 flow rate = 400 seem, CH4
flow rate = 6 seem, CO2 flow rate = 2 seem, p ressu re = 35 Torr, and
absorbed microwave power = 2.16 kW.
The effects of su b stra te location on su b stra te tem perature Ts,
growth rate and uniform ity are displayed in Table 6.7, where the
uniformity w as judged by visual inspection.
Table 6.7
Q uartz tube length
5.05 cm
5.09 cm
5.20 cm
Effects of S ubstrate Location (II)
Ts (°C)
791
784
777
Growth Rate
8.7 m g /h o u r
9.3 m g /h o u r
8.6 m g /h o u r
Uniformity
fair
good
fair
As shown, the b etter choice of the quartz tu b e lengths is again
5.09 cm, judging from both growth rate and uniformity.
5.09 cm long q uartz tu b e is m ost frequently u sed in the
experiments where diam ond films are deposited on 4” silicon wafers.
6.5.5
Cavity Mode Excitation
The microwave cavity applicator can be excited by TMqih modes for
diamond film deposition, where n is a positive integer. In the present
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
229
experimental reactor, th e excitation of TM qh m ode is not possible due to
the conflict between cavity sh o rt length and excitation probe length
requirem ents. TM0 i 2 a n d TM013 modes were investigated for their
potentials in diamond film deposition experim ents. W hen the cavity is
excited with TM012 mode, the n ear field effect from th e probe is slightly
present, which causes th e plasm a to be attracted slightly towards the top
of quartz dome. W hen th e cavity is excited w ith TM013 mode, the near
field effect is alm ost non-existent. Higher order modes, su ch as TM014
and TM015 modes, etc., were n o t used since microwave power loss on the
cavity walls is of concern.
The following two experim ents were conducted to quantitatively
com pare the effects of TM012 and TMq13 modes. The experimental
conditions were: quartz tu b e length = 5.09 cm, su b stra te m aterial =
graphite, H2 flow rate = 400 seem, CH4 flow rate = 6 seem, CO2 flow rate
= 2 seem, pressure = 35 Torr, and absorbed microwave power = 2.06 kW.
The effects of cavity m ode excitation on su b strate tem perature, Ts,
growth rate and uniform ity are displayed in Table 6 .8 , where the
uniform ity is judged by visual inspection.
Table 6.8
Excitation Mode
™012
™ 013
Effects of Cavity Mode Excitation
Ts (°C)
806
790
Growth Rate
12 m g /h o u r
10.8 m g /h o u r
Uniformity
good
better
As shown, the diam ond film deposited w ith TM013 mode is slightly
more uniform th a n th a t w ith TM012 mode. B ut th e average growth rate of
th e diamond film deposited w ith TM013 mode is slightly lower th an th a t
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
230
w ith TM012 mode.
Comparing to TM012 mode excitation. TM013 mode excitation has
th e following advantages: (1) there is nearly no n ea r field effect from the
excitation probe, w hich m eans th a t the quartz dome is coated slower
during the diamond film deposition process, an d (2 ) the deposited
diam ond film is more uniform. Hence, TM013 mode excitation is used in
m ost experiments described in th is chapter.
6.5.6
Seeding Density
In th e photo resist seeding method, different seeding densities
resu lt from different m ixtures of diamond powder, photo resist and photo
resist thinner. The effect of seeding density on diam ond film growth was
examined by depositing diamond films u n d er otherwise identical
experim ental conditions. The two seeding m ixtures examined are:
Seeding mixture #1:
(a) mix 250 mg 0.1 pm diamond powder
w ith 25 ml of photo resist, (b) ultrasonically mix th e m ixture for an hour,
(c) apply the mixture onto a silicon wafer, (d) spin th e wafer a t 4000 rpm
for 30 seconds, (e) bake the wafer a t 135 °C for 30 m inutes.
Seeding mixture #2:
(a) mix 250 mg 0.1 pm diamond powder
w ith 53 ml of photo resist and 20 ml of photo resist thinner, (b)
ultrasonically mix the m ixture for an hour, (c) apply the m ixture onto a
silicon wafer, (d) spin th e wafer a t 4000 rpm for 40 seconds, (e) bake the
wafer a t 135 °C for 30 m inutes.
The deposition conditions are: H2 flow rate = 400 seem, CH4 flow
ra te = 4 seem. CO2 flow rate = 2 seem, p ressure = 47.5 Torr, absorbed
microwave power = 2.58 kW, quartz tube length = 5.2 cm, su b strate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
231
holder m aterial = graphite and su b strate tem perature - 860 °C.
The average growth rates are sim ilar w ith 9.46 m g /h o u r for
seeding m ixture #1 an d 9.8 m g /h o u r for seeding m ixture # 2 .
6.5.7
Sum m ary
From the study of experim ental set-ups for diamond film
deposition on 4” silicon wafers, the resu lts can be summarized as
follows, in diamond film deposition on 4” silicon wafers,
(1)
Cavity shell geometry influences diamond film growth. Cavity
shell (I) is a preferable choice.
(2)
Variation in quartz dome geometry influences diamond film
growth rate and morphology.
(3)
S ubstrate holder m aterial influences discharge geometry and
su b stra te heating.
(4)
S ubstrate location influences film growth rate and
uniformity. 5.09 cm long quartz tu b e is m ost frequently used su b strate
su p p o rt tube.
(5)
Cavity excitation modes influence film growth rate an d
uniformity. TM0i 3 mode is a preferable choice.
(6)
Diamond film growth is n o t sensitive to slight variation in
seeding density.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
232
6.6
6 .6 .1
D iam ond Film Deposition on 4" Silicon Wafers
Introduction
D iam ond films w ith various physical properties are deposited on 4"
silicon w afers u n d e r various experim ental conditions. In order to
optimize the diam ond film deposition process to achieve the objective of
depositing diam ond films with high growth rate an d good uniform ity
(judged by visual inspection), it is useful to stu d y th e co-relation between
the physical properties of the deposited diam ond films and th e
experim ental conditions. Hence a param etric stu d y of diamond film
deposition on 4 ” silicon wafers w as conducted. In th is study, th e
following experim ental param eters are considered:
I
The independent in p u t experim ental param eters:
( 1)
absorbed microwave power, Pt,
(2)
pressure, p,
(3)
(4)
II.
total flow rate, ft.
The in tern al experimental param eters:
(1)
(2)
III.
gas (H2, CH4, CO2) composition, and
su b strate tem perature Ts, and
gas residence time, tr.
The o u tp u t experimental param eters:
( 1)
(2)
(3)
film linear growth rate (pm /hour),
film weight gain u (m g/hour),
film growth efficiencies:
(i)
growth rate vs. power flux, i.e., u vs. S,
(ii)
gas flow efficiency: k* (mg/liter) = u /f t,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
233
(ili)
(4)
carbon conversion efficiency, k s (%).
film morphology,
(5)
typical film uniformity, and
(6)
typical film R am an spectrum .
In th is study, diamond films are deposited on 4” silicon wafers
using the forced flow configuration displayed in Figure 3.16. The
supporting quartz tube (28) h a s a n inside diam eter of 110 mm, an
outside diam eter of 115 mm an d a length of 50.9 mm. The physical
properties of the 4” silicon wafers are th e sam e as th a t described in
section 6.5.1.
Description of relevant experim ental chemistry space is also
included whenever possible.
6.6.2
Film Growth Rate and Growth Efficiency
6.6.2.1
Effects of S ubstrate Tem perature
Figure 6.35 displays linear growth rate and weight gain vs.
su bstrate tem perature and power flux u n d er the following experimental
conditions: H2 flow rate = 400 seem, CH4 flow rate = 6 seem, and CO2
flow rate = 2 seem. The pressures an d absorbed microwave powers used
to achieve these tem peratures are listed in Table 6.9 located above Figure
6.35.
As shown, the average linear growth rate and weight gain are
increased a s the substrate tem perature is increased from 813 °C to 900
°C. B ut the film uniformity (from visual inspection) decreases as the
substrate tem perature is increased from 813 °C to 900 °C.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
234
Table 6.9
Substrate Temperature Conditions (4")
Tem perature, Ts (°C)
900
836
825
813
Pressure, p (Torr)
45
40
40
37.5
Power, Pt (kW)
2.43
2.28
2.24
2.2
Power Flux, S (W/cm2)
28.6 29.1 29.6
0.6
^
o3
1
1
1
31.6
1
H2 = 400 seem
CH4 = 6 seem
C02 = 2 seem
4" silicon wafer
0.55
16.22
■14.86
B
£
ja
0.5 ■
|
I
0.45 m
O
■13.51 |
f
■12.16 O
2
£
■10.81 &
0.4 ■
0.35 ______ 1___ 1__ 1_________ _________ 1___
813 825 836
900
9.46
Substrate Temperature, Ts (°C)
Figure 6.35 Linear Growth Rate and Weight Gain versus
S ubstrate Tem perature and Power Flux
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
235
Figure 6.36 displays carbon conversion efficiency and gas flow
efficiency versus substrate tem perature an d power flux. The carbon
conversion efficiency and gas flow efficiency are increased as substrate
tem perature is increased from 813 °C to 900 °C.
In th e present reactor, there exists a compromise between the two
objectives, i.e., high growth rate and good uniformity, in diamond film
deposition on 4" silicon wafers. When a good uniformity is required, the
upper limit on growth rate comes from th e upper limit on the substrate
tem perature. The substrate tem perature is primarily limited by the safety
operation tem perature limit and the coating speed of the quartz dome. As
is shown in Figure 6.35, higher growth rate is achieved with higher
su bstrate tem perature. To improve the film uniformity in the higher
sub strate tem perature case, more absorbed microwave power is required.
The h o t discharge is th en in close contact w ith the quartz dome, heating
the q uartz dome up to its unsafe operation tem perature region and
coating the quartz dome at the same tim e. A compromise range for the
su bstrate tem perature to serve both objectives is between 800 °C to 850
°C.
In th e growth rate vs. gas composition and total flow rate
experim ents th a t follow, the substrate tem perature is chosen to be fixed
at 845 °C. The CH4 concentration is varied first in th e growth rate vs. gas
composition experiments.
6.6.2.2
Effects of CH4 Concentration
Figure 6.37 displays the film linear growth rate, weight gain and
carbon conversion efficiency vs. CH4 gas concentration under the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
236
Power Flux, S (W /cm2)
£
g
•PN
5.5
28.6 29.1 29.6
H2 = 400 seem
CH4 = 6 seem
CO2 = 2 seem
4" silicon wafer
0.58
0
Cd
a
e
5
0.52
4.5
0.47
4
0.42
o1
a
a
o
*§
o
813 825 836
Gas Flow Efficiency (mg/liter)
6
900
Substrate Temperature, Ts (°C)
Figure 6.36 Carbon Conversion Efficiency and Gas Flow Efficiency
versus S ubstrate Temperature and Power Flux
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
237
£
6*
S
hH
0
%
Ed
.2
03
u
1
o
o
g 35
la
O
5.5
6.5
s
12.15
0.45
5
H2 = 400 seem
C 09 = 3 seem
Ts ~ 845 °C
p = 40 Torr
Pt = 2.24 kW
4" silicon wafer
0.4
2
o
0.35
5.5
6
6.5
10.8
9.45
CH4 Flow Rate (seem)
(b)
Figure 6.37 Linear Growth Rate, Weight G ain and Carbon
Conversion Efficiency versus CH4 concentration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
(a)
238
following experimental conditions: H 2 flow rate = 400 seem. CO2 flow rate
= 3 seem, pressure = 40 Torr, absorbed microwave power = 2.24 kW,
su b strate tem perature - 845 °C and deposition time = 8 hours.
As shown, the film linear growth rate and weight gain experience
peak values w ith CH4 to H2 volume ratio of about 6 .5 /4 0 0 = 1.625%. B ut
the film uniformity (judged by visual inspection) in this case is not as
good as th a t in the case where the CH4 to H2 volume ratio is 6 /4 0 0 =
1.5%. The following experiments of growth rate versus CO2 concentration
sta rt from the gas composition of th e later case. Also shown in Figure
6.37 is th a t the carbon conversion efficiency increases initially w hen CH4
flow rate is increased from 5 to 5.5 seem an d decreases w hen CH4 flow
rate is further increased. The location of these gas compositions on the
B achm ann C-H-O phase diagram is show n in Figure 6.38.
6.6.2.3
Effects of CO2 Concentration
Figure 6.39 displays the film linear growth rate, weight gain and
carbon conversion efficiency versus CO2 gas concentration under the
following experimental conditions: H2 flow rate = 400 seem, CH4 flow rate
= 6 seem, pressure = 40 Torr, absorbed microwave power = 2.24 kW,
su b strate tem perature ~ 845 °C and deposition time = 8 hours.
As shown, the film linear growth rate and weight gain experience
peak values with CO2 to H2 volume ratio of about 2 .5 /4 0 0 = 0.625%. B ut
the film uniformity (judged by visual inspection) in this case is n ot as
good a s th a t in the case where the CO2 to H 2 volume ratio is 3 /4 0 0 =
0.75%. The following experiments of growth rate vs. total flow rate sta rt
from th e gas composition of the later case. Also shown in Figure 6.39 is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in
o
n
N
oI oo
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.38
Location of Gas Compositions on the C-H-O Phase Diagram
(IV)
239
240
vP
p^r
0
I
45
&
W
C
2
2
I
°
CJ
-5 C
3.5
fi
0
1
O
3
3
2
(a)
3
12.15
0.45
Pt = 2.24 kW
4" silicon wafer
s
t>
0.4
10.8
0.35
9.45
2
o
C 02 Flow Rate (seem)
Figure 6.39 Linear Growth Rate, Weight Gain and Carbon
Conversion Efficiency versus C 0 2 concentration
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weight Gain (mg/hour)
seem
241
th a t th e carbon conversion efficiency experiences a peak value a t the
peak growth rate point an d decreases as the CO2 flow rate is further
increased. The location of these gas compositions on the B achm ann C-HO phase diagram is show n in Figure 6.40.
6.6.2.4
Effects of Total Flow Rate
Figure 6.41 displays the film linear growth rate, weight gain and
carbon conversion efficiency versus total gas flow rate and gas residence
time u n d er the following experimental conditions: CH 4 to H2 volume ratio
= 1.5%, CO2 to H 2 volume ration = 0.75%, pressure = 40 Torr, absorbed
microwave power = 2.24 kW, su b strate tem perature - 845 °C and
deposition time = 8 hours.
As shown, the film linear growth rate and weight gain initially
experience an increase as the H2 flow rate is increased from 100 to 300
seem. They experience a slight decline as the H2 flow rate is further
increased from 300 to 400 seem. Also show n in Figure 6.41 is th a t the
carbon conversion efficiency decreases w ith increasing total gas flow rate
and decreasing gas residence time.
6.6.3
Film Morphology
6.6.3.1
Introduction
Scanning Electron Microscopy (SEM) is used to identify the
morphologies of films deposited u n d er various experimental conditions.
The films deposited u n d er the various experimental conditions described
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
«>th
* - CM
o o
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.40
CM
Location of Gas Compositions on the C-H-0 Phase Diagram
(V)
242
243
12.15
s 0.45
%
£
I
Ts ~ 845 °C
p = 40 Torr
CH4/H2 = 1.5%
C02/H2 = 0.75%
Pt = 2.24 kW
0.4
2
o
0.35
100
200
300
400
100
200
300
400
10.8
Weight Gain (mg/hour)
Residence Time, t,. (seconds)
20
10
6.7
9.45
5*
s
•H
CJ
ea
w
c
o
2
1
o
o
C2
o
%
O
H2 Flow Rate (seem)
(b)
Figure 6.41
(a) Linear Growth Rate, Weight G ain and (b) Carbon
Conversion Efficiency versus Total Flow Rate and
Gas Residence Time
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
244
in section 6.6.2 are characterized.
6.6.3.2
Effects of S ubstrate Temperature
Figure 6.42 displays the SEM photographs of four films deposited
w ith su b strate tem peratures: (a) 900 °C, (b) 836 °C, (c) 825 °C, and (d)
813 °C. The p ressures and absorbed microwave powers used to achieve
these tem peratures are listed in Table 6.9 located above Figure 6.35. The
other deposition conditions are: H2 flow rate = 400 seem, CH4 flow rate =
6 seem, CO2 flow rate = 2 seem, and deposition tim e = 8 hours.
As shown, the films deposited u n d er these conditions are
crystalline films with different average grain sizes. The film deposited a t
900 °C h as the largest average grain size and th e film deposited a t 813
°C h as the sm allest average grain size.
6.6.3.3
Effects of CH4 Concentration
Film morphology versus CH4 concentration is shown in Figure
6.43 where th e SEM photographs of five films are displayed. These films
are deposited u nder the following experimental conditions: H2 flow rate =
400 seem, CO2 flow rate = 3 seem, pressure = 40 Torr, absorbed
microwave power = 2.24 kW, substrate tem perature ~ 845 °C, deposition
time = 8 hours and CH 4 flow rate (a) 7, (b) 6.5, (c) 6 , (d) 5.5, and (e) 5.
As shown, the film deposited with 7 seem CH4 gas flow h as very
sm all average grain size. The average grain size ju m p s up when th e CH 4
flow rate is reduced from 7 to 6.5 seem. It th en decreases gradually a s
th e CH4 flow rate is fu rth er reduced from 6.5 to 5 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.42 Effects of Substrate Temperature on Film Morphology
H2 = 400 seem, CH4 = 6 seem, CO2 = 2 seem, an d
Ts = (a) 900, (b) 836, (c) 825 and (d) 813 °C.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
246
(c)
1 3 K U
Figure 6.42 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
247
(a)
Figure 6.43 Effects of CH4 Concentration on Film Morphology
H 2 = 4 0 0 seem, CO2 = 3 seem, p = 40 Torr, Pt = 2.24 kW, Ts = 845 °C,
an d CH4 = (a) 7, fb) 6.5, (c) 6 , (d) 5.5, (e) 5 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6.43 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 3K U
Figure 6.43 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
250
6.6.3.4
Effects of CO2 Concentration
Film morphology versus CO2 concentration is shown in Figure
6.44 where th e SEM photographs of five films are displayed. These films
are deposited under th e following experim ental conditions: H2 flow rate =
400 seem, CH 4 flow rate = 6 seem, p ressu re = 40 Torr, absorbed
microwave power absorbed = 2.24 kW, su b strate tem perature - 845 °C,
deposition tim e = 8 hours and CO2 flow rate (a) 1, (b) 2, (c) 2.5, (d) 3 and
(e) 4 seem.
As shown, the films deposited w ith 1 and 2 seem CO2 gas flow
have very sm all average grain size. The average grain size ju m p s u p w hen
CO2 flow rate is increased to 2.5 seem. It th en experiences a gradual
decline as th e CO2 flow rate is further increased to 4 seem.
6.6.3.5
Effects of Total Flow Rate
Film morphology versus total flow ra te is show n in Figure 6.45
where th e SEM photographs of four films are displayed. These films are
deposited u n d er the following experim ental conditions: CH4 to H2 volume
ratio * 1.5%, CO2 to H2 volume ratio = 0.75%, pressure = 40 Torr,
absorbed microwave power = 2.24 kW, su b strate tem perature ~ 845 °C,
deposition tim e = 8 hours and H2 flow rate (a) 400, (b) 300, (c) 200, and
(d) 100 seem.
As shown, the average grain size stays about the same w hen the
H2 flow rate is decreased from 400 to 300 seem. It th en experiences a
gradual decline as the H2 flow rate is fu rth er decreased from 300 to 100
seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
251
Figure 6.44 Effects of CO2 Concentration on Film Morphology
H2 = 400 seem, CH4 = 6 seem, p = 40 Torr, Pt = 2.24 kW, Ts - 845 °C,
an d C 0 2 - (a) 1, (b) 2 , (e) 2.5, (d) 3 and (e) 4 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A —
--
* J'fysrI
-.
-■
rr
j/
i
■-' v ■ V
V
^
>
*
-
*a
v
’
/fc \
***>' A,/"'
:k
,*>
>
‘ ^ r
r
^
\
✓
—
A 1 * 7
" > .;W '.
/
W
1 S KU
X 10 , 0 0 0
1 Hm
.
-■
1 17501
Figure 6.44 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
253
Figure 6.44 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
254
13KU
Figure 6.45 Effects of Total Flow Rate on Film Morphology
CH4 / H 2 = 1.5%, C 0 2 / H 2 = 0.75%. p = 40 Torr, Pt = 2.24 kW,
Ts ~ 845 °C, and H2 = (a) 400, (b) 300, (c) 200 a n d (d) 100 seem.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
255
1 SKU
15KU
Figure 6.45 (cont’d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
256
6.6.4
Typical R am an Spectrum
Figure 6.46 displays the R am an spectrum <76> of a diamond film
deposited on a 4” silicon wafer u nder the following experimental
conditions: H 2 flow rate = 400 seem, CH4 flow rate = 6 seem, CO2 flow
rate = 3 seem , p ressure = 40 Torr, absorbed microwave power = 2.24 kW
and su b stra te tem perature ~ 845 °C. The sh a rp peak n ear 1332 cm ' 1
indicates th e excellent quality of the deposited diamond film.
6.6.5
Typical Uniformity Profile
The thickness profiles of a diamond film deposited on a 4” silicon
wafer are displayed in Figures 6.47 and 6.48.<71> They were obtained by
laser interference/reflection technique. The diam ond film was deposited
und er the following experimental conditions: H 2 flow rate = 400 seem,
CH4 flow rate = 6 seem, CO2 flow rate = 3 seem, pressure = 40 Torr,
absorbed microwave power = 2.24 kW and su b stra te tem perature ~ 845
°C.
6 .6.6
Sum m ary
From th e param etric study of diamond film deposition on 4 ” silicon
wafers,th e resu lts can be summarized as follows:
(1)
There exists a compromising su b stra te tem perature range
for diam ond film deposition with high growth rate an d good uniformity.
When a good uniform ity is required, the u p p er lim it on film growth rate
is prim arily th e safety operation tem perature lim it and the coating speed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
i
1 1 7 0 0 .0
H2 = 400 seem
CH4 = 6 seem
C 0 2 = 3 seem
p = 40 Torr
Pt = 2.24 kW
T* ~ 845 °C.
9 6 0 0 .0
7 5 0 0 .0
UJ
5 4 0 0 .0
257
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
i
3 3 0 0 .0
1 2 7 4 .0
4 3 0 .0 0
7 0 0 .0 0
9 8 0 .0 0
1 2 6 0 .0
1 5 4 0 .0
1 8 1 0 .0
RECIPROCAL CM
Figure 6 .4 6
Ram an Spectrum of a 4" Diam ond Film D eposited in MCPR7-3
i
t
i
<
C/5>
u
s
tA
A
4X
>
J9
o
258
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
i
s
e
£
H 2 = 4 0 0 seem
CH 4 = 6 seem
C 0 2 = 3 seem
p = 4 0 T o rr
Pt = 2 .2 4 kW
T s ~ 8 4 5 °C.
i
-40
-20
20
0
Wafer Surface ( Millimeters )
i
40
Figure 6.47 Cross-sectional Thickness Profile of a 4" Diamond Film
Film ( microns)
259
H2 = 400 seem
CH4 - 6 seem
c q 2 = 3 seem
D s 40 Torr
p1 tI * 2 .2. 4— kW
~ 845 °c .
Wafer Surface ( m m )
Profile of a 4" Diamond Film
Figure 6.48 Three Dimensional Thickness
260
of th e quart? dome.
(2)
There exists an optimum CH4 concentration range for high
rate diam ond film growth at a suitable su b strate tem perature.
(3)
Addition of CO2 to H 2/CH 4 discharge tends to dilute some of
the reactive species in the discharge.
(4)
The growth rate increases w ith increasing total flow rate in
the region of 100 -200 seem and satu rates in th e region of 300 - 400
seem, keeping gas composition and su b strate tem perature constant. This
suggests:
(i)
a t the low flow rates of 100 - 200 seem, the deposition
rate is limited by chemically active species hitting the substrate.
(ii)
at th e higher flow rate of 300 - 400 seem, some of the
input gas by-passes the deposition process.
(5)
Carbon conversion efficiency decreases with increasing total
flow ra te and decreasing gas residence time.
(6)
Film morphologies depend on su b strate temperature, gas
com position an d total gas flow rate.
(7)
Ram an spectrum exhibits the excellent characteristics of a
diam ond film deposited.
(8)
Diamond film with excellent uniform ity (better th an 2%) h as
been deposited.
(9)
Diamond film with excellent uniform ity and growth rate is
deposited on a 4" silicon wafer u nder th e following experimental
conditions: H2 flow rate = 400 seem, CH4 flow rate = 6 seem, CO2 flow
rate - 3 seem, absorbed microwave power = 2.24 kW, substrate
tem perature ~ 845 °C, pressure = 40 Torr and growth rate ~ 0.43 p m /
hour.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER SEVEN
CONCLUSIONS AND RECOMMENDATIONS
7.1
Sum m ary of Significant Results
7.1.1
Introduction
The research and development described in this dissertation have
lead to the successful creation of a working prototype reactor (MCPR7-3)
of a n improved microwave cavity plasm a reactor concept an d a U. S.
p aten t application<17>. Diamond films with uniformities better th a n 2%
have been deposited on 3 ” and 4” silicon wafers with th is prototype
reactor. The linear growth rates obtained are 0.89 |xm /hour on 2” silicon
wafers, 0.67 p m /h o u r on 3" silicon wafers and 0.43 p m /h o u r on 4”
silicon wafers. The perform ance “figures of m erit” of diam ond film
deposition reactors were developed to compare the MCPRs and other
reactors. Three generations (MCPR7-1, MCPR7-2 and MCPR7-3) of
microwave cavity plasm a reactors have been investigated a n d /o r
developed for diamond film deposition over larger su b strate areas. Three
generations (MCJR-1, MCJR-2 and MCJR-3) of microwave cavity je t
reactors have also been developed a n d /o r designed to deposit diamond
films. These reactors have th e potential to deposit diamond films with
higher growth rates. A m ethod to obtain the operational characteristics of
261
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
262
the MCPRs w as developed. The diamond films deposited in the MCPR7-3
prototype reactor were investigated w ith respect to their growth rate and
quality as the in p u t experimental param eters were varied. The electric
fields in the microwave cavity p lasm a reactor (MCPR7-1) were m easured
to develop a better understanding of the electromagnetic field/plasm a
interactions during diamond film deposition process. These im portant
accom plishm ents are sum m arized in greater detail in the sections th a t
follow.
7.1.2
MCPR Development
An improved microwave cavity plasm a reactor was developed. This
reactor u ses the microwave plasm a technology th a t w as developed a t
MSU in earlier investigations. T hat is, the new reactor still employs single
mode excitation and two internal independent tuning variables, i.e.,
cavity short and excitation probe length, are used to minimize the
reflected microwave power from the cavity applicator.
Improvements over earlier microwave diamond film reactors are as
follows:
(1)
The excitation probe is placed on the sliding short end of the
cavity applicator instead of on the side of th e cavity. Near field effects are
eliminated w ith the use of higher order axial modes. Only TM modes are
excited and th e interference from nearby TE modes is eliminated. TM012
and TM013 modes are u sed for diamond film deposition. The reactor
operation is m ore reliable and the reactor is suitable for industrial
applications.
(2)
A resonance breaker w as placed in the applicator to create a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
263
stable discharge above th e su b strate surface.
(3)
A larger (5") diam eter quartz dome is u sed and discharges u p
to 12.5 cm in diam eter are created.
(4)
Higher power in p u t (from 1.5 kW to 4.5 kW or more) capacity
is achieved w ith th is reactor.
(5)
An improved internal gas flow configuration w as developed.
Reactive gases are forced to flow close to th e su b stra tes and the gas
efficiencies are improved.
This reactor h as been used to deposit diam ond films in the 20 to
95 Torr p ressu re range. It can be used to deposit uniform diamond films
on su b strate surfaces u p to 10 cm in diameter, even w ithout substrate
heating or cooling. Diamond film uniform ity b etter th a n 2% over 10 cm
diam eter silicon wafer h a s been achieved.
The typical experimental conditions for diam ond film deposition on
3" and 4” silicon wafers are a s follows:
I.
Diamond film deposition on 3” silicon wafers:
p ressu re = 51 Torr,
H 2 flow rate = 400 seem, CH4 flow rate = 6 seem,
absorbed microwave power = 2.34 kW,
su b strate tem perature - 900 °C,
weight gain = 10.6 m g /h o u r an d linear growth rate = 0.67
(im /hour.
II.
Diamond film deposition on 4" silicon wafers:
pressu re = 40 Torr,
H2 flow rate = 400 seem, CH4 flow ra te = 6 seem, CO2 flow
rate = 3 seem,
absorbed microwave power = 2.24 kW,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
264
su b strate tem perature ~ 845 °C,
weight gain = 11.7 m g /h o u r and linear growth rate = 0.43
pm /hour.
7.1.3
Microwave Plasm a J e t Reactor Development
7.1.3.1
Microwave Cavity J e t Reactors
Three generations of microwave cavity je t reactors have been
developed for high rate diamond film deposition.
The first two generation reactors were investigated for diamond
film deposition. One inch diameter quartz tubes were u sed as the plasm a
confinem ent chamber. The operating pressure an d in p u t microwave
power were limited due to quartz tube over-heating. The total gas flow
rate w as limited by the pum ping speed of the vacu u m system.
The third generation reactor was designed in which the discharge
confinem ent cham ber w as enlarged to be a dome. This design should
overcome th e confinement cham ber over-heating problem. It was not
bu ilt du e to lack of funding.
7.1.3.2
J e t Configuration in MCPR
Prelim inaiy investigation of diamond film deposition using a je t
concept w as conducted in the new MCPR7-3. Increased deposition rate (~
1 pm /h o u r) over a 1.5 cm diam eter area was obtained w ithout substrate
cooling. It is expected th a t higher deposition rates are possible with
proper su b strate cooling.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
265
7.1.4
Reactor Operational Characteristics
A m ethod to experimentally characterize a microwave plasm a
reactor was developed. It w as described in sections 6.2 an d 6.3 of this
dissertation. This method probably can also be used to experimentally
investigate other plasm a reactors. It can be used to locate th e required
experim ental conditions w hen diam ond films with desired properties,
su ch a s film thickness, morphology, and uniformity, etc. are deposited.
W hen the discharge occupies only p a rt of th e quartz dome volume,
th e following conclusions were reached from the stu d y of operational
characteristics of MCPR:
(1)
Substrate tem perature, Ts, is primarily determ ined by
operating pressure. S ubstrate tem perature Ts increases sharply with
increasing pressure (~ 6 °C/Torr) and a t constant pressure, Ts increases
gradually (~ 0.02 °C/W) w ith increasing absorbed microwave power. Ts is
n o t sensitive to variation in gas composition and total flow rate.
(2)
For a constant in p u t power, the discharge volume, Vj,
decreases w ith increasing p ressure. At a constant pressure, V j increases
w ith increasing absorbed microwave power.
is n o t sensitive to
variations in gas composition an d total flow rate.
(3)
Average discharge power density, w, increases w ith
increasing pressure an d a t co n stan t pressure, it is not sensitive to
variation in absorbed microwave power. T hat is, u n d er a co n stan t
pressure, a s the absorbed microwave power increases, the discharge
volume also increases, resulting in only small variations in absorbed
power density. Power density is n ot sensitive to variations in gas
composition and total flow rate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
266
(4)
Power flux S to the su b strate increases with increasing
absorbed microwave power. It is n o t sensitive to variations in gas
composition an d total flow rate.
7.1.5
Diamond Film Deposition on 3" Silicon Wafers
A param etric stu d y of diamond film deposition on 3” silicon wafers
w as conducted. The results of this study can be summarized as follows:
(1)
There exists a n optimum su b strate tem perature range (~ 800
- 1000 °C) for high rate diamond film growth with a suitable gas
composition. The growth rate is gas composition dependent.
(2)
There exists a n optimum CH 4 concentration range for high
rate diamond film growth at a suitable su b strate tem perature.
(3)
Addition of CO2 to H2/C H 4 discharge tends to dilute some of
th e reactive species in the discharge.
(4)
The growth rate increases w ith increasing total flow rate in
the region of 50 -200 seem and satu rates in th e region of 300 - 400
seem, keeping gas composition and su b strate tem perature constant. This
suggests:
(i) a t the low flow rates of 50 - 200 seem, the deposition rate
is limited by chemically active species hitting the substrate,
(ii)
at the higher flow rates of 300 - 400 seem, some of the
in p u t gas by-passes the deposition process.
(5)
Carbon conversion efficiency decreases w ith increasing total
flow ra te and decreasing gas residence time.
(6)
Film morphologies depend on su b strate tem perature, gas
composition an d total gas flow rate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
267
(7)
W ithin all th e CH4 concentrations used, the average grain
size of deposited films decreases as CH4 concentration is increased.
(8)
W ithin all th e C 0 2 concentrations used, th e average grain
size of deposited films increases as C 0 2 concentration is increased.
(9)
R am an spectrum exhibits th e excellent characteristics of a
diamond film deposited.
(10)
Diamond film with excellent uniform ity (better th a n 2%) h as
been deposited.
(11)
Higher deposition rates m ay require changing the gas flow
configuration.
(12)
The maxim um diamond film growth rate obtained on a 3"
silicon wafer is ~ 0.67 |im /hour. The experimental conditions are: H 2 flow
rate = 400 seem, CH4 flow rate = 6 seem, absorbed microwave power =
2.34 kW, su b strate tem perature - 900 °C an d p ressu re = 51 Torr.
7.1.6
Diamond Film Deposition on 4" Silicon Wafers
7.1. 6 .1
Study of Experimental Set-ups
A stu d y of experimental set-ups for diamond film deposition on 4 ”
silicon wafers w as conducted, the results of th is stu d y can be
sum m arized as follows:
(1)
A slight variation in cavity shell geometry influences
diamond film growth. Cavity shell (I) is a preferable choice.
(2)
A slight variation in quartz dome geometry influences
diamond film growth rate and morphology.
(3)
S u b strate holder material influences discharge geometry and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
268
su bstrate heating. G raphite is a preferable su b stra te holder material.
(4)
S ubstrate location influences film growth rate and
uniformity. 5.09 cm long quartz tube is m ost frequently used substrate
support tube.
(5)
Cavity excitation modes influence film growth rate and
uniformity. TMq13 mode is a preferable choice.
(6)
Diamond film growth is not sensitive to slight variations in
seeding density.
7.1.6.2
Diamond Film Deposition on 4” Silicon Wafers
A param etric stu d y of diamond film deposition on 4" silicon wafers
was conducted, the resu lts of this study can be sum m arized as follows:
(1)
There exists a compromising su b strate tem perature range
for diamond film deposition with high growth rate and good uniformity.
When a good uniformity is required, the u pper lim it on film growth rate
is primarily the safe operating tem perature limit an d the coating rate on
the quartz dome.
(2)
There exists an optimum CH4 concentration range for high
rate diamond film growth a t a suitable su b strate tem perature.
(3)
Addition of CO2 to H2/CH 4 discharge tends to dilute some of
the reactive species in the discharge.
(4)
The growth rate increases w ith increasing total flow rate in
the region of 100 -200 seem and saturates in th e region of 300 - 400
1
seem, keeping gas composition and substrate tem perature constant. This
suggests:
(i) at th e low flow rates of 100 - 200 seem, the deposition
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
269
rate is limited by chemically active species hitting the substrate.
(ii)
at the higher flow rate of 300 - 400 seem, some of the
in p u t gas by-passes the deposition process.
(5)
Carbon conversion efficiency decreases w ith increasing total
flow rate and decreasing gas residence time.
(6)
Film morphologies depend on substrate tem perature, gas
composition and total gas flow rate.
(7)
Raman spectrum exhibits the excellent characteristics of a
diam ond film deposited.
(8)
Diamond film w ith excellent uniformity (better th a n 2%) has
been deposited.
(9)
Diamond film w ith excellent uniformity and growth rate is
deposited on a 4" silicon w afer u n d er the following experimental
conditions: H2 flow rate = 400 seem, CH4 flow rate = 6 seem, CO2 flow
rate = 3 seem, absorbed microwave power = 2.24 kW, substrate
tem perature - 845 °C, p ressu re = 40 Torr and growth rate ~ 0.43 pm /
hour.
7.1.7
Reactor Performance “Figures of Merit”
7.1.7.1
Performance “Figures of Merit" of MCPR
Performance “figures of merit" were developed for comparison
between different group of reactors and for com parison of the sam e
reactor under different operating conditions.
I.
The MCPRs have the following performance “figures of merit"
a s listed in Table 7.1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
270
Table 7.1 Sum m ary of "Figures of Merit" of
Diamond Film Deposition Experim ents in MCPR
Weight
gain
(mg/hr)
Energy
effic.
(m g/
kW-hr)
Gas
flow
effic.
(mg/
liter)
Carbon
conv.
effic.
(%)
2.254.5
4.59
0.190.38
3.57
2 .8 -
5.6
1.753.5
0.150.3
2.95.8
20
6.2
3.3
0.26
2.4
0.67
45
10.6
4.5
0.44
5.5
WT#29
0.65
45
10.3
4.4
0.57
7.1
WT#30
0.62
45
9.75
4.2
0.81
10.1
WT#31
0.51
45
8
3.4
1.3
16.6
W F#75
0.43
78
11.7
5.2
0.49
4
W F#84
0.44
78
11.9
5.3
0.66
5.5
W F#85
0.42
78
11.4
5.1
0.95
7.9
WF #86
0.37
78
9.9
4.4
1.65
13.7
Linear
growth
rate
Oim/hr)
Dep.
area
(cm2)
MCPR
7-1
0.4-0.8
16
MCPR
7-2
0.4-0.8
20
MCPR
7-3
NT#4
0.89
WT#27
Exp.
No.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
271
II.
Comparing the experiments conducted in the MCPRs, the
following conclusions can be reached:
(1)
With th e sam e deposition area, th e diam ond films deposited
in MCPR7-3 have th e highest linear growth rate an d weight gain.
(2)
Among the uniform diamond films deposited in MCPR7-3, as
the deposition area increases, the linear growth ra te decreases and the
weight gain increases.
7.1.7.2
Com parison w ith other Reactors
Comparing w ith other reactor concepts reviewed in chapter 2 and
referring to Table 2.1, th e following conclusions can be reached:
(1)
The linear growth rate in MCPR is lower th a n th a t in m ost
high p ressu re (higher th a n 100 Torr) reactors.
(2)
The uniform diamond film deposition a re a (78 cm2) in MCPR
is larger th a n th a t in th e other reactors reported in the literature.
(3)
The weight gain of the diam ond films deposited in MCPR is
higher th a n th a t in the other low pressure (less th a n 100 Torr) reactors.
It is lower th a n th a t in some high pressure (more th a n 100 Torr) reactor.
(4)
The energy efficiency of MCPR is com parable to m ost other
diamond film deposition reactors.
(5)
The gas flow efficiency of MCPR is higher th a n m ost high
pressu re diamond film deposition reactors.
(6)
The carbon conversion efficiency of MCPR is higher th a n th a t
of other diam ond film deposition reactors.
(7)
The excellent uniformity (better th a n 2%) of the diamond
films deposited in MCPR is better th a n th a t in th e other reactors reported
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
272
in the literature.
7.1.8
M easurem ent of Electric Fields in MCPR
The (relative) spatial variations of the exciting electromagnetic field
pattern s in MCPR7-1 u n d er diamond film deposition conditions were
m easured. These spatial electric field m easurem ents and intensities are
related to other im portant experimental param eters su ch a s gas mixture,
flow rate, input power, su b strate tem perature, an d discharge pressure,
etc.
The following conclusions were reached from th is study, u nder
diamond film deposition condition in MCPR7-1:
(1)
The excitation mode w as identified to be discharge loaded
TMq h mode.
(2)
Maximum electric field intensity in the cavity was ~ 150 V /
cm.
(2)
Discharge loaded cavity quality factor w as - 60.
(3)
The tangential component of the E field is the m ain
discharge excitation field.
(4)
Average power density in the discharge w as ~ 6 W /cm 3 a t 20
Torr and 20 W /cm 3 a t 70 Torr.
(5)
Under a co n stan t pressure, the square of the cavity electric
field strength is approximately proportional to the power in p u t and only
slightly influenced by the variation in flow rates.
(6)
At a co n stan t gas flow, the square of th e cavity electric field
strength is approximately proportional to the power in p u t and only
slightly influenced by th e variation in pressure.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
273
(7)
Under a co n stan t pressure, the su b strate tem perature
increases with the power in p u t (~ 0.4 °C/W) and is only slightly
influenced by the variation in flow rates.
(8)
At a constant gas flow, the su b strate tem perature increases
w ith th e power input (~ 0.4 °C/W) under each working pressure. Also, at
th e constant gas flow an d in p u t power, the su b strate tem perature
increases (~ 3.5 °C/Torr) w ith increasing pressure.
7.2
Recommendations for Future Research
(1)
For diam ond film deposition in MCPR7-3, higher growth
rates m ay be achieved by using higher p ressure (100 - 200 Torr), higher
power ( 3 - 6 kW) and su b stra te cooling. A su b strate cooling stage has
been built. Different gas flow configurations should also be investigated
for their potential to improve the diamond film growth rates.
(2)
The MCPR concept m ay be scaled up to larger diam eter (14 -
20 inch) cavities by dropping the microwave in p u t power frequency to
915 MHz. Larger area su b strates may be covered w ith increasing input
microwave power. Also, w ith proper reactor design, the diam eter of the
cavity operated with 2.45 GHz power may be enlarged to 8-12 inches.
(3)
The improved microwave cavity je t reactor (MCJR-3) should
be built and investigated for its potential in diam ond film deposition with
higher growth rates.
(4)
The je t configuration in MCPR7-3 offers an interesting
alternative approach to deposit diamond films w ith higher growth rates.
Experim ents should be conducted to investigate th is potential, using a
su b stra te cooling system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
274
(5)
In order to obtain a better understanding of the
electromagnetic iield/plasm a interactions, electric field m easurem ents in
MCPR7-3 should be conducted. A new cavity w ith electric field
m easurem ent probe holes h as been built.
(6)
In order to improve the knowledge of th e fundamental
plasm a/chem ical reactions of diamond film deposition and to be able to
better control the diamond film deposition process, it is necessary to
identify the discharge species essential for diam ond film deposition. It is
helpful to characterize the microwave discharge u n d er diamond film
deposition conditions by emission spectroscopy, actinom etiy and laser
induced fluorescence (LIF) spectroscopy. The knowledge obtained in this
investigation will be beneficial to achieve single ciystal diamond film
growth which is essential for some industrial applications, such as
diamond electronic and optical devices.
(7)
For some industrial applications, it is necessaiy to deposit
diamond film on su b strates a t low substrate tem perature. The research
on diamond film deposition a t lower su b strate tem peratures a n d /o r low
pressures should be conducted.
(8)
For different industrial applications, diamond films with
different morphologies are required. Since only a preliminary
investigation of diamond film morphologies versus experimental
conditions was conducted, more detailed investigation of diamond film
morphologies versus experimental conditions should be conducted.
(9)
Alternative gas compositions, su ch as carbon monoxide (CO),
oxygen (O2), acetone (CH3COCH3) and acetylene (C2H2), etc., should be
considered to improve the gas flow efficiency and diamond film growth
rate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF REFERENCES
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF REFERENCES
[1]
S. Tennant, “On the n atu re of the diam ond,” Phil. Trans. Roy.
Soc. London, 87, 123 (1797).
[2]
J.C . Angus an d C.C. Hayman, Science 241, 913 (1988).
[3]
R. Orr, "The im pact of th in film diamond on advanced
engineering systems," presented a t th e global business and
technical outlook for high performance inorganic thin film
coatings. Gorham advanced m aterials institute, Monterey,
California, October 30 - November 1, 1988.
[41
F. P. Bundy, H. T. Hall, H. M. Strong, and R. H. Wentof, J r.,
“M an-made diam ond,” Nature, 176, 51 (1955).
[51
W. G. Eversole, U. S. Patents 3030187 and 3030188.
[61
J . C. Angus, H. A. Will, an d W. S. Stanko, J . Appl. Phys. 39,
2915 (1968).
[71
B. V. Deiyaguin, D. V. Fedoseev, V. M. Lukyanovich, B. V.
Spitsyn, V. A. Ryabov, and A. V. Lavrentyev, J . Ciyst. Growth 2,
380 (1968).
[81
B. V. Spitsyn, L. L. Bouilov and B. V. Deryagin, J . Cryst.
Growth 52. 219 (1981).
[91
S. M atsumoto, Y. Sato, M. Kamo, N. Setaka, Jp n . J . Appl. Phys.
21, p art 2, 183 (1982).
[101
J . Rogers, “Properties of steady-state, high pressure argon
discharges,” Ph.D. D issertation, Michigan State University,
275
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
276
1982.
[11]
S. J . Whitehall*. “Experimental development of a microwave
electrotherm al th ru ster,” Ph.D. D issertation, Michigan State
University (1986).
[12]
M. Dahimene, “Development of a microwave ion and plasm a
source immersed in a m ulticusp electron cyclotron resonant
m agnetic held,” Ph.D. Dissertation, Michigan State University
(1987).
[13]
J . A. Hopwood, “Macroscopic properties of a m ultipolar electron
cyclotron resonance microwave cavity plasm a source for
anisotropic silicon etching,” Ph.D. D issertation, Michigan State
University (1990).
[14]
J . Zhang, B. Huang, D. K. Reinhard and J . A smussen, J . Vac.
Sci. Technol. A 8 . 2124 (1990).
[15]
J . Zhang, G. King, T. Grotjohn, an d J . Asmussen, “Diagnostic
m easurem ents of a resonant cavity microwave plasm a diamond
deposition reactor,” presented at Third International Conference
on th e New Diamond Science and Technology, Heidelberg,
Germany, A ugust 31 - Septem ber 4, 1992.
[16]
P. Mak, G. King, M. Ulczynski, J . Zhang, T. Grotjohn, an d J .
A sm ussen, “Experimental diagnosis of low pressure microwave
discharges during diamond th in film deposition,” presented a t
Third International Conference on th e New Diamond Science
an d Technology, Heidelberg, Germany, A ugust 31 - Septem ber
4. 1992.
[17]
J. A sm ussen an d J . Zhang, "Improved ap p aratu s for th e coating
of m aterial on a substrate using a microwave or UHF plasma,"
U. S. Patent Application, (May 28, 1992).
[18]
S. M atsumoto, Y. Sato, M. Kamo an d N. Setaka, Jp n . J . Appl.
Phys. 21, 183 (1982).
[19]
S. M atsumoto, Y. Sato, M. T sutsum i and N. Setaka, J . Mater.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
277
Sci., 17, 3106 (1982).
[20]
Y. H. Lee. P. D. Richard, K. J . Bachm ann, and J . T. Glass, Appl.
Phys. Lett. 56, 620 (1990).
[21]
K. Baba, Y. Aikawa, and N. Shobata, J . Appl. Phys. 69, 7313
(1991).
[22]
A. Masood, “Technology of electronic properities of CVD
diamond him m icrosensors for therm al signals,” Ph.D.
dissertation, Michigan S tate University (1992).
[23]
K. Suzuki, A. Sawabe, H. Y asuda and T. Inuzuka, Appl. Phys.
Lett. 50, 728 (1987).
[24]
K. Suzuki, A. Sawabe, an d T. Inuzuka, Jp n . J . Appl. Phys. 29,
153 (1990).
[25]
S. Matsumoto, I. Hosoya an d T. C hounan, Jp n . J . Appl. Phys.
29, 2082 (1990).
[26]
K. V. Ravi, C. A. Koch. H. S. Hu, an d A. Joshi, J . Mater. Res., 5,
2356 (1990).
[27]
L. M. Hanssen, W. A. Carrington, J . E. Butler and K. A. Snail,
Mater. Lett. 7, 289 (1988).
[28]
S. Matsumoto, J . Mater. Sci. Lett., 4, 600 (1985).
[29]
S. Matsumoto, M. Hino and T. Kobayashi, Appl. Phys. Lett., 51,
737 (1987).
[30]
M. Kamo, Y. Sato, S. M atsumoto and N. Setaka, J . Cryst.
Growth 62, 642 (1983).
[31]
H. Kawarada, K. S. M ar an d A. Hiraki, Jp n . J . Appl. Phys. 26,
L1032 (1987).
[32]
J . J . Chang, T. D. Mantel, R. V uppuladhadium an d H. E.
Jackson, Appl. Phys. Lett. 59, 1170 (1991).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
278
[33]
T. D. Mantei, Third Diamond W orkshop, Detroit, MI, March,
1992.
[34]
Y. Liou, A. Inspektor, R. Weimer an d R. Messier, Appl. Phys.
Lett. 55, 631 (1989).
[35]
P. K. Bachm ann, W. Drawl, D. Knight, R. Weimer, and R. F.
Messier, in Extended A bstracts Diamond and Diamond-like
M aterials Synthesis, edited by G. Jo h n so n , A. Badzian, and M.
Geis, M aterials Research Society, Pittsburgh, PA, 1988, p. 99.
[36]
P. Bachm ann, D. Leers and H. Lydtin, Diamond and Related
M aterials, 1, 1 (1991).
[37]
Y. M itsuda, T. Yoshida and K. Akashi, Rev. Sci. Instrum . 60,
249 (1989).
[38]
K. Takeuchi and T. Yoshida, J . Appl. Phys. 71. 2636 (1992).
[39]
L. Mahoney, ‘T h e design an d testing of a compact electron
cyclotron resonant microwave-cavity ion source," M.S. Thesis,
Michigan State University (1989).
[40]
J . T. Salbert, "Anodic growth an d cathodic removal of silicon
dioxide layers utilizing an electron cyclotron resonant
microwave plasm a disk reactor,” Ph.D. Dissertation, Michigan
S tate University (1992)
[41]
The com puter monitor system w as designed and built by Z. M.
Rum m ler w ith help from M Ennis. The monior program was
w ritten by Z. M. Rummler a t M ichigan State University, E ast
Lansing, Michigan.
[42]
J . A sm ussen, J . Root, "Ion generating apparatus and method
for th e use thereof,” U. S. P atent No. 4,507,588, M arch 26,
1985.
[43]
J . A sm ussen, S. Nakanishi, J . Malinkey an d S. W hitehair,
“Microwave electrothruster," p a te n t applied for by NASA-Lewis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
279
[44]
J . Asm ussen, D. K. Reinhard, “Method for treating a surface
w ith a microwave plasm a and improved apparatus,” U. S.
P atent No. 4,585,688, April 29, 1986.
[45]
J . Asm ussen, D. K. Reinhard, “Microwave or UHF plasm a and
improved apparatus,” U. S. Patent No. 4,630,566, Dec. 23,
1986.
[46]
T. Ropple, D. K. Reinhard and J . A sm ussen, “Dual plasm a
microwave app aratu s an d m ethod for treating a surface,” U. S.
Patent No. 4,691,662, Sept. 8 , 1987.
[47]
J . Asm ussen, D. K. Reinhard an d M. Dahimene, “Improved
plasm a generating apparatus u sin g m agnets and m ethod,” U. S.
P atent No. 4,727, 293, Feb. 23, 1988.
[48]
J . A sm ussen, “Method for treating a m aterial using
radiofrequency waves," U. S. P atent No. 4,777,336, Oct.
11,1988.
[49]
J . A sm ussen, “Improved microwave ap p aratu s,” U. S. Patent
No. 4,792,772, Dec. 20, 1988; also Canadian Letters Patent No.
1,287,666, Aug. 13, 1991.
[50]
J . A sm ussen, D. K. Reinhard, “Improved plasm a reactor
ap p aratu s and method for treating a su b strate,” U. S. Patent
No. 4,943,345, Ju ly 24, 1990.
[51]
J . Root and J . Asmussen, Rev. Sci. Instrum ., 56, 154 (1985).
[52]
T. Roppel, D. K. Reinhard. and J . A sm ussen, J . Vac. Sci.
Technol., B4, 295 (1986)
[53]
J . A sm ussen and M. Dahimene, J . Vac. Sci. Technol., B5, 328
(1987).
[54]
J . A sm ussen, J . Vac. Sci. Technol., A7, 883 (1989).
[55]
L. Mahoney, M. Dahimene, and J . A sm ussen, Rev. Sci.
Instrum ., 59, 448 (1988).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
280
[56]
S. W hitehalr, J . A sm ussen and S. Nakanishi, J . Propul. Power,
3, 136 (1987).
[57]
This figure w as provided by J . Asmussen, Michigan State
University, E ast Lansing, Michigan.
[58]
B. Huang, “Electrical properties and physical characteristics of
polycrystalline diamond films deposited in a microwave plasm a
disk reactor," Ph. D. Dissertation, Michigan S tate University
(1992).
[59]
Designed an d b u ilt by W. Richardson in 1990 a t Michigan State
University, E ast Lansing, Michigan.
[60]
Designed and b u ilt by Union Carbide Coatings Service Co.,
Cleveland, OH.
[61]
Developed by W avemat Inc., Plymouth, Michigan.
[62]
Designed and b u ilt by F. C. Sze at Michigan State University,
E ast Lansing, Michigan.
[63]
J. Asmussen, R. Mallavarpu, J . R. Hamann, an d H. C. Park,
“The design of a microwave plasm a cavity", Proc. IEEE, 109
(1974).
[64]
R. Mallavarpu, M. C. Hawley and J . A sm ussen, “Behavior of a
microwave cavity discharge over a wide range of pressures and
flow rates," IEEE Tran, on Plasma Science, PS-6, 341-354
(December, 1978).
[65]
R. F. Harrington, Time-Harmonic Electromagnetic Fields
(McGraw-Hill, New York, 1961)
[66]
R. L. Kapitza, Sov. Phys. JETP 30, 973 (1970).
[67]
L. M. Baltin, V. M. Batemin, I. I. Devyatkim, V. P. Lebedeva, N.
I. Tsemko, High Temp. 9, 1019 (1971).
[68]
Y. Arata, S. Miyake, A. Kubayashi, and S. Takuchi, Jp n . J .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
281
Phys. 40, 1456 (1976).
[69]
R. Chapm an, J . Filpus, T. Morin, R. SneUenberger, J .
Asm ussen, M. Hawley, and R. Kerber, J . Spacecraft 19, 579
(1982).
[70]
J . A sm ussen, H. H. Lin, B. Manring, and R. Fritz, Rev. Sci.
Instrum . 58, 1497 (1987).
[71]
Film uniform ity m easurem ents by laser interference/reflection
technique were conducted by T. T heissen a t Michigan State
University, E ast Lansing. Michigan.
[72]
A. Masood, M. Aslam, M. A. Tamor and T. J . Potter, J .
Electrochem. Soc. 138, L67 (1991)
[73]
A. Masood, private com munication, Michigan State University,
E ast Lansing, Michigan.
[74]
Amplex, 16 Britton drive, bloomfield, Connecticut 06002
[75]
Shipley company, 91 B artlett Street, Marlboro, MA 02162
[76]
The R am a n sp ectra were provided by Scott M artin a t Norton
Diamond Film, Northboro, MA.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
9 892 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа