A STUDY OF GaAs,
. InP,
GROWN BY ORGANOMETALLIC
Fred
Institute,
1977
Center,
1984
submitted
to the faculty
Center
in partial
fulfillment
requirements
for
the degree
Doctor
EPITAXY
R. Bacher
B.S. Worcester Polytechnic
M.E.S. Oregon Graduate
A dissertation
Graduate
and InGaAs
VAPOR PHASE
of
Philosophy
in
Applied
June,
Physics
1987
of
the Oregon
of the
The dissertation "A Study of GaAs, InP, and InGaAs grown
by organometallic
vapor phase epitaxy" by Fred R. Bacher
has been examined and approved by the following Examination
Committee:
J.-S~- Bla~~more:
Professor
Thesis
Advisor
and
Wallace B. Leigh, Assiitant Professor
Nicholas G. Efor, Professor
Jack S. Sachitano, Adjunct Assistant
Professor (Tektronix, Inc.)
ii
ACKNOWLEDGEMENTS
The author
contributions
and Wallace
throughout.
crystal
Leigh,
whose
Especially
advice
persons
Foremost
and assistance
for their
are John Blakemore
were
invaluable
Dr. Leigh, who taught me so much
about
growth.
relied
I am especially
Anderton
to the following
to this dissertation.
This work
Inc.
is indebted
and Jack
George
contributed
Howard
heavily
grateful
support
for the ongoing
from Tektronix,
commitment
of Harry
Sachitano.
ably performed
numerous
useful
source
many
of the crystal
growths,
and
Corinne,
who
suggestions.
A final note of gratitude
was a constant
on financial
is sounded
of guidance
Hi
to my wife,
and wisdom.
TABLE OF CONTENTS
ABSTRACT.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
CHAPTER
I:
CHAPTER II:
CHAPTER III:
INTRODUCTION.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
CRYSTAL GROWTHMETHOD
5
CHARACTERIZATIONTECHNIQUES
A. Photoluminescence(PL)
.18
B. Hall Mobility
.26
C. Auger Electron
D. Secondary
Ion
Spectroscopy
(AES)
Mass Spectrometry
32
(SIMS)
35
E. Optical Absorption
CHAPTER IV:
37
SUBSTRATE AND UNOOPEDEPILAYER CHARACTERIZATIONRESULTS
A. GaAs and InP Substrates
49
B. GaAs and InP Undoped Epilayers
55
CHAPTER V:
CHAPTERVI:
CHAPTERVII:
InP: Mg EPILAYERS
In
1 -x
69
Ga As
76
x
CONCLUSION.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103
REFERENCES.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105
APPENDIX A:
OHMIC CONTACT PREPARATION
115
APPENDIX B:
EPILAYER THICKNESS MEASUREMENT
117
APPENDIX C:
77 K PL Linewidth
Raw Data from the
versus Net
Literature
APPENDIX D:
RAWSIMS DATA FOR InGaAs/InP
APPENDIX E:
List
APPENDIX F:
COMPUTERPROGRAMS
of
Chemicals
Carrier
(Table
Concentration:
XIII)
LAYERS
and Equipment
(Table
119
120
XIV)
125
128
VITA
..
iv
.141
LIST OF FIGURES
FIG. 1.
InGaAsP Bandgap at 300 K
FIG. 2.
OMVPE
Reactor General Layout.
FIG.
3.
OMVPE
Reactor
FIG.
4.
OMVPE Reactor
Cabinet.
FIG.
5.
OMVPE
Tube.
FIG.
6.
Schematic
FIG.
7.
Light
FIG. 8.
Reactor
of
Output
3
Gas Schematic.
. . . . . . . . . . . . . . . . . . .
. . . . . . . . .11
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
PhotoluminescenceApparatus
19
for Various MonochromatorSlit Widths
2l
Optical Detector and Filter Spectral Response
22
FIG. 9. PL FWHM for Various Doping Concentrations
24
FIG. 10. PL for OMVPE InGaAs
25
FIG. 11. Schematicof Hall MobilityMeasurementApparatus
27
FIG. 12.
30
Electromagnet Power Supply Calibration
FIG. 13. PL Spectrum for InO.53GaO.47As"Standard"
33
FIG. 14. AES Profile for InO.53GaO.47As"Standard"
34
FIG.
15.
Methods of Mounting Epilayersfor SubstrateEtching
38
FIG.
16.
Optical AbsorptionApparatus
40
FIG.
17.
InO.53GaO.47AsRefractiveIndex
42
FIG. 18.
Schematicfor Epoxy RefractiveIndex Measurement
43
FIG. 19.
Raw TransmissionData for Optical Absorption
47
FIG. 20.
PL for GaAs
FIG.
21.
Substrates.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
PL for N-type InP Substrates............................53
FIG. 22.
PL for P-type InP Substrates............................54
FIG. 23.
OMVPE GaAs Growth Rate
6l
v
FIG. 24. Hall Mobility for GaAs at CryogenicTemperatures
62
FIG. 25. Hall Mobility for OMVPE InP
65
FIG. 26. PL for OMVPE GaAs
66
FIG. 27. PL for Undoped OMVPE InP
68
FIG. 28. PL for OMVPE InP:Mg
72
FIG.
29.
FIG.
30.
PL for
FIG.
31.
Variation
Net Acceptor
Concentration
in InP:Mg versus [CPZMg]/
[TMln] Rat io. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
InGaAs 10 at 77 K
of InGaAs
78
Defect
PL with
Excitation
Light
Inten.sity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80
Variation
of In -x Ga xAs PL with
1
FIG.
32.
FIG.
33.
FIG.
34.
FIG.
35.
AES Profile
FIG.
36.
Absorption
FIG.
37.
Bandgap
PL for
InO.53GaO.47As
at
PL for
InO.53GaO.47As
at
Ra
t io.
for
InGaAs
85
77 K
87
300 K
88
65
Coefficient
Estimates
x
90
for
In1_xGaxAs
at
10 K
91
for OMVPE InGaAs versus [TMIn]/[TMGa]
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
92
FIG. 38.
Near-bandgap Absorption Coefficient for InO.53GaO.47As..95
FIG. 39.
AbsorptionCoefficientfor InO.53GaO.47As at 300 K......96
FIG. 40.
FIG. 41.
FIG. 42.
FIG. 43.
FIG. 44.
FIG. 45.
FIG. 46.
FIG. 47.
FIG. 48.
Absorption
Ga As at 300 K....98
Coefficientfor OMVPE In1_x x
Absorption Coefficient
Hall
Mobility
for InO.53GaO.47As at 77 K.......99
for OMVPE
InO.53GaO.47As.................101
InO.53GaO.47As...............121
Cs Beam SIMS Data for MBE
o Beam SIMS Data for MBE
InO.53GaO.47As................122
Cs Beam SIMS Data for OMVPE
InO.53GaO.47As.............123
InO.53GaO.47As..............124
Flowchartfor Hall Mobility MeasurementProgram........129
o Beam SIMS Data for OMVPE
Flowchart for Absorption Coefficient Calculation Program
...136
vi
LIST OF TABLES
TABLE
I.
Comparison
of Atmospheric
Pressure
OMVPE
Growth
8
Conditionsfor InP Using TMIn
TABLE
II.
Comparison
of Atmospheric
Pressure
OMVPE
Growth
Conditionsfor InGaAs Using TMIn and TMGa
TABLE
III.
Comparison
of OGC Hall Mobility
9
Results
with Other Labs
TABLE IV.
TABLE V.
TABLE
8...31
Proportional Error Estimates for Optical Absorption
44
Substratesand OMVPE Growth Numbering
VI.
Hall Mobility
and Resistivity
for OMVPE
50
InP with
Various Total Gas Flow Rates
56
TABLE VII. Hall Mobility for OMVPE GaAs
TABLE
VIII.
Undoped
Result
s.
OMVPE
InP Electrical
57
Characterization
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
64
TABLE IX. Growth Conditionsand PL FWHM for OMVPE InP:Mg
71
TABLE X. SIMS Results for MBE and OMVPE InO.53GaO.47As
82
TABLE
XI.
TABLE
XII.
PL Peak
energies
Hall Mobility
for OMVPE
and Carrier
In l-xGaxAs
Concentration
84
for
In Ga As/lnP
l-x x
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
TABLE
XIII.
77 K PL FWHM versus
Net Carrier
100
Concentration:
Raw Data From the Literature
119
TABLE XIV. List of Chemicalsand Equipment
125
TABLE XV.
130
TABLE XVI.
Hall Mobility Measurement and Calculation Program
Optical Absorption Coefficient Calculation program...137
vii
ABSTRACT
Presented
a study
based
used
of the semiconductors
organometallic
reactor.
parameters
were
obtained
pressure
optical
with
doped
system
(PL) peak full width
InP is reported
devices
of magnesium
here.
cussed.
junctions
mechanisms
because
growth
beyond
of InGaAs
An empirical
of the parasitic
zone.
scribed.
epitaxy
and
including
the
GaAs and InP epilayers
of 6480 and 4820
at half maximum
require
cm2/Vs,
values
and
of 12
at 77 K suggest
A study of Mgalternative
depletes
grown
Mg incoracceptor.
is dis-
in the present
trimethylindium
lattice-matched
For these, Hall mobilities of 9000
viii
several
compositions
the extent
to
PL peaks
substitutional
of various
documents
epilayers
InP and InGaAs,
of its low diffusivity.
epilayers
which
p-type
an attractive
the simple Mg1n
study
reaction
InGaAs
lasers
here,
as a dopant.
Mg offers
seen at 1.0 and 1.3 eV from InP:Mg
The growth
phase
were
at 77 K.
use has been made
poration
vapor
is described
150+ epilayers.
some optoelectronic
zinc for abrupt
The methyl-
particularly
300 K Hall mobilities
and 10 meV respectively
little
of
and trimethylgallium
organometallic
applications,
for growing
photoluminescence
While
trimethylindium
The OGC OMVPE
used
InP, and InGaAs.
results
Evaluation of this growth technique was conducted
an eye toward
photodetectors.
and characterization
GaAs,
sources
in an atmospheric
(OMVPE)
with
here are the growth
prior
system
to the
to InP are de-
and 26,500 cm2/Vs at 300
K and 77 K respectively
The optical
energy
range
were achieved.
absorption
for In
coefficient
was mP~red
1-xGa xAs with compositions
:h>etweenx = 0.45 and
x = 0.51, at 10, 77, and 300 K.
The absorption
abruptly
the bandgap~
to 30,000
scarcity
to 6000 cm-l just
cm-l at 1.5 eVe
of published
An intrinsic
emission
near
this defect,
and these
Motivation
optical
InGaAs
0.7 eV.
above
ov<er a broad
~d
coeffic2ent
increases
for this work
a~sorptian
rises
gradually
was provided
by the
rlata f~r InGaAs.
defect was found by 1'1..
at 71 ~~ giving
A variety
results
of methods
were used tu characterize
are pre£ehted.
ix
1
INTRODUCTION
The motivation
properties
for this work
of InGaAs.
an established
to reproduce
was to study the growth
Organometallic
technique
for GaAs,
vapor
phase
and material
epitaxy
and so the first
(OMVPE)
efforts
results
seen
in other
labs and to characterize
ment.
InP epilayers
were
required
as a prelude
growth
because
ternary
growth.
InP, GaAs/GaAs,
compounds
tronic
devices
InGaAs/GaAs,
perfection.
samples,
The OMVPE
desirable
communications
using
materials,
described
its reproducibility
and efficiency
a popular
the propagation
technique
method
InP:Mg/
III-V
speed
elec-
and Hall
(PL)
purity
and
to compare
techniques.
in Chapter
relative
II, is
structures
ease.
in growing
for producers
with
high quality
Based
fibers
on
GaAs layers,
of optoelectronic
communication
a
It is the
for GaAs photocathodes.
loss in optical
for the
Characterization
used
Multilayer
can be grown with
manufacturing
on high
of material
in detail
of its versatility.
InGaAs/lnP
and other
photoluminescence
and growth
the equip-
of InP/lnP,
These
are commonly
here were
good surface
networks.
give an indication
technique,
because
it is becoming
a very
to those working
These methods
of compounds
established
Since
interest
and optical
both of which
different
provide
and InGaAs/InP.
layers was performed
mobility,
crystal
layers
OMVPE was used to grow thin layers
are of intense
of these
variety
InP buffer
to OMVPE
is
devices.
is mini-
2
mized
at - 1.55 ~
desirable
wavelengths,
indium-containing
than GaAs for some applications.
elements
In, Ga, As, P in single
depicted
in Fig. 1.
the GaAs bandgap.
easily
Indium
is added
than molecular
of the two methods
of each
[1].
growth
depends
on chemistry
difficult
to understand.
the utility
of OMVPE
as necessary
Many
type dopant
are desired
studied.
compared
developed
used
of OMVPE
dynamics
with
more
hybrid
some of
is that
the
that are
is to examine
GaAs and InP growth
require
which
[2].
While
p-type
doped
used zinc.
Magnesium
can be important
when
for a later
herein.
devices
study.
InP layers
The pto
of a low
or interfaces
have been grown
by any method
New incorporation
here,
in contrast
abrupt junctions
grown
or pn
InP is investigated
has the advantage
At OGC, eight Mg-doped
of InP:Mg
regions
was to build
for part of this work was magnesium,
to the results
level
to retain
of this report
of InGaAs,
are cited
mechanisms
and
and
are
for Mg in InP.
As a part of the InGaAs/lnP
deep
(MBE), although
and fluid
InGaAs must wait
The few reports
postulated
below
epitaxy
disadvantage
devices
epilayers.
and p-type
commonly
diffusivity,
response
are generally
A long term goal of this research
doping
the more
for growth
bandgaps
compounds
beam
So the focus
semiconductor
with n and p doped
n-type
A major
of the four
preliminaries.
useful
junctions.
to GaAs to obtain
are being
the advantages
process
Solid mixtures
are more
form have the energy
Phosphorus-containing
grown by OMVPE
combinations
crystal
materials
defect was observed
growth
characerization,
in PL data and extensively
an unexpected
analyzed.
3
~
2.0eV
o
o
rot)
I- 1.5eV
ex
Q..
ex
C>
C
Z
I.OeV
ex
CD
>-
/InP
(x= O,y= I,
0= 5.87A)
C> 0.5eV
a::
I.LJ
Z
I.LJ
O.OeV
InAs
(x=O,y=O,
0= 6.06A)
x = 0.5
GaAs
(x=I,y=O,
o=5.65A)
COMPOSITION (x.y) AND
LATTICE CONSTANT (0)
FIG. 1.
InGaAsP Bandgap at 300 K.
GaP
(x=I,y=1
a=5.45A)
4
Accurate
absorption
temperature
and composition
data are compared
by liquid
phase
Chapter
techniques
measurement
VI.
epitaxy
InGaAs
is discussed
grown
Topics
over a wide
published
when
report
reports
as a function
energy
of
range.
for material
reported
on the various
characterization.
appropriate.
in this study
are presented
on include:
absorption
ion mass
The theory
The results
These
grown
experimental
behind
each
and discussion
in Chapters
IV
PL of GaAs and InP substrates;
of GaAs and InP undoped
(AES), and secondary
layers.
obtained
(LPE).
used for epilayer
and PL, Hall mobility,
ternary
of
to the several
PL and Hall mobility
copy
data were
III and the appendixes
of the crystals
through
coefficient
epilayers;
coefficient,
spectrometry
Auger
PL of InP:Mg;
electron
(SIMS),
of In
spectros-
Ga As
l-x x
5
CHAPTERII
CRYSTAL GROWTHMETHOD
Within
the family
of vapor
phase
epitaxial
techniques, organometallic vapor phase epitaxy
crystal
growth
is an important
(OMVPE)
is the elevated
method both for industry and research [3].
OMVPE
temperature
on a substrate
growth
the same lattice
of single
constant,
for source materials.
substrates
layer
with
using
organometallic
Interestingly,
lattice
constants
layers
growth
quite
compounds
with nearly
and hydrides
can also occur on
dissimilar
to the desired
[4].
This
applied
chapter
to GaAs,
and its history,
Other
crystal
growth
will
outline
the lattice-matched
InP, and InGaAs.
the reader
processes
For further
growth method
information
as
on OMVPE
is referred
to two recent
reviews
that are related
yet distinct,
namely
[5,6].
low
pressure OMVPE [7], chloride/hydride VPE [8], adduct OMVPE [9], vapor
levitation epitaxy [10], and laser-induced chemical vapor deposition
[11], will not be further
InP and InGaAs were
discussed
here.
among the first materials
to be grown by
Manasevit in the early OMVPE investigations [12,13], dating to 1973.
The increased
use of this technique
can be seen in the proceedings
three International Conferences on OMVPE:
and in 1986
[4].
The early attempts
in 1981 [14],
to grow InGaAs
1984
on GaAs
[15],
of
6
[16-19],
pheric
or on InP
pressure
reaction
growing
[20-25], using
OMVPE met with
in the vapor
particularly
and Bass
zone of TEIn
(TMIn)
The reaction
x(CH3hGa
using
producing
InP is:
ing the growth
over a wide
temperature
ranged
The reactions
[42].
above
range,
to 700°C
[26].
for In-containing
are poorly
understood
of Pz over InP is in excess
up to 100:1
is a loss of TMIn prior
less severe
there
reagent
[5], and are
is a significant
temperatures.
are used.
must
to the growth
Arsenic
loss from InGaAs
zone due to reaction
loss of TEIn,
[44].
reagent
be maintained.
still be maintained.
than the corresponding
The
of 10-4" T at 650°C
must
growth
compounds.
of the phosphorous-containing
to the indium-containing
not as bad, but an overpressure
While
First,
optimum
minimiz-
In this study,
from InP at elevated
an overpressure
although
For some applications,
is desirable
from 600°C
given
temperature
of phosphorous
[TMIn] ratios
[33-39]
In l-xGa xAs + 3CH4.
evaporation
compared
the use
TMIn and TMGa.
mechanisms.
Consequently,
with
of Stringfellow
due to several
pressure
reported
success
at OGC was aimed
inefficient
vapor
consistent
is:
have been reported
temperatures
The effort
+ (I-x) (CH3) 3In + AsH3 -+
can take place
temperatures
has been
co-workers,
reaction
for InGaAs
compounds
More
the growth methods
[41-43] and their
(TEIn) by atmos-
due to the parasitic
and PH3.
[26-32].
at reproducing
The pyrolytic
Growth
problems
the indium-containing
of trimethylindium
triethylindium
as
[PH3J!
or GaAs
Second,
there
with PH3.
it does mean
is
7
that
an excess
Third,
quantity
po1ycrysta11ine
or contaminated
of both
of TMIn must
material
be introduced
growth
can occur
areas of the substrate.
the group
the substrate
III a1ky1s
is less than
Magnesium
and group
was used for p-type
bis(cyc1opentadieny1)magnesium
on the hot susceptor
Fourth,
the cracking
V hydrides
100% for the typical
doping.
(CPZMg) was
into the reactor.
efficiency
in the hot zone near
OMVPE
reactor.
The organometallic
introduced
compound
into the reactor,
and by the reaction
(CSHS)ZMg--+
Mg was available
growth
their
points
solids
-lS.8
instead.
growth
the published
is a liquid
determine
values
growth
the
(boils at
HZ through
conditions
for InP and InO.S3GaO.47As
in Table
I and Table
samples
grown
II, which
at OGC for comparison.
for various
a
was varied
of the TMGa bath.
conditions
conditions
over
of Ludowise
at room temperature
to the gas stream DY bubbling
used here are shown
give the typical
specific
to accurately
1-xGa xAs was grown, the composition
the temperature
by the method
HZ was passed
mixtures.
in this study
TMGa
as
at room temperature
When In
Published
also
was made
°C), and was added
adjusting
for incorporation
and 88.4 °c respectively),
of the organometa11ics;
used
TMGa vessel.
of the epi1ayer
CpZMg and TMIn are solids
to obtain vapor
pressure
[S] were
Since
176°C
No attempt
vapor
at the surface
proceeded.
(melting
Mg + ZCSHS'
are given
(The
in later
by
8
TABLE
I.
TMIn
source
Lab
RSREc
1983
Comparison of Atmospheric
for InP Using TMIn.
Alfa
60°C
PH3
source
Carrier a
gas flow
(10% in H2)
(l/min.)
Air Prod.
]
RSREc AHa
1984
60°C
[41,42] [10-4]
4xlO
[2.4xlO
-3
Growth
Conditions
0.0420.13
2xl015
3700
mIll dia.
24
638
43
650
0.033 1000
8xl014
40 mm dia.
J
-3
[3.l9xlO
5
]
12.2 °c
[7.43xlO-5]
France Telecom
1983
[30]
4 H2
8 N2
65
Fr. Tele:
1984
[28]
10-30 750
7.5
Air Prod.
Cornellc
1985
[27]
40
]
OMVPE
[PH3] Temp. -GROWTHbBest
[TMIn] (OC) rate.effic. n_3
(~m/m1n.)
(cm)
5 (H2 &
He)
-4 [0.3--3
[0.5-2xlO
[40]
Pressure
Air Prod.
-3
Plessey AHa_5
1984
[4xlO]
[31]
[3.7xlO
0.06
650
0.033
...
600
0.03
4200
650
0.08
mIll dia.
Matheson
[2.525xlO-3]
AHa
25°C
650
4
2
4 cm
93
]
British Telecom
1985
[32J
d
U. of Utah
[0.015]
17°C
1983
-4
[34-6J [2.28xlO ]
2
2x5 cm
66
625
0.083
OGCd
1987
4
80
600
0.03 2000
Alfa
-5 Phoenix
2.4x4 em
[9.25xlO ] Research_3
[7.43xlO ]
aH2 gas unless
otherwise
bGrowth efficiency
assuming
the vapor
2xl016
noted.
Growth
tube cross-section
4500
noted.
=
(growth rate)/(TMIn molar flow), in pm/mole,
pressure of TMIn given in Ludowise [5].
cInP (100) substrates were etched in bromine-methanol,
then baked
in the reactor at over 650°C
for at least 10 prior to growth.
dInp (100) substrates
no pre-bake.
were
etched
in H2S04:H202:H20,
but experienced
TABLE II.
Lab
RSRE
1986
Comparison of Atmospheric Pressure OMVPE Growth Conditions for InGaAs Using TMIn and TMGa.
TMln
source
TMGa
source
AsH3
source
Queen Mary
AHa
Phoenix Res.
Coll'='460°c
[42]
[10
HP, 1984
[26]
17.3 °C_5
[3.48x10
]
[27]
1983
[28]
Bell Labs,
1986
P1essey
Alfa
-12°C
-5
[2.3x10 ]
[29]
Alf a
_5
[2.3x10
]
80
2.02
660
. ..
. . .
4
x 1014
13.5
173
(18)
1.52
540
(640)
.. .
. . .
2
x 1015
]
5
14.3
1.95
650
. . .
. . .
2 x 1015
Liquide
50% N2
. . .
650
0.06
...
50% H2
5
. . .
. . .
625
. . .
. . .
2 x 1015
4.4
59
1.74
600
0.06
5300
9 x 1015
. . .
. . .
. . .
650
0.13
. . .
. . .
80
. . .
530
.. .
2
31
. ..
650
0.05
6800
5 x 1015
Phoenix Res.
4
80
2.07
600
0.057 2600
3 x 1015
Air
Pro
[3.7x10
5%
]
in H2
[32]
U. of Utah:
1983 [37-8]
10°C
-12°C
1984 [33]
10°C
-11.
1986
-3
[1.6x10
. . .
. ..
7.5
Bes!3n
(cm )
]
[0.01] -
Air
. . .
British Telecom
OGC
[1.2x10
[0.01]
[4x10-5]
1984 [31]
1985
]
Alfa
AHa
25°C
Fr. Te1e.
[TMln] Temp.
[TMGa] (OC)
-3
-5
-5
[7.43x10 ] -14°C -5
12.2 °c
[3.8x10
]
Cornell
1985
-9°C
[4.95x10
]
[V]
[III]
H2 flow
(l/min.)
Alfa
[9.25x10
_5
Alfa
]
5
°c
_5
. . .
20,000
[4.5x10 ]
\0
10
chapters
along with
represented
in the tables
reactor
surface
to grow
defect-free
What
reactor
produced
design
material
for growth
The general
layout
When
contained
The reactor
on a compilation
The toxic
to remove
lines
hours/day.
of published
exhausted
itself were vented
or joined
Arsine,
cylinders
phosphine,
unifor-
through
was changed,
purged
a single
and toxic
through
VCR gaskets
reactor
were
tube,
and
the individual
during
blower
gas regulator
vacuum
prior
bodies,
air, 24
and the growth
barrel
316 S.S.,
to
seal, and
to outside
charcoal
line
the change.
with nitrogen
reactor
an activated
O-rings
(See
(shown in the top view
All gas tubing was 1/4 inch O.D.
tube,
in Fig. 2.
any room air introduced
by VCR fittings.
thermocouple
The
two-dimensional
is shown
The four cabinets,
All hydrogen
air.
list.)
could be individually
to room air.
box were
efficiency,
in gas cabinets
each of these
was evacuated
Plated
on the ability
for this study, but was not
of the reactor
were
outside
impact
abruptness.
hydrogen
tube
were based
rate, growth
E for an equipment
glove
and interior
of the OGC reactor.
acceptable
Appendix
opening
have a large
is a description
or interface
only).
which
purity
are not
It was built by Crystal Specialties, Inc. in 1985.
optimized
mity,
which
epilayers.
and apparatus
reports.
Major variables
are source material
cleanliness,
follows
geometry
results.)
to the
either welded
were used only to seal glassware:
and reactor
used throughout.
exhaust
Valves
oil trap.
contained
teflon
seal
11
H
Organometallic
2
Source Baths Cabinet
TOP VIEW
Process
Vent
\
1 m
Air
Cabinet
FRONT
Lucite
Arsine/PhosphO
Exhaust
Cabinet
VIEW
Glove
Box
Keactor
Gas Pipe
Cabin?
p
rf
III
I
Flow Meter
/
/'
-.
Indicators
III
T
Abort
Int50rs
Power
n
H
Bus
U~~
~PU~if
Generat~..
4m
FIG. 2.
CaDinet
& Valve
.#'
/
& Valve
OMVPE Reactor General Layout.
2m
"
>f
12
gaskets.
Prior
to installation,
hot trichloroethylene
acetone
wall
and hot acetone,
and methanol.
10-8 T.
The assembled
The two gas cylinder
flammable-liquid
inside
connecting
tubing
box was employed
components
or substrates.
at least
10 minutes
not used
switching
A schematic
Gas flow
the organometallic
output
valve
the tubing
and valve
in this figure.
source
at about
or condensation
to the reactor
tube.
lines
Details
tube as desired.
tube
components.
to left.
N2 for
door.
the reactor
was
over flow rates;
Heat
leading
is given
in Fig. 3.
tape was wrapped
40 em of the source
into the reactor
of the reactor
in.the
cabinet
tubing
are shown
oven could be positioned
The bake-out
tube,
This was done to
of the organometa11ics
the reactor
phosphorus
cabinet
40 °C at all times.
The rf coil and bake-out
white
and
tube
low purity
with
from within
manifold
in Fig. 4.
to convert
ports were
reactor
no control
doub1e-
bulkheads
load/unload
supplied
it provided
to the 3-way valve
plating
the reactor
sequencer
of the gas pipe
is left to right
maintaining
and unloading
to
was done manually.
around
prior
opening
process
leak checked
and vacuum
through
with
with
by OGC, using
This box was purged with
in this study because
all valve
were built
flowed
degreased
Gas flow in Fig. 2 is right
for loading
before
The programmable
system was helium
gases
to the reactor.
were
by cold rinses
All regulators
and process
The glove
followed
cabinets
cabinets.
the cabinets,
prevent
all gas line components
along
oven was used occasionally
to red, and to otherwise
clean
the reactor
Plant
Air
PH3
..
Vac
Seal
aust
Oil Trap
Vent
FIG. 3.
OMVPE Reactor Gas Schematic.
....
w
TOP
VIEW
--_._I
I
i
Water Cooling for rf Coils
I
,.
End Cap _,
I
Vent
I
I
I
Thermocouple
3 - Way
Valves
Bake-out
Oven
I
~
~
FIG. 4.
OMVPE Reactor
Cabinet.
15
The reactor
quartz
tube itself
ramp in front of the susceptor
susceptor
was SiC-coated
and held substrates
containing
which
(ungrounded).
the reactor
whenever
1000 to 3000 liters
then rinsed
gloves
were
until
were worn,
regia,
insert
and rinsed
turbulence.
possible
cross
quartz
tube
solely with
of H2 were
through
The
section),
13% Rh in an inconel
to sweep
of the
in the rear of the
were measured
100 cc/min.
kept
this
sheath
flowing
away contaminants.
the system
after
long
no HZ flow could be maintained.
a clean water
sample
break
edges were
in a teflon
and ramp were
in water
five to ten growths,
in a hole
etched just prior
etch or rinse was done
The quartz
fitted
of HZ were purged
during which
The purpose
A 5 mm diameter
was Pt vs. Pt with
Approximately
Substrates
was to reduce
at a 10° angle.
was
in Fig. 5.
(Z.4 x 0.6 em front
The growth temperatures
thermocouple,
shutdowns
graphite
a thermocouple
susceptor.
through
is detailed
the susceptor
at 40°C
was obtained.
held with
beaker
cleaned
followed
for 30 minutes in H2 at 900°C.
to growth
teflon
dedicated
prior
by acetone
was cleaned
Talc-free
latex
tweezers,
and the
to the purpose.
to each growth
in aqua
and methanol.
Every
in aqua regia
on InP, a two minute
and start
of AsH3
remaining
PH3.
and baked
GaAs substrates were baked at 800°C
for 5 minutes in 10-3.atm of AsH3 just prior to growth.
was grown
in a 5:1:1
gap was kept between
flow during which
II/min.
This was to prevent
InGaAsP
When InGaAs
the end of PH3 flow
of HZ flushed
growth.
out the
AXIAL VIEW OF END CAP AND
RECTANGULAR INSERT
SIDE
End Cap
VIE'"
Rectangular
Insert
TMIn &
CpZMg
PH3
Round
Insert
AsH3
Substrate
(On Susceptor)
~
Gas Flow
6mrn
I
Outer
Tube
4cm -1
a-rings
40 cm
48 cm
FIG. 5.
OMVPE Reactor Tube.
j
Pressure
Gauge
.....
0\
17
Numbering
GaAs
of samples
1 was grown
grown
InGaAs
3/30/86
was sequential
10/17/85,
while
GaAs
and InP 47 1/16/87.
69 was grown
and the resulting
2/2/87.
epi1ayer
Growth
by epi1ayer
62 was grown
InGaAs
times
thicknesses
9/15/86.
1 was grown
ranged
were
material.
InP 1 was
4/25/86,
while
from 30 to 120 minutes,
1 to 6 microns.
18
CHAPTER III
CHARACTERIZATIONTECHNIQUES
A. Photoluminescence
(PL)
spectra were taken on nearly all epilayers and
Photoluminescence
substrates
levels.
for the purpose
A good general
has eluded
the author,
[45], and good
information
setup
An argon
discussion
in Fig.
provided
afterward.)
of the sample
surface
point
the low temperature
In order
focal
=
length
with
monochromator.
of PL
is in reference
to the quaternary
radiation
of above
The laser was focused,
=
alloys
bandgap
InGaAsP
energy
of a large
diameter
The same was mounted,
checked
unfiltered,
120 mm, 42 mm diameter
using
by
on the
lens.
The same
(30 rom), short
rubber
cement,
focal
on
pad of a MMR cryostat.
the need
respect
to focus
for precise
to the sample
of the monochromator,
34 mm) was used
background
in 1981 and was occasionally
by a f
to obviate
monochromator
and measurement
(Power level calibration data was
was at the focus
(10 rom) lens.
and defect
6.
by the manufacturer
surface
length
purity
For the following description, refer to the
ion laser
OGC personnel
of the theory
in regard
(514.5 ~m) at up to 20 W/cm2.
provided
relative
but some theoretical
is in reference [46].
experimental
of establishing
and to match
a third
the light
location
lens
(f
=
of the
more
32
onto the entrance
rom,
closely
the
diameter
slit of the
A filter was introducedinto the collimatedportion of
19
Path
ChoppeT
Lock-in
Amplifier
Detector
Acplifier
FIG. 6.
Schematic of Photoluminescence Apparatus.
20
the light
path
wavelengths
minimized
width),
width,
to remove
laser
from passing
to obtain
but light
emission
through
good energy
throughput
the monochromator.
resolution
dropped
as can be seen in Fig.
and to prevent
7.
with
and output
order
Slit width
(dispersion
off rapidly
Input
lower
was
was 6 nm/mm
decreasing
slits were
slit
slit
always
set to the same width.
The monochromator
the Ar+ ion laser
(632.8
nm).
lenses
used
possible
to the output
clean N2.
used.
The well
The larger
detector.
was
both
laser
by the windows
as close
or
as
in between.
The signal
by a conventional
lock-in
for absorption
for all PL data since
Ge detector
fitted with
1500 nm.
a gas inlet
to purge
the PL studied
absorption
H20 absorption
assumed
response
water
vapor
peak at 1375
8.
Since
this
here, N2 was not typically
drop at 1410 nm was obtained
and was therefore
it had a
beyond
atmospheric
(The
measurements.)
as can be seen in the inset of Fig.
The detector
peak heights
was used
known
did not affect
PbS detector
lamp as used
than the cooled
[47] was observed
absorption
and a He-Ne
was positioned
slit, with no optics
using
Detector response curves are given in Fig. 8.
The monochromator
nm
was observed
The detector
PbS detector
response
was calibrated
and 488 nm lines)
absorption
source was the Oriel
better
indicator
and fed to an X-Y recorder
(uncooled)
with
496.5,
in this study.
detection system.
The
(514.5,
No selective
was amplified
light
wavelength
only with
to be characteristic
of this
curve was used to correct
relative
in some of the PL data; whether
noted in the text with each figure.
or not this was done
An equipment list for
PL
is
the
21
>-
100
.....
(f)
80
z
w
.....
60
.....
::>
a..
.....
::>
40
z
o
20
o
o
0.5
1.0
I. 5
2.0
MONOCHROMATOR SLIT WIDTH (mm)
FIG. 7.
Light Output for Various Monochromator Slit Widths.
22
PbS
300K
t
1400 nm
:::J
o
77K
Ge
Unfi Itered
>r(f)
z
w
rz
300K
PbS
Unfiltered--...
400
600
800
1000
WAVELENGTH
FIG. 8.
1200
1400
1600
1800
(nm)
Optical Detector and Filter Spectral Response.
23
measurements
When
is given
determination
photoluminescence
values
in Appendix
the impurity
of impurity
data,
[46,48-50].
a simple
peaks
of least
(Raw data are in Appendix
report
as full width
Data
samples.
from
in PL.
squares
C.)
in Chapter
16
-3
ing levels in excess of 10
em
in this
ratios
employed
important
directed
data.
throughout
this
NA-ND
for InP:Mg
in the determination
A method
is valid
of car-
only
has also been
un-
for dop-
claimed
to
from PL [51], but that was not attempted
to increase
obliquely
face and cryostat
Secondly,
although
wavelengths
greater
substituted
for looking
Si filter
the sample,
window
a Si filter
in Fig.
transmission
50%).
laser
6, the laser
to enter
is desirable
methods
One step that was
scattered
were unable
emissions
is only about
and several
thus the specular
than 1200 nm, a glass
at weak
seen,
ratio.
was to direct
As indicated
toward
easily
the signal-to-noise
in early measurements
the sample
slits
referenced
IV to estimate
This method
was not always
away from the detector.
tor.
9, which
study.
Photoluminescence
were
in Fig.
of
from the data in Fig. 9 that considerable
by this method.
compensation
width
(FWHM).
a factor of two--exists
rier concentrations
the
can be seen
are given
using
to literature
broadens
fits to the above
Peak widths
are used
It is obvious
determine
This
was attempted
was made
of impurities
at half maximum
[46]
certainty--perhaps
concentrations
comparison
The presence
transition
is a compilation
E.
beam light
beam was
reflections
the mono chroma-
for order
sorting
filter was occasionally
in this
Third,
were opened as wide as possible (consistent
range
from
(because
the
the monochromator
with reasonable
at
24
>
Q)
E
45
r
40
77K
35
30
25
:I: 20
u.. 15
...J 10
5
0
1016
1017
(cm-3
INA-NDI
FIG. 9.
energy
when
the laser was
its surface.
laser
Fourth,
focused
An example
all parameters
beam.
surface
)
PL FWHM for Various Doping Concentrations. a) LPE
InGaAs:Zn [48], b) LPE InGaAsP:Zn [46],
c) VPE InP [49],
d) LPE InP:Zn [50].
resolution).
62 with
1019
1018
Variations
constant
this
a larger
was greatly
enhanced.
on the edge of the epi1ayer
is shown
of any samples.
the edge excites
PL emission
in Fig. 10:
except
It is presumed
thickness
PL results
for location
large were never
as opposed
of the InGaAs
from InGaAs
of the incident
observed
that laser
to
along
the
light entering
layer.
(This
enhancement was seldom needed. and all data reported in subsequent
25
-
::;
c
OMV PE
I nGaAs 62
77K
>-
fen
z
a
LL.I
.....
Z
LL.I
U
Z
LL.I
U
en
LL.I
z
:E
=>
.-J
o
.....
o
x
n.
1500
1600
1700
WAVELENGTH
FIG.
10.
chapters
the
center
generated
of the
This
accuracy
resolution
spacing
with
the
laser
edge-focused
input.
striking
laser
the
input,
surface
near
sample.)
When estimates
the
(nm)
PL for 'OMVPEInGaAs.
a)
b) surface-focused
laser
were
1800
of peak
was assumed
is
shown
of two vertical
energy
or FWHMwere
to be limited
in each
lines.
figure
by the
made from
PL data,
monochromator
containing
PL spectra
resolution.
as the
26
B.
Hall Mobility
Measurements
the van der Pauw
from this
Appendix
of the sheet
technique
reference.)
[52].
Ohmic
A, and were
placed
a diamond
lines)
was used
These
by the method
of samples-that
by
are
given
in
had been
For inhomogeneous samples, the epilayer
between
the contacts
to form a
The normal configuration (without scribe
for all data reported
measurements
effect were made
in this section
were made
in the corners
scribe midway
cloverleaf (see Fig. 11).
and Hall
(All formulae
contacts
cleaved into 3-4 mm squares.
was cut with
resistance
herein.
give the low field response
of the majority
carriers (Hall mobility), information about compensation (when the
data cover
a suitable
and net carrier
quantity
temperature
concentration
that is measured
and D in Fig. 11 when
resistivity,
current
I is passed
age V is measured across C and D.
where
R
=
VII.
edges with
tances
current
are averaged
of contacts
F depends
edges
In practice,
R.
and the resulting
F
~
1
_
R - Rs
(
R+R
s)
[
'lrdR/ln2
for all four
from adjacent
(ln2) 2 -
4
a volt-
resis-
F, in the calculation.
4
ln2 _
2
For
in the placement
as:
R - Rs 2
R+R
(
s)
=
eight
Any lack of symmetry
R and R s obtained
A, B, C,
A and B while
is repeated
a factor,
The
into the sample.
Then the resistivity p
by including
d).
the contacts
points
this measurement
to obtain
on the resistances
between
through
as "the resistivity
thickness,
is injected
of both polarities
is corrected
as well
(given the epilayer
is the voltage
a known
a current
range),
(ln2) 3
12
J
sample
27
A
D
~
.
.
.
.
B
ELECTROMAGNET
C
Inhomogeneous
(With Scribe
Sample
Lines)
SAMPLE
APPLE
lIe
CONTROLLER
N2 GAS
CURRENT
CYLINDER
SOURCE
DISC
STORAGE
SCANNER
PRINTER
ELECTROMETER
PLOTTER
TEMPERATURE
CONTROLLER
APPLE
INTERFACES
IEEE 4BB BUS
FIG. 11.
Schematic of Hall Mobility Measurement Apparatus.
28
For nearly symmetrical contacts, F = 1.
p
=
TIdRF/ln2.
Calculation
resistivity
of the Hall mobility
measurement
b, perpendicular
across
there
of these
value
on d.
assuming
the Hall
factor,
temperature
These
then n
formulae
controlled
Shown
schematically
were
The Hall
current
polarity
from
to give four
or n-type
con-
of the absolute
Note that since p
values
a:d, IlH
can then be ca1-
by one type of carrier
for the material,
dominates.
magnetic
If
induction,
and
rH/ellHP.
in a program
Hall mobility
(see Appendix
measurement
F)
system.
11, this system was set up as described
note
#501, omitting
E.)
set a particular
From any starting
stabilized after a 10°C
diagonally
i injected
is zero,
(the average
concentration
list is in Appendix
temperature--or
ture controller.
~r
induction,
is to be subtracted
p-type
indicates
incorporated
in Fig.
Each
and with
~
a semi-automated
application
(The equipment
=
vIi,
current
field
Carrier
rH, is known
used,
the Keithley
=
IlH = d ~r/bp.
conductivity
which
increase.
of
four differences),
cu1ated,
v is measured
the magnetic
From the resultant
does not depend
sample
when
(The polarity
ductivity.)
Now r
p, and a further
of a known magnetic
The voltage
are then four r values.
its corresponding
values.
requires
from B to D (or A to C) with
A and C (or B and D).
reversed
#H
in the presence
to the sample.
the sample
through
~r
Including possible asymmetries,
the unity
The computer
va1ue--via
temperature,
increase in 30 s.
factor was assumed
For GaAs, values were used from [53].
in
gain amplifiers.
monitored
the
the MMR temperathe sample
temperature
It took 50 s after a 50°C
to be unity
for InP and InGaAs.
(A recent report estimated the
29
Hall
factor
for InP
[54].)
All measurements
within
were made
the particular
voltage
drop was in the range
not measured
(except
have no deviation
control
although
setting
The resistance
Mobility
this was a small
For 5000 Gauss
changed
effect,
for both field
data
mobility apparatus.
OGC data agree with
parameters
the data,
changed
were
was usually
it was found
to
of the magnet
(see Fig. 12),
To see how the
the magnetoresistance
a 3000 Gauss
the same induction
of GaAs 27
induction
change.
Since
study was done of magnetic
sought
effects.
was set to 4.0 and
to confirm
the accuracy
These data are summarized in Table III.
other
independent
of temperatures,
and n-type
which
current
polarities
data the control
chosen
such that the
Calibration
was used.
2% with
no further
Gauss
because
were
(See Fig. 12.)
Corroborative
range
polarity
0.3% with
and 10,000
9.0 respectively.
both p-type
affect
value.
values
regime
The actual
samples)
from the programmed
might
Current
conduction
for some early
only one magnetization
was tested.
a broad
ohmic
10-100 mV.
was performed
field magnitude
change.
sample's
in the dark.
range
sample
materials.
over
several
of the Hall
The recent
measurements
to within
10% over
thicknesses,
and mobilities,
This is quite good agreement
orders
of magnitude.
for
for
30
-
10,000
-
8000
It)
It)
::»
D
C)
z
-0
U
::>
0
I
6000
z
u
-
4000
.+
POSITION f112
POLARITY
/"'0
I
-
__
.
-
POSITION
4' I
POLAR ITY
w
z
C)
<I
2000
0
0
2
ALPHA
FIG. 12.
4
POWER
6
SUPPLY
8
SETTING
Electromagnet Power Supply Calibration.
10
31
TABLE
III.
Comparison
of
OGe Hall Mobility
Results
with
Other
Labs.
Data
source
Test
date
Temp.
(K)
OGe
1/79
300
0.018
112
3 x 1018
GaAs:Zn (111)
t = 500 JIm
OGe
6/86
292
0.0202
128
2.7 x 1018
OGe GaAs55
n-type (100)
t = 1 JIm
DEI
10/86
77
0.0733
88,700
9.6 x 1014
OGe
9/86
79.5 0.0816
86,100
9.1 x 1014
7/86
300
0.0267
10.3
2.3 x 1019
3/86
292
0.0267
11.2
2.5 x 1019
Sample
identity
Laser Diode
P
JIH
(nem) (em2/Vs)
n or p
-3
(em )
Lot IIA.0809
Tektronix 117
B-doped (100)
po1ysilieon
t = 0.4 JIm
Tek
OGe
32
C. Auger
Electron
Auger
relative
Spectroscopy
analysis
amounts
matched.)
for the sample
diameter
was used.
sample
(Optical
to verify
absorption
that this sample was nearly
InGaAs
in January,
65.
The beam
the
and
lattice-
1987, at Charles
A 10 kV primary
Evans
beam of 10-20
current was approximately
1 ~A,
rate was 100 A/min.
A comparative
"standard"
laboratories,
sample
and gallium.
had predicted
and the sputtering
Evans
on one InGaAs
The AES work was performed
& Associates
microns
was undertaken
of indium
photoluminescence
(AES)
under
OSU 1-29-7-86,
ternary
sample was also analyzed
the same conditions.
grown
by molecular
beam
at the
This was the InGaAs
epitaxy
and obtained
from Professor J. R. Arthur of Oregon State University.
The bandgap of
this
at 77 K and
"standard"
300 K.
been
These
for detector
at 0.789
Eg(300
the relations
K)(x)
+ 0.53x2)
=
(0.35
i.e. lattice-matched
occur
=
0.468
response
13.
The peak heights
with wavelength.
is centered
VI, the PL results
have
The 77 K peak
at 0.736
for InGaAs
eV.
65 were
is
(As
ambiguous.)
for E (x) of In l
-xGa xAs given in [55], namely
2
g
+ 0.6lx + 0.46x
eV, the composition
match is x
in Fig.
eV and the 300 K peak
be seen in Chapter
Using
at OGC by photoluminescence
data are presented
corrected
centered
will
was measured
to InP.
and
),
parameter
are different
for InGaAs
proposed
g
K)(x)
(At 300 K, the precise
because
for a given
the coefficients
and InP.)
Alternative
for E (x) at room temperature
=
x for this sample
[56]; lattice-matching
at one temperature
Eg(77
value
(0.41
is x
of thermal
[57,58] and at 2 K
=
0.47,
for a lattice
sample
relations
+ 0.57x
can only
expansion
have been
[59], as well
33
-
::t
CJ
MBE
In I-X Gox As
OSU 1-29-7-8
UJ
o
77K
z
UJ
o
(/)
UJ
Z
~
::>
I
o
~
o
J:
Cl.
1500
1600
1700
WAVELENGTH
FIG. 13.
PL Spectrum for
as a relation
for
The value
calibrate
data
of
the
that
As content,
the
top
for
=
x
0.47
atomic
23.5%
4000
A of
(nm)
InO.53GaO.47As
for
the
MBE "standard"
data
The result
derived
is
and 26.5%
the
That
OMVPEsample
layer.
MBE sample
InGaAs
65,
sample
from
given
Ga content,
of the
"Standard".
[56].
percent
sample.
PL/ AES analysis
data
E (T)
g
1800
the
14,
In content
was then
as reported
which
uniformly
scaling
used
in
to
AES raw intensity
in Fig.
compositional
was used
to
VI.
50%
throughout
derived
interpret
Chapter
shows
the
from
AES
34
100
90
I
MBE
80
70
E-< 60
:z
t:J
u
p:: 50
t:J
p..
U
H
::E:
0
E-<
<::
In1 -x GaxAs
.
OSU 1-29-7-86
I
As
40
II
In
30
20
10
1000
2000
3000
DEPTH FROM SURFACE (A.)
FIG. 14.
AES Profile
for InO.53GaO.47As
"Standard".
4000
35
D. Secondary
Ion Mass
Spectrometry
The SIMS technique,
tered
ion mass
to measure
spectrometry,
Cs+ ion bombardment
was employed
0+ ion bombardment
used
to measure
and positive
sputtered
Fe, Mg, Hg, Na, W, Cu, Ni,
Cs+ ion bombardment
was used
to measure
monitored
to locate
the interface
Evans
& Associates
in InGaAs/InP.
SIMS
ion spectrometry
sputtered
was
ion spectrometry
distributions.
between
sput-
Cr, and Al distributions.
and positive
the Zn (as ZnCs+)
and negative
by Charles
0, C, S, Si, CI, and Ag distributions
with
SIMS with
using
(SIMS)
Arsenic
the epilayer
was also
and the sub-
strate.
Appendix D contains the processed data, plotted semi-Iogarithmically
as "concentration"
processed
data plot
versus
is labeled
are referenced
to this scale.
of ion current
to concentration
standards
in GaAs,
GaAs matrix.
and should
For InGaAs
larger.
The quantitative
sitivity
factor
The accuracy
depth.
The right-hand
"secondary
With
ion counts,"
the exception
they were not
be accurate
to a factor
of two in a
Ag data were
is unknown
obtained
electron
be within
have been
for possible
in the correct
an interference
of Ag, the conversion
of ion implant
from the relative
to check
and the As data
on analyses
about
S, Si, Cu, Zn, and Hg, were monitored
per element
of each
was based
the uncertainty
for Ag should
axis
molecular
isotopic
and probably
by estimating
affinities
an order
using
a sen-
of Ag and As.
of magnitude.
at least
two isotopes
ion interferences.
ratio when measured,
at one or all of the isotopes
The profile which yields the lowest concentration
much
If
there must
of that element.
after this comparison
36
will
be closer
because
no ion implant
are near
rather
to the actual
those
of Hg.
concentration.
standard
Thus
The Hg data are unreliable
was available,
and isotopes
the Hg data probably
track
of In-Ga-As
the ternary
than Hg!
The SIMS data are used only for comparison
and no conclusions
concentrations
are drawn
of expected
from them regarding
or unexpected
purposes
in Chapter
the absolute
contaminants.
VI,
37
E. Optical
Absorption
Optical absorption coefficient data for InGaAs (a) were sought
for two reasons.
First,
ternary
Also,
samples.
of photodetectors
While
interest
intensity
reductions
1500 nm light passing
removing
polishing,
InGaAs,
etch
While
substrates
fixture
mounting
from the heat,
Also,
when
would
flow between
Despite
and yielded
free-standing
etched,
the InGaAs
for
methods
for
thinned
by
wax are depicted
to work,
not always
layer
were
attached
very
completely
low surface
samples
the InP substrate
that an
ways
of
in Fig.
15.
there was difficulty
and the glass,
early
InP but not
meant
Several
the wax solidified
several
After
which
the ternary.
be made
and would
samples
Several
in this
observed
Since HC1 etches
hot, the wax had an extremely
some data.
is desirable
used
were mechanically
slide using
because
such difficulties,
for a substrate
For substrates
destructive.
could
the process,
of
tried.
to support
to a glass
in the wavelengths
of the InP substrate
could be selectively
all these methods
removed
light
the substrate.
on some samples
was needed
samples
controlling
away,
were
but this proved
devices.
is large enough
ratio.
of
the performance
of up to a 2:1 factor were
through
the substrates
The substrates
absorb
that removal
the signa1-to-noise
the bandgap
of a helps predict
the absorption
thickness
to estimate
optoelectronic
InP does not strongly
of 400 microns
study,
knowledge
and other
for InGaAs,
to improve
they were used
quickly
tension,
etched
edges
and
the sample.
was entirely
by their
when
seal all edges.
ruining
were
in
this way,
etched
to a glass
38
METHOD A
Pull
Sticks
Heat Until Wax
Lay Strands
at
Sample Edges
of Wax
Melts and Seals Edges
Hot Plate
InGaAs Layer
Facing Glass
Step 3
Step 2
METHOD
B
Seal Wax to Glass,
Remove Rod
Layer
Rubber
Cement
~
Rod
\.]ax
Attach
Sample
to Holder
Hot Plate
Step 1
Hot Plate
Step 2
METHOD
Step 3
C
Pour
Hot
Wax,
Trim
When
Cool
Fixture
Surrounding
Sample
FIG. 15.
Methods of Mounting Epilayers for Substrate Etching.
39
slide with
ideal
GE #7031 varnish.
because
strain
of the sample.
However,
in the epilayer
In addition,
it was difficult
this mounting
could
cause
scheme was not
cleavage
and curling
if the InP was only partially
to quantify
scattering
and reflection
etched,
losses
at the
InP exit face.
The most
successful
sealing
the InGaAs
epoxy.
While
layer
sealing
hours
the epoxy was not
to a small
glass
slide by means
immune
noted
developed
slide
with
samples
(6 x 10 rom), etching
remove
GE #7031
the thin
So for the
were
prepared
by
in 38% HCl for 1-2
the substrate,
varnish
involved
of a transparent
from attack by HCl,
otherwise,
to fully
in the cryostat
method
to the glass was not attacked.
here, unless
at room temperature
mounting
preparation
layer to a glass
the epilayer
data presented
epoxying
sample
rinsing,
as shown
and
in the inset
of Fig. 16.
Figure
16 also shows
for an equipment
depicted
list.)
in the inset.
for the short
measurement was
The stability
+ 1 °C.
was checked
data were
and sample
aperture
that the intensity
time;
using
sample
was maximized
and glass
each.
is
adequate
no more
than
temperature
filter
were
respectively.
at 1300 nm to ensure
positioned
through
of the
nm ranges
each filter.
alternately
it drifted
The accuracy
E
refrigerator
of the lamp was quite
nm and 800-1300
on each
(See Appendix
of the Displex
As in PL, a Si filter
obtained
were
apparatus.
measurement
a two hour warmup.
for the 1300-1800
Transmission
sistent
The cryotip
(about 5 min.)
1% per hour after
employed
the optical
The reference
within
that
aperture
the beam
The light
con-
from
such
40
Reference
Beam
Apertures
lass Slide
G. E.
7031
Varnish
ryotip
Position
Adjustment
Lenses,
diameter
Lens,
f =
65 mm,
45 mm Dia.
Chopper
FIG. 16.
Optical Absorption Apparatus.
= 32 mm,
f
=
34 mm
41
the monochromator
imaged
not
onto
an (uncooled)
identical
through
was imaged onto
in area,
both apertures
either aperture,
PbS detector.
a scaling
when
factor
Since
and was
in turn
the two apertures
(the ratio
open) was measured
were
of intensities
and applied
to the raw
transmission data.
Transmission data (T) were extracted graphically
at various
energies
discrete
wavelength.
The resolution
measurements
(0.5 mm slits).
To convert
refractive
The data
of both
of InGaAs
from these,
from Burkhard
between
calculations
by 40 meV
of bandgap
those
a smooth
respectively,
temperature.
index with
changing
data were
0.47;
of 1300 nm light
two glass
no epoxy.
slides
x was attempted
through
measurement
replaced by an InP substrate.
to higher
energy
the variation
on LPE material
here.)
by comparing
a layer of epoxy
to the transmission
A similar
with
For
no estimate of variation of
The index of the epoxy, n e, was measured
mission
In the
interpolated.
(These data were measured
=
Data
1.2 eV, and those
shifted
to correspond
[60,61].
in Fig. 17.
above
curve was
Only
to the author
are used for energies below 0.76 eV.
with the composition parameter x
refractive
of the
and InGaAs was necessary.
index are known
at 77 K and 10 K, these
with
versus
3 nm for all
knowledge
at 300 K, are reproduced
energies,
and 60 meV,
was
coefficient,
et al [61] are used for energies
of Chandra et al [60]
range
the epoxy
refractive
valid
of intensity
of the monochromator
T data to absorption
indexes
two studies
from the X-Y plots
through
sandwiched
the transbetween
the same two slides
was made with
one of the glass
A schematic of each is given in
with
slides
Fig
18.
42
3.75
LPE
100.53 G00.47As
300 K
3.70
3.65
x
w
o
z
"
3.60
w
>
I-
,I
,
,, ''
U
<X:
a::
LL
W
a::
"
,,
"
,..
,..
"
,,,
,
,,,
,
,I ,
,...'
3.55
3.50
3.45
3.40
0.5
0.7
0.9
ENERGY
FIG.
17.
1.1
1.3
1.5
(eV)
InGaAs Refractive
Index.
Solid line below 0.76 eV
is data from [60], above 1.2 eV is data from [61].
Broken line is an interpolation estimate.
43
I.J.
I
1111
I.J.
I
Glass
S
I s /1 0
= 1. 05
I /1
s 0
=
0
Glass
Epoxy
InP
Glass
I.
J.
I
1-*1
I.
J.
FIG. 18.
Schematic
In the first
case,
the
reflectivity
of
the ratio
interface
the
1.19
0
for Epoxy Refractive Index Measurement.
of intensity
that through the air gap was 1.05:1.
of each air-to-g1ass
>1
S
through
the epoxy to
Assuming that the reflectivity
is 0.04
epoxy-to-glass
(i.e. a glass with ng
interface,
R
eg
can
be
=
1.5),
estimated
from
so that
R
eg
=
0.16
=
2
e
This implies that n e
In the second
2
(1.5 - n ) /(1.5 + n )
e
.
= 1.2.
case, the measured
intensity
Assuming an InP refactive index of 3.21 [61],
reflectivity
of 0.276,
a slightly
different
ratio was 1.19:1.
which gives an InP-to-air
expression
results:
44
1.19 where
This
R
eg
-
(1.5
expression
epoxy
R
eg
is
n ) /(1.5
e
well
on these
of this
interest.
The refractive
all
regard
)
+ n)
e
and R
es
satisfied
if
ne
refractive
index
the
method
TABLE IV.
2
n e ) /(3.21
refractive
index
temperatures.
was observed
of the
glass
of proportional
error,
all
to sign,
for the various
parts
cient measurement are given in Table IV.
the
(3.21
2
+ n e ).
= 1.3.
for all
index
=
No spectral
in the
slides
of the
wavelengths
was assumed
of
to be
temperatures.
The values
with
es
to be n e = 1.3
was assumed
for
R
two measurements,
dependence
1.5
-
)(1
(1 - 0.04)(1 - 0.276)
2
2
=
Based
-
(1
_
of least
squares
independent
and
of the absorption
symmetrical
coeffi-
Combining these errors by
gives a simplifiedestimate of the total
Proportional Error Estimates for Optical Absorption.
parameter
n
e
proportional error (€ )
0.04
n( A)
0.004
t
0.05
T
0.005
A
0.005
S
0.005
n
g
(glass)
0.004
45
[62],
error
contains
equal to + 6.5%.
the largest
teclmique
is given
To calculate
[45]
from Moss
addition
contributing
in Appendix
the absorption
=
[1
=
n(A)
-
of vacuum-to-InGaAs
n( A)]
Ret
=
InO.53GaO.47As
=
N
8
=
A =
=
scaling
light
n ( A)
ik( A ), which
is
refractive
temperature
index,
1. e.
dependent,
measurement,
wavelength,
=
(21ft/
tP
=
angle
t
=
sample
1/1
=
arctan [2k( A) /n2 ( A)
the
,
from transmission
6
ex =
interface
part of the complex
factor
is temperature
2
/[1. 3 + n( A)]
+
which
above,
of epoxy-to-InGaAs
[1. 3 - n( A)]
imaginary
interface
,
2
k( A) =
+ 4/Ru..R_..sin (0 + 1/1)}
index of refraction,
as noted
reflectivity
coherent
2
/[1 + n( A)]
dependent
layer with
A )/n2( A)]
22.
R_.. exp(-at/2)]
2
=
was used
components:
- IR
reflectivity
measurement
(a), a formula
(T) of an absorbing
- Rvt) (1 - Ret) [1 + k2(
8(1 - 0.04) {[exp(at/2)
the thickness
coefficient
for transmission
(1
Rvt
error;
B.
of multiply-reflected
T=
Here,
The measurement of sample thickness
A )n( A
)cos
tP,
of incidence
=
0°,
so cos
tP
=
1,
thickness,
InGaAs
absorption
are trying
+ k2 ( A)
- 1)],
coefficient
to measure.
=
and
finally
41rk(A) / A,
1.e. what we
46
This
formula
was derived
and transmission
position
in a plane
were
significant
example
coefficients
of the amplitudes
tors
from the Fresnel
parallel
at each
interfacet
of reflected
plate
is given
for reflection
and a linear
and transmitted
(see e.g.
in the data for thint
of the raw data
formulae
[63]).
highly
in Fig. 19.
super-
electric
Interference
re1ective
effects
samples.
The intensity
vec-
An
of light
through the reference aperture (top) and sample aperture (bottom) is
shown in the figuret with interference peaks on the left.
Note that even for quite large values of at k2«
examplet
=
if A
4
1 ~m
then k2
=
further
that since
be
than
less
etc. [64]
[2k(
0.012.
The
series
2
A )In (A)
=
cm
two orders
and n
=
3.65 at 300 K)t
of magnitude
larger.
of the arctangent
expansion
will
= z
arctan(z)
-
Note
always
+ z515
z3/3
- 1)]
so our formula
1378
2
-
1)]
nmt
a
= 2k( A) I (n
4
10
=
In this caset 2k/(n2
2
arctan[2k/(n
T
13.3t
k < nt the argument
unity.
For examp1et if A
=
=
0.134 while n2
-1
4.6 x 10
=
For
then holdst and we can safely assume:
arctan
and k2
(where a
n2.
=
0.0185777.
- 1).
-1
cm
t
=
and n
2
3.58t
- 1) = 0.0185798t
FinallYt
can be simplified
(A)
note
then n
=
12.8
while
that cos2z
=
.
1
2
SID Z t
to:
(1
=
0.96sreat+ R R e-at
L:
vt et
Since
T and
this formula
At a computer
iteratively
matched
could not be solved uniquely
program
was written
to measured
T values
which
for a in terms of
allowed
to three
for trials
significant
of a
digits.
47
OMVPE
InGaAs 57
IOK
-
:J
C
>-
I-
CJ)
Z
W
tZ
1800
1700
1600
1500
WAVELENGTH
FIG.
19.
1400
1300
(nm)
Raw Transmission
Data for Optical
light
through
reference
aperture;
through sample aperture.
Absorption.
Top curve,
bottom curve, light
48
This program
is given in Appendix F.
The author
determining
methods
of %T
has despaired
bandgaps
from absorption
in the literature,
[65].
versus
[25,66,67].
theory
including
Some authors
hv to zero,
ignore
and ascribe
Others
include
[68] to their
data
There
taking
the "midpoint"
exciton
of "bandgaps"
study,
not
A number
can be fitted
of
of
in the rise
and extrapolate
energy
effects,
method
are a variety
effects
the resultant
[59,69].
for an accurate
data.
the exciton
and so a variety
I decided
in his search
a2
to the bandgap
and fit Elliot's
of assumptions
to the data.
to take time for the computation
are required,
For this
involved
in ac-
curately fitting absorption data to Elliott's theory, choosing instead
a graphical
technique.
The bandgap
higher
in energy
off rapidly.
straight
their
lines
of samples
than the energy at which
This point
fitted
intersection
proved
bandgap
was typically
to above-bandgap
at the "knee".
and the below-bandgap
was assumed
appearance
easier.)
in this study was assumed
of strong
exciton
and below-bandgap
experimental
absorption
peaks
the absorption
begins
found by extrapolating
When
to be 5 meV higher
to be somewhat
was nearly
in energy
would
to fall
two
absorption
technique
a vertical
imline,
than this point.
have made bandgap
to
the
(The
determination
The accuracy of this method is estimated at + 5 meV.
49
CHAPTER IV
SUBSTRATE AND UNDOPED EPILAYER
CHARACTERIZATION
RESULTS
A. GaAs and InP Substrates
The effect
is
of
a fascinating
scope
of
this
study.
was useful
This
section
topic
However,
presents
with
analysis
was done
data
discussed
PL spectra
the
at
PL
conjunction
presented
InP
Weak PL was observed
No PL was observed
three
figures,
21, n-type
for
the
no peak
in
the
these
in
figures:
Chapter
Fig.
and Fig.
detector
VI.
20,
22,
some substrates
for
sub-
SIMS
at
InP substrate.
correction
used
results.
lists
analysis
InP substrates;
semi-insulating
height
V
the
substrates
substrate;
three
for
the
beyond
manufacturer.
InP:Fe
InGaAs
is
epi1ayer
Table
by the
with
Fig.
to
data.
semi-insulating
GaAs substrates;
substrates.
but
of
comparison
provided
77 K are
on OMVPE epi1ayers
[70],
research
substrate
the
in
of
for
information
for
and defects
PL characterization
as a benchmark
along
are
impurities
and useful
here
strates
substrate
p-type
300
K.
For these
spectral
response
was employed.
Detection
of
and Buehler
[71]
luminescent
emitter
times
smaller
1ating
for
this
in the
PL proved
reported
than
GaAs substrate
PL
system.
Zn-doped
more
that
because
that
of
in
InP
the
is
GaAs than
100 times
20 is
recombination
near
of
corresponds
the
for
InP.
more efficient
The luminescence
The 31 meV shift
GaAs substrate
for
surface
GaAs.
Fig.
difficult
limit
as a
velocity
from
of
luminescence
to the well
the
Casey
is
100
semi-insu-
reliable
detection
to
energy
lower
known binding
50
TABLE
v.
Substrates
and OMVPE Growth
Material &
Orientationa
Manufacturer
GaAs
Spectrum
13252
+
(l00)
0.25°
GaAs:Zn
~ (110)
(100)
InP:Fe
~ (110)
(100)
& Wafer
Crystal
/1
Tech.
Spec.
Numbering.
Resistivity
(n cm)
n O!3P
(cm)
107
0.0025
GaAs 10-29, 31-62
InGaAs 2, 4-11
1018
E2652
ICI Americas
Rl163
InP
(100)
Atomergic
InP:S
(100)
Atomergic
InP:S
Sumitomo
506216
0.20
1016
InP 19-26, 32-4
InGaAs 20-3
0.001
1018
InP 17-8
InGaAs 3-7, 24, 31-6
0.0005
1019
InP 27-31
InGaAs 25-30, 37-46,
48-9, 57-64, 68-9
H1668
~ (110)
GaAs 2-9, 30
InGaAs 3
InP 37, 39-47
InGaAs 50-6, 65-7
H1667
(l00)
OGC b
usage
InP:Zn
(100)
Atomergic
H1669
InP 1-16
InGaAs 1-2
InP:Zn
(100)
Atomergic
11662
InP 35-6, 38
InGaAs 47
aA11
were
b
substrates
polished
by the manufacturer.
Of the first 18 InGaAs layers grown, 7 were on InP substrates and 11
were on GaAs substrates, both numbering sequences starting with
"InGaAs 1".
Growth numbers begining with InGaAs 20 are unique and
represent layers grown on InP.
51
21
HB
>-
GaAs
A
-
77K
-1r-
(f)
2
W
21
w
u
b\
I \a
2
W
U
(f)
W
2
::>
.....J
0
.....
0
I
a..
700
800
WAVELENGTH
FIG. 20.
900
1000
(nm)
PL for GaAs Substrates. a) GaAs:Zn, Crystal Specialties
#E2652, b) semi-insulating GaAs, Spectrum Technology #13252.
52
energy
of the Zn acceptor
in these
as "undoped"),
observed
marked
emission
acceptor
been
With
at 1.37 eV.
the moderately
sulfur-doped
at 1. 39 eV.
the peak now located
FWHM of 150 meV.
reported
elsewhere
the variation
These
1nP:Zn
possible
high
higher
substrate
considerably
shift
defect
possible
has not
in energy
1nP
resolvable
lines was not
1.0 eV.
to phosphorus
(to 40 meV) ,
substrate
complex.
used
similar
sources
#506216,
a very
to those
in this study
data, but several
emission
peaks
near
investigated.
Luminescence
are seen in each sub-
of this emission.
is pro-
Substrate #11662
The origin
of these
#H1669 has broad
in this energy
V for a more
22 does not
1.40 eV and what
Substrate
[74,75],
demonstrated
Figure
at 1.33 eV and 1.29 eV.
See Chapter
seen in
at 1.48 eV, and with
are quite
(b)
[73].
substrates
vacancies
(see curve
is clearly
from a single manufacturer.
resolution
also has peaks
tributed
the band-
emission
bably band-to-acceptor (Zn) emission at 1.36 eV [50].
sion at about
(sold
to be donor-to-
#H1668
doped n-type
spectra
Both have band-to-band
emission
#H1667
These were
This latter
The Burstein
for n-type
The two p-type
represent
1nP substrate
presumed
FWHM broadened
(c) of Fig. 21 for the highly
strate.
was observed
identified.
and was centered
broad
emission
(a) in Fig. 21.
at 1.40 eV, and emission
of Fig. 21), the emission
with
in the n-type
as curve
(or donor-to-band)
clearly
curve
No deep-level
substrates.
Two peaks were
to-band
[72].
but quite
detailed
range
is often
likely
emisat-
involves
discussion
of
a
53
:J
I
LEG
InP
....
77K
(f)
z
w
....
zw
-1r-
u
z
w
u
(f)
w
Z
-
=>
-I
0
....
0
:I:
0..700
800
WAVELENGTH
FIG. 21.
PL for N-type lnP Substrates.
1000
900
(nm)
a) lnP, Atomergie #RI667,
18
b) lnP:S, Atomergie #RI668, ND- NA = 10
19
-3
Sumitomo#506216,ND- NA = 10 em .
-3
em
, e) InP:S,
54
:J
21
LEC
InP
-(f)
77K
.
1r
z
w
z-
w
u
z
w
u
(f)
w
z-
::>
...J
0
0
:r:
Cl.700
800
900
1000
1100
WAVELENGTH
FIG.
22.
PL for P-type InP
#H1669. b) InP:Zn.
Substrates.
Atomergic
1200
1300
1400
(nm)
a) InP:Zn,
#1662.
Atomergic
55
B.
GaAs and InP Undoped
Most
of the growth
of the reactor
Epi1ayers
parameters
were worked
necessary
out by growing
GaAs prior
growth
of In-containing
compounds.
here.
(These represent
the characteristics
later
InGaAs
regarding
ample
epi1ayers.)
growth
exists
One important
samples
to GaAs
A rectangular
The insert
growth
parameter
quartz
provided
three
flow rate
(more efficient
than the outer
quired).
VI suggest
higher
although
samples
round
tube, which
for all
24 (i.e. through
7/21/86).
for subsequent
and 3) was much
disassembly
samples.
gas flow over the
grown with
crystal
easier
the rectangular
samples.
structure
These
did vary
the data in
and purity
result with
InP epilayers
to a minor
in the rectangular
than those
in Table
insert
Also,
tube with H2 flow rate as the principal
fraction
to clean
of the end caps was not re-
over the substrate.
flow velocity
and so
As delivered,
was used
1) near-laminar
gas usage),
of samples
subsequently
flow rate was less.
growth
than that of previous
the TMIn mole
grown
compounds,
geometry.
2.4 x 4 em, was used
that improved
gas velocity
in the large
higher
better
for
2) higher gas flow velocities for a given
(because
The morphology
was noticeably
Table
tube
layers used
body of literature
binary
was reactor
advantages:
rectangular susceptor [76],
reported
the OGC reactor.
45, InP 26, and InGaAs
insert,
to attempting
of buffer
of these
to evaluate
operation
InP data are also
is a considerable
had a 55 mm diameter
prior
Undoped
and characterization
opportunity
the reactor
There
for successful
degree.
were
variable,
Most
insert benefited
VI, even though
grown
from a
the H2 input
56
TABLE VI.
Hall Mobility and Resistivity
Total Gas Flow Rates.
[TMIn] H2 flow rate
(l/min. )
x 10-5
InP sample
II
for OMVPE
InP with
PH2 (300 K) P (300 K)
(cm /Vs)
(n cm)
6
2.3
5
1600
0.071
5
1.9
6
2550
0.048
13
2.8
8
4820
0.54
A second
subject
of concern
gas cylinders
require
frequent
with
a regulater
the toxic
evacuation
gas sources.
observations
during
in-line).
For example,
(listed
Room temperature
line,
was a suspicion,
data,
that
Hydrogen
were not fitted
as was built
based
the change
(despite
immediately
the presence
into
on morphology
after
the growth was poor due to the atmospheric
introduced
ditions
and initially
port or N2 purge
and Hall mobility
the H2 cylinder,
was the gas supply purity.
change,
There
Various
changing
gases
of the H2 purifier
GaAs 11 and 12 were grown under similar con-
below),
but the H2 tank was changed
Hall mobility
prior
to GaAs
for GaAs 11 was 4050 cm2/vs
12.
but only
2
3300 cm Vs for GaAs 12.
just
prior
to growth
As a further
=
600 °C, [PH3]/[TMIn]
=
lator
after
of 4820
a vacuum
cm2/Vs.
for
-5
80, 5 l/min.
line and the H2 cylinder
mobility
Hall mobility
InP 2 was grown with the same parameters
but the mobility increased to 2660
Immediately
the H2 tank was changed
of InP 1; the room temperature
2
this sample was 1980 em /Vs.
(T
example,
H2, and [9.12
x 10
] of TMIn),
em2/vs.
exhaust
changed,
Sample numbers
port was added
to the H2 regu-
InP 13 was grown
begining
with
yielding
GaAs
a
35, InP 13,
57
and InGaAs
exhaust.
12 were
That
A) evacuated,
and B) were
new
is, after
B) purged
repeated
cylinder
Table
VII shows
three
groups
VII.
N2
grade
using
cylinder,
a separate
for at least three
nitrogen
E, with
the vacuum
the regulator
vent,
cycles before
connecting
Hall mobility
of GaAs
samples.
low purity
nitrogen
Hall Mobility
GaAs
amO\mts
was used
for OMVPE
process,
it was hoped
--
(cm )
300 K (cm /Vs) 77 K
6290
low purity
40 to 52
. . .
118
380
low
53c
8 x 1014
6480
37,700
54 to 62
8 x 1014
4860
70,500
~ata
shown are the averages of about
group and represent all data taken.
measurable
Since
that the
PL
FWHM
(meV)
Hall mobilitya
2
2710
No PL was
the
the reactor.
7 x 1016
b
during
for
GaAs.
n
-3
II
concentration
group was grown
for purging
in the growth
sample
and carrier
The middle
are
of 02 and H20 in each.
1 to 39
UPC
the
Two types
UPC
purity
was
then steps A)
was also a concern.
the relative
the average
N2 was not used directly
TABLE
a new
with H2 through
of process
in Appendix
time when
installing
installed
to the reactor.
The purity
listed
grown with hydrogen
13
half of the samples
in each
on GaAs 48, GaAs 50, and GaAs 51.
cGrowth conditi~~s were: T
and [2.95 x 10
] of TMGa.
=
700°C,
[AsH3]/[TMGa]
= 52,
2 limine H2,
58
lower
purity
as expensive
clear
type would
as the low purity
decision
during
to be made
the same
troduced.
be acceptable.
cleaning
the low purity
that the rectangular
the round
outer
that the O-rings
the end caps were
sealed.
into the reactor
the bulk
during
growth
of the improvement
to a good reactor
seal,
in samples
insert was being
which
This allowed
for samples
prior
after
a
N2 was tried
tube just prior
of GaAs 53, it was discovered
not tightly
N2 was ten times as
The data do not allow
on this, because
time period
While
type.)
(High purity
to the growth
seal
this tube to
atmospheric
to GaAs
in-
53.
gases
While
GaAs 53 may have been
the use of high purity
N2 was continued
due
as a
precaution.
The hydride
used
initially
from Phoenix
dustry
used
were
GaAs
40 and InGaAs
that of GaAs
high
purity
also a concern.
The arsine
These were
the latter vendor
gases.
and phosphine
replaced
with
was reputed
The Phoenix
Research
11; the new phosphine
gas
in the inarsine
was
was used after
InP
PL at 77 K for GaAs 11 gave a FtmM of 20 meV, while
36 (new arsine)
of H2' with
was
15 meV.
[AsH3]![TMGa]
Both
=
samples
20.
was 10 meV.
the same--see
Table
(Note that the growth
l--except
that 2 l!min.
were
grown
Regarding
77 K PL for InP 40 gave a FWHM of 15 meV, while
(new phosphine)
growing
because
34 and InGaAs
66.
°c in 4-5 l!min.
were
were
from Air Products.
Research
to produce
after
phine,
sources
at 700
the phos-
that of InP 46
conditions
of these
of UPC N2 was used
in
InP 46.)
Because
new sources,
the best binary material
it was concluded
was grown
that Phoenix
after
Research
changing
hydrides
to the
give the
59
potential
of growing
growth;
too many
other
the improvement.
a half,
material.
grown weeks
Through
GaAs
in a clean
reactor
much
apart must
glove
in the center
because
from the tube walls
A similar
improvement
insert was removed
moved
improvement
was
etching
taken
cleaned
insert.
unloading,
significant
in particu-
was done
covered
allowing
as
insert
the substrate
re-
was con-
(With the round
fell on the substrate
tube,
this.
to and from the
Use of the rectangular
loading,
from
and achieved.
and were kept
seen during
from the insert without
was sought
glass
from the outer
drawn
reduction
were
during
cleanliness
by acknowledging
continuing
dish,
and
in gas line clean-
The samples
of the newly
particles
in reactor
such
to quantify
of a year
11, all pre-growth
in the glove box.
mainly
study
and so any conclusions
of this work,
box in a covered
in this
variations
be qualified
air hood.
do not assure
over a period
on or in the epilayers
filtered
particles,
tained
over time,
They
varied
a general
34, InP 6, and InGaAs
as possible
duced
grown
Also,
the course
late contamination
were
were
the inevitable
is to be expected
samples
After
parameters
Samples
and there were
and substrate
liness
high purity material.
during
when
loading.)
the entire
the sample
motion,
tube,
to be re-
vibration,
or air
movement.
The expected
direct
proportional
and group III alkyl partial
characterization
rates
phase,
pressure
relationship
[5] was observed
as can be seen in Fig. 23.
seen in this study was 0.01 to 0.1 pm/min.
ditions
were
between
kept relatively
constant
early
growth
in the GaAs
The range
for GaAs.
for InP layers;
rate
of growth
Growth
the typical
congrowth
60
rate for this material
type.
No attempt
10:1)
[V]/[III]
[V]/[III]
ratio
was
0.05 ~m/min.
was made
ratios.
to grow p-type
No extensive
or growth
All undoped
material
epilayers
using
effort was made
temperature
for the binary
were n-
low (below
to optimize
the
compounds.
The
typical literature values [5,77] of [V]/[III] = 20-50 and T = 650-700 °c
seemed
to be good for GaAs
growth.
InP required
ratio
to compensate
for P2 evaporation
Table
I for typical
InP growth
Good morphology
microscopically.
tube was cleaned.
better
crystals
improved
Pinholes
of varying
specular.
areas
ing/rinsing
or contamination
noted
approaching
erature
those
in growth
could
aGC samples
due to polar
scattering,
deformation
purity
mobility
13
limit
procedures
etch-
purity
be grown with properties
The Hall mobility
in Fig.
The limits
potential,
dashed
concentration
14
of early
samples,
] of TMGa) had a low mobility (n
impurity
line for the ionized
4 l/min. H2'
=
temp-
of mobility
and ionized
-3
-2
versus
24, along with
cm ,while the lowerone represents5 x 10
[1.25 x 10
or poly-
and gas source
to an impurity
670°C,
growth
due to poor
5 x 10
=
and
tube.
for comparison.
The upper
corresponds
GaAs 7 (T
See
but the
swirl patterns
are plotted
sample
round
observed,
samples,
routinely
high purity
[78].
were
Occasionally,
a well-known
are noted
II).
visually
the outer
density
of very pure material.
data for two
scattering
whenever
in the growth
GaAs epilayers
in Chapter
as observed
could be seen on inferior
the improvements
above,
obtained,
Morphology
crystalline
With
(mentioned
[V]/[III]
conditions.
was generally
were
a higher
of
-3
em
Typical
.
[AsH3]/[TMGa]
16
8 x 10
im-
=
20,
-3
em
).
However,
61
OMVPE
T
-.
GaAs
= 670°C
=
[AsH3]/[TMGa]
H2 Flow
=4
22
l/min.
s::
-EE
:t.
........
t1.1
E-<
0.05
::I:
0.04
a
c:x:
0.03
0.02
0.01
0.5
1.0
1.5
[TMGa]
FIG. 23.
OMVPEGaAs GrowthRate.
2.0
2.5 x 10
-4
62
-.
"
-
~
D£FO~MATION
>
POTENTIAL.
MOBILITY
N
:IE
%
--8
-
.
\'
."
/
\ ,,
>
.
. },
\
Stillman[78y
105
MOBILITY
\
~.
\
.
1'-
///
IONIZED
IMPURITY.
m
o
MOBILITY
~
..J
~
/'
/'
/'
/
:l
>
/
/
'POLAR
Y~
'~/..-
.
u
~"
~
/
/
/
-
'
I
~/
10."
~
.10
TEMPERATURE
FIG. 24.
'100
1000
T (K)
Hall Mobility for GaAs at Cryogenic
Temperatures.
63
-4
GaAs 55 (T = 640°C, 4 l/min.H2, [AsH3]/[TMGa]
TMGa,
with
the benefit
of a cleaner
growth
= 20,
[1.5
x 10
mobility
behavior
of
tube, H2 line and arsine
source) achieved a 77 K Hall mobility of nearly 105 cm2/vs.
the typical
]
for later GaAs
samples,
This was
as can be seen in
Table VII.
Resistivity values were calculated for 25 GaAs epi1ayers, the number of samples for which thickness was measured.
Resistivity did
not depend critically on growth conditions for the first 40 growths.
The average resistivity of these samples was 0.0224 ncm at room temperature,
ranging
0.0169 ncm,
and 0.02
ncm
from 0.007
to 0.03
ncm.
(The average
at 77 K was
ranging from 0.004 to 0.02 r2cm.) GaAs 53 had 0.15 ncm
resistivities
at 300 K and 77 K respectively.
The best
quality GaAs epi1ayers, GaAs 54-62, averaged 3 ncm at room temperature
(the range was 1-6
A comparison
epi1ayers
grown
(Resistivity
ncm
of Hall mobility
data are also given
InP layers were
of less
consistent
and are discussed
V.)
concentration
is presented
grown
than GaAs layers.
in Chapter
at 77 K).
in Table
for the OGC samples.)
intermittently
quality
ncm
and carrier
at OGC and elsewhere
undoped
13.
at 300 K, and 0.08-0.4
for InP
VIII.
Thirty-eight
over many months,
(InP 18-25 were
and were
p-type
The best of the early samples
was InP
The last batch, InP 42-5, were uniform and generally an improvement
over earlier samples and comparable to recent literature reports [28,31,
34,40].
(The highest reported mobilities [79,80] are often not re-
peatab1e [81].)
The room temperature mobility values for InP 32-40 were
the same as for later samples.
However, the liquid nitrogen mobility was
64
TABLE VIII.
Undoped OMVPE InP Electrical Characterization Results.
Sample
Hall mobility
2
300 K (cm jVs) 77 K
II
or
reference
Resistivity
2
300 K (em /Vs) 77 K
n
-3
(cm )
[40]
4260
29,800
. . .
. . .
2 x 1015
[41]
5000
70,000
. . .
. . .
8 x 1014
[28]
3400
28,500
[31]
3660
. . .
. . .
. . .
2 x 1016
[34]
3670
4820
13,100
0.54
0.60
2 x 1015
2150
5220
0.20
0.10
1 x 1016
2100
10,000
0.30
0.10
7 x 1015
OGC:
InP 13
InP 32-40
InP 42-5
a
a
a
All data are averages
of the numbered
samples.
65
-
14,000
>. 12,000
"N
E 10,000
-0
8000
>.....
-
...J
- 6000
OMVPE
0
InP 43
m
4000
...J
...J
2000
<:[
::r:
0
10
100
1000
TEMPERATURE
FIG. 25.
about half,
Hall mobility for
probably
due
InP 43, in Fig.
impurity
scattering
presumed
to contribute
strates
etched
substrates
to
with n
inferior
=
8 x 10
most of the
themselves
phosphine
15
Hall mobility
[82].
impurities
probably
for
ionized
The TMIn source
[40,83].
is
The InP sub-
less pure than the GaAs ones.
yet not unexpected,
As the OGe crystal
-3
cm
source.
more slowly than GaAs substrates,and the InP
were
did not come as easily
showed improvements
OMVPEInP..
25, shows the expected
and rinsed
was disappointing,
the
(K)
that growth
of high quality
It
InP
as it did for GaAs.
growth
for both
from GaAs 11 and GaAs 61.
"learning
curve"
GaAs and InP.
These
samples
progressed,
PL data
Shown in Fig. 26 are data
were
ditions, except for temperature (GaAs 11: T
=
grown under
similar
755 °e, GaAs 61:
T =
con-
66
-
::J
21
I -1r-
z
"
OMVPE
1\
GoAs
77K
LLJ
Z
LLJ
<..>
z
LLJ
<..>
(/)
LLJ
Z
-.J
0
bl
I
0-
Go As II
rr \ ,
=>1
JMl'
750
800
tV""\J\Jl
850
WAVELENGTH
FIG,
26.
PL for
OMVPEGaAs.
GoAs61
900
(nm)
950
67
640 DC).
The emission
that of GaAs 11.
corporation
FWHM
GaAs
of carbon
for GaAs 61 is significantly
61 also had very good Hall mobility:
at lower growth
in comparing these samples [84],
quality
is improved
comparable
in these
corresponds
closely
Growth
over early
samples.
the intensity
clear,
16 vs. 10 meV.
1.404
bandgap
are available
which
temperatures.
data
The peaks
27 gives
The classic
from a variety
The effects
Recently,
Good general
[93,94,95].
has been
references
are centered
given
at 1.407
eV [57].
references
are
techniques
interest
for the growth
VI and
two samples.
is quite
eV (InP 46) and
of references
of InP at various
[49,50,73,86].
More
recent
in [71,72,74-7].
are discussed
in Si-implanted
[85].
liquid nitrogen
are contained
treatments
eV, which
in Tables
A variety
the photoluminescence
of growth
at 1.501
is
of GaAs at 77 K, 1.5076
close to the reported
of heat and surface
there
are centered
of emission
and the FWHM difference
of 1.4046
discuss
be considered
the 77 K PL data for these
The peaks
value
should
The intensity
bandgap
is comparable,
eV (InP 6), acceptably
temperature
samples.
for InP 6 and InP 46 were
Figure
Again,
temperatures
5090 and
it seems clear that the material
to the reported
parameters
I respectively.
less than
in [87-90].
InP for PETs
and characterization
[91,92].
of InP are
68
-
::J
21
(f)
z
I
OMVPE
InP
77K
- I
-1r-
w
u
z
w
u
(f)
w
z
::>
I
JI ,
\.
InP6
....J
0
,I
51
J:
Cl.750
800
850
WAVELENGTH
FIG. 27.
PL for Undoped OMVPEInP.
In
"-
900
(nm)
P 46
950
69
CHAPTERV
InP:Mg EPILAYERS
As a prelude
type layers were
reason
to the growth
grown
onto InP substrates.
for this approach.
established
capability
ly, information
layers
about
First,
reliable
regime
described
ternary,
of p-type
procedures
a number
There was more
it was desirable
for the growth
is far from plentiful.
Mg-doped
of the InGaAs
buffer
That is particularly
Much
than one
to have a well-
for p-type
in this chapter.
of p-
layers.
doping
Second-
of InP epi-
the case for the
of the material
here
has now been provided in published form [96].
In considering
ceptors
as a p-type
been used,
the possible
dopant
affect
also provides
a shallow
for providing
in InP, it may be noted
but is undesirable
can adversely
elements
device
in applications
performance.
monovalent
where
its high
Reported
doping
here
in this semiconductor,
on magnesium
is a photoluminescence
doping
However,
of GaAs is scanty
study of OMVPE
and
InP:Mg
[98,99].
over a wide
range.
This work
of MgIn
tions
literature
diffusivity
in the form MgIn'
has been reported as being much less mobile than zinc [97].
even the OMVPE
ac-
that zinc has often
Magnesium,
acceptor
shallow
has demonstrated
as a shallow
forms
complexes
acceptor,
with
that,
in addition
magnesium
deep-level
at higher
flaws.
pared to PL data for bulk InP:Mg [100-2],
to the expected
doping
role
concentra-
The data here
can be com-
Mg-implanted InP [103], and
70
molecular
beam
Table
rates
epitaxial
IX gives
for Mg-doped
depending
samples
on a par with
the surface
Figure
centrations
width [46]
derived
the CP2Mg
input
flux,
The broken
adduct
OMVPE technique,
mole/min.
[97].
of a near-quadratic
poration
see e. g.
emissions
[
97, 104]
highly
was generally
at OGC.
During
qualitative
sample.
PL
uniformity.
Net acceptor
correlation
in these
samples
line is a least
data from CP2Mg
a "typical"
slope
and Westbrook
molar
samcon-
with PL peak
as a function
squares
work.
into the crystal
in the
flow of 6 x 10-5
to the slope
Such a slope
flow alone.
of
fit to the
incorporation
of 1.82 is similar
of Mg incorporation
than on CP2Mg molar
on
is indicative
[CP2Mg]/[TMIn]
The mechanism
of incor-
is not postulated
here,
.
doped
observed
the Mg acceptor
seen
assuming
from the gas phase
achieved,
Figure 29 shows a log-log plot of the
achieved
dependence
The band-to-band
the most
from the reported
Our power-law
Growth
for four of the five InP:Mg
reference
line reproduces
of 1.85 in the Nelson
rather
to demonstrate
and the solid
data.
The morphology
PL results
concentration
per hour were
the best GaAs grown
(see Chapter III).
net acceptor
ratio,
fraction.
IX, and an undoped
were
for five samples.
of 3 or 6 microns
was scanned
28 exemplifies
pIes of Table
[100].
the growth parameters
on the TMIn mole
good, nearly
testing,
InP:Mg
peak at 1.41 is observed
sample.
in undoped
concentration
in all Mg-doped
samples,
Band-to-band
material,
was
transitions
but became
increased.
resulting
in Fig. 26 for all but
were
the only
less significant
A peak at 1.37 eV was
from conduction
band and/or
as
71
TABLE
IX.
Growth
Mole
InP:Mg
Sample
II
Conditions
and PL FWHM for
Fractions
TMln
-5
CP2Mg
-8
15
4.1
InP 23
-5
8.3 x 10
-8
4.6 x 10
17
InP 20
-5
4.1 x 10
-8
3.2 x 10
22
InP 25
-5
8.3 x 10
-7
2.3 x 10
25
InP 24
8.3 x 10
-5
1. 2 x
Growth temperature 600°C,
in the 55 rom diameter
10
4.6 x 10
-7
[PH3]/[TMln]
growth
tube.
InP:Mg.
PL FWHM
(meV)
InP 19
x 10
OMVPE
72
=
80, 8 1/min. H2
72
ENERGY (eV)
1.4
,
1.1
1.2
1.3
1.0
0.9
OMVPE
InP
77K
w
p = B X 1018cm-3
,
)--
InP 24
-
(/)
Z
i&J
Z.
L&J
"
. II
-
-- --
p=:3
--
X
InP 25
1017 cm-3
U
Z
I I, I L
InP 20
p = 1017 cm-3
i&J
Z
-
.::>
..J_
P=4
X 1016cm-3
InP 23
-U900
1000
1100
WAVELENGTH
FIG. 28.
PL for OMVPEInP:Mg.
1200
(nm)
I
1300
I
1400
73
1019
1018
I
I
r
101J
1016
I
/
I
I
/
I
r
I
I
I
I
/
r
/
I
/
/
I
I
/
I
I
/
/
.
/
y
I
I
InP 24
InP 20
InP 23
InP 19
10-4
10-3
10-2
[CP2 Mgl
[TMIn 1
FIG. 29.
Net Acceptor Concentration
[TMIn] Ratio.
Solid
data from [97].
line,
in InP:Mg versus
[CPzMg]/
this
study;
broken l1ne,
74
shallow
acceptor
IX for each
transitions.
sample.
the polarity
were weakly
The FWHM of this peak
The Mg-doped
of thermoelectric
n-type.
samples
were
power, whereas
The lowest net acceptor
is given
confirmed
"undoped"
in Table
as p-type
control
concentration
by
samples
we obtained
16
-3
was 2 x 10
cm ,and no minimum p-type threshold (as reportedby
Nelson and Westbrook [97])
There
was a shift
increased.
in the emission
For Mg doping
observed.
Moreover,
1.0 eV, becoming
grown
dominant
at this energy
InP:Mg
reference,
eV emission
change
the intensity
this represents
could
Since
InP:Mg
100],
a large
shift toward
In that work,
++ - +
be (Cd.
1.-Cd In ).
++
An explanation
to reports
-
+
.
InP:Mg
[105].
was varied,
higher
energy
transition,
near
of
[103], MBE-
In the latter
and the 1.30
(50 meV per decade
argued
that
and that the donor
(See also [106].)
is seen consistently
by annealing
occur in the dopJ.ngrange 10
occur.
sample.
it was extensively
in InP:Mg
(Mg.
1.-Mg In )
at 1.30 eV was
centered
InP:Cd
excitation
a deep donor-to-acceptor
and is removed
doped
was
appeared
be found by comparison
of laser
and not consistently
reports
PL band
as OMVPE-grown
the 1.3 eV emission
the transition
cited
additional
might
concentration
1017 cm-3, a peak
for the highest
[100], as well
showed
as acceptor
(at 77 K) from ion-implanted
in excitation).
complex
exceeding
a broad
for the 1.30 eV emission
emission
was observed.
in ion-implanted
grown by other
[103], I postulate
techniques
[97,
that it represents
-
MgIn .
17 18
That
~
-10
-3
em
and that the concentration
this emission
only begins
to
is consistent with the above
of MgIn must
be high
for it to
75
There
is more
74,74,100,103].
complex
literature
discussion
All authors
involving
attribute
a phosphorus
this defect have included iron [74]
to support
this emission
vacancy.
InP with high impurity content [75].
have no evidence
of the 1.0 eV peak
This
is seen
in InP
to an impurity
in In-rich
sion only
phosphorus
cies.
vacancies
and copper [70].
identification
appears
in the highest
over-pressure
As noted
nominally
are involved,
in Table
undoped.
during
LPE
The impurities identified with
In our case, we
of the impurity
other
to assume it is MgIn, with the transition being (Mg~n-Vp)+
phosphorus
[70,
~
than
MgIn'
If
then the fact that the 1.0 eV emis-
doping
growth
levels
is sensible
is maintained
V, the substrates
for these
since
to prevent
samples
a
vacan-
were
76
In
This chapter
epi1ayer
growth
portion
discussed
matched
at
An unexpected
in detail.
data
and
the
Finally,
are
describe
those
at
are
samples
of this
absorption,
important
InGaAs/GaAs
in InGaAs/lnP
presented
for
Hall
with
properties
defect
mobility,
both
is
1attice-
various
for
InGaAs
The first
including
characterization
results
of the
OGC.
growth,
In 1 -x Ga xAs/lnP
for
aspects
was found
optical
given
several
ternary
defect
range,
Next,
InGaAs/lnP.
data
of
undertaken
efforts
composition
resistivity
These
discussion
early
results.
a wide
Ga As
x
and characterization
discusses
growth
over
presents
1-x
and
x values.
cryogenic
and
room temperatures.
As mentioned
occurs
in Chapter
by a parasitic
a white
polymer
which
of this
reaction
is
proper
growth
first
step
grow
ternary
These
would
layers
not
small.
InGaAs
reactor
tube
a loss
of TMIn as compared
reaction
of TMIn with
PH3.
deposits
on the
of the
the
major
conditions
towards
II,
for
finding
unknown
walls
factor
lattice-matched
the
correct
on GaAs substrates
be lattice-matched
in the
ternary
[TMIn]/[TMGa]
with
to GaAs,
9, 10, and 11 were all
This
to TMGa
reaction
reactor.
The extent
determination
input
The
value
percentage
but
mismatch
grown in the
of
material.
a small
the
forms
was to
of
In.
would
be
55 mm diameter
-4
with
5 l/min.
of TMGa, and [Z.Ol x 10
-5
HZ' T
= 675°C,
] of TMIn.
[V]/[III]
=
A one micron buffer
20,
[1.80
layer
x 10
of GaAs
]
77
was grown
prior
to the InGaAs
composition
near
too large.
The value
ternary
would
InO.lGaO.9As
0.9 was chosen
separate
PL data from all three
sample
1.384
a 20 meV FWHM.
variation
i.e., x
= -0.538
Thus
is shown
with
meaning
no parasitic
The peak
the formula
TMIn
purposes,
constant.
radically;
were
loss
above,
factor
of the
identical.
is centered
1/2
,one
A
at
input was equivalent
be 0.53/0.47
calibration,
could be expected
flow rates,
certainly
=
deduces that x
=
1.13
[55],
0.92.
to 10%,
reaction.
To
assuming
a 20% loss of TMIn.
that 20% of the TMIn
the temperatures,
and that factor would
were not
for the In l-xGaxAs
to 1.36 assuming
is a reactor-dependent
A different
same equipment
would
but increases
described
effects
so that the PL peak
30.
- 0.544)
the required
reaction,
give a
at 77 K given by Wu and Pearson
[TMIn]/[TMGa]
The calibration
OMVPE
g
reaction
that 20% of the TMIn was lost to the parasitic
grow InO.53GaO.47As,
growth
Using
+ 0.943(2.l2E
would
at 77 K were nearly
in Fig.
composition
to get 8% indium,
parameters
from the GaAs peak.
samples
representative
bandgap
These
if the parasitic
=
of x
be clearly
eV, with
growth.
is lost
for
not a universal
even with
etc.,
be different
the
to be changed
using
another
facility.
In fact,
severe
the loss factor
than estimated
InP substrates
as deduced
required
with x close
above,
a ratio
scientific
above,
those
when
out to be considerably
In l-xGa xAs layers were
to 0.47.
Rather
layers nearest
slightly
calibration
turned
in excess
were
than
grown
onto
[TMIn]/[TMGa]
to lattice-matching
of 2:l!
more
onto
The best laid plans
thus frustrated--but
= 1.36
InP
for a
the indium-to-gallium
78
OMVPE
InGoAslO
77K
-.
-ci
::J
--1
I-
en
z
LLJ
IZ
LLJ
U
Z
LLJ
U
en
LLJ
Z
=>
...J
0
I-
0
I
CL
600
800
1000
WAVELENGTH
FIG.
30.
PL for
the
InGaAs 10 at
extent
of
the
77
K.
1200
(nm)
This epi1ayer
TMIn + PH3 parasitic
1400
was grown
reaction.
to
test
79
ratio was eventually
determined
The morphology
appearance
reduction
lattice
of the InGaAs/GaAs
to GaAs
temperature
on quite
layers.
and 2680
epilayers
Hall mobility
cm2/Vs
mobility
was good,
similar
3700 em2/Vs
temperature.
may be explained
The minimum
grounds.
averaged
at liquid nitrogen
at low temperatures
mismatch.
empirical
in
at room
The
by the significant
for these
samples
was 2500
2
em /Vs at 100 K.
Early
the above
ranging
InGaAs
ternary
TMIn depletion
was obtained
grown with
[TMIn]/[TMGa]
of these
The highest
samples,
of InGaAs
energy
peak
is the band-to-band
(if present
as involving
53 shown
until
As can be noted
between
in Fig.
optical
Typical
31.
Three
emission
of samples
(InGaAs
later
were
29-57).
of these was the
peaks
are present.
was so broad
data
was
= 2, but no PL
the lowest
that it was not
in Fig.
which
all,
Two samples,
1.60 and 1.77
absorption
deep-level
series
was usually
This
samples
The two large peaks
one or more
An extensive
(#1), which
transition.
layer,
[TMIn]/[TMGa]
PL was obtained.
at all) in most
transition
pretation.
between
buffer
good luminescence.)
then grown near
from them either.
77 K spectrum
bandgap
on an InP OMVPE
in obtaining
27 and 28, were
For most
on the
(Most, but not
grown
employed
1-26) based
from 1.1 to 1.6, but showed no PL at all.
were
estimate
(InGaAs
of [TMIn]/[TMGa]
to be important
InGaAs
on InP substrates
values
samples
found
samp1e~
in intensity,
and weak
interpreted
suggested
as the
that inter-
31 (#2 and #3) came to be seen
defects.
in Fig. 31, there was no variation
in energy
either deep-level emission with varying laser excitation.
of
If the tran-
80
/12
I
OMVPE
InGaAs 53
77K
= Laser Current
-
/13
~
-o
>I-
-
fJ)
Z
UJ
I-
Z
UJ
(.)
Z
UJ
(.)
fJ)
I
,
UJ
Z
= 20.5A
~
:)
-I
e
le
:I:
a..
111
1500
1600
WAVELENGTH
FIG. 31.
1700
1800
(nm)
Variation of InGaAs Defect PL with Excitation Light Intensity.
81
sitions
shift
were
donor-to-acceptor,
toward
The defect
buffer
higher
PL spectrum
layers.
InP:S
InP:Fe
was present
substrates.
65) was analyzed
substrate
clearly
too high,
marized
in Table
for samples
An impurity
by SIMS.
concentrations
These
given
Values
in this table
of the concentration
versus
depth
were
the substrate,
while
and surfaces.
Within
about
of two, no significant
were
a factor
observed
between
InGaAs "standard",
No Hg-imp1ant
tally
spurious.
tope,
since
in III-V materials,
tion
standard
Perhaps
listed
in InGaAs
is interpreted
on both the Sumitomo
suspected,
data were
and a sample
compared
to both the
by the SIMS analysis
represent
were
graphical
differences
averages
Measured
or within
at the interfaces
of the SIMS analysis,
65 and either
D.
the epi1ayer
observed
are
The data are sum-
in Appendix
within
was available,
the "Hg" signal
in both epi1ayers
represent
which
in impurity
the InP:Fe
was
levels
substrate
or the
the most
so the Hg is probably
tracked
an As-containing
and low in both
common
to its substrate
as an indication
that the defect
toiso-
substrates.
and expected
and the lack of any significant
65 as compared
than an impurity.
and without
OSU 1-29-7-86.
it is high
The elements
constant
transients
InGaAs
was
data given
concentrations
the accuracy
grown
for comparison.
X.
large
grown with
[107].
sample OSU 1-29-7-86.
but they are useful
generally
to see a measurable
of Zacks and Halperin
for samples
and the MBE-grown
The elemental
expect
from the theory
and was present
and InP:Fe
(InGaAs
energy
one would
contaminants
impurity
concentra-
or the "standard"
is structural
InGaAs
rather
82
TABLE x.
SIMS Results for MBE and OMVPE InGaAs/InP.
Element
MBE -EpilayersOSU 1-29-7-86
OMVPE
InGaAs 65
-Substrates OSU 1-29-7-86
InAsAs 65
C
2 x 1017
2 x 1017
1 X 1017
1 x 1017
0
3 x 1017
3 x 1017
5 x 1017
3 x 1017
Na
1 x 1015
2 x 1015
1 x 1015
4 x 1015
Mg
1 x 1016
8 x 1015
4 x 1015
4 x 1015
A1
4 x 1017
4 x 1017
2 x 1017
2 x 1017
Si
4 x 1016
4 x 1016
4 x 1016
2 x 1016
S
4 x 1015
5 x 1015
6 x 1016
1 x 1019
C1
4 x 1016
6 x 1015
4 x 1015
3 x 1015
Cr
2 x 1016
3 x 1014
3 x 1014
5 x 1015
Fe
6 x 1016
4 x 1016
2 x 1017
3 x 1016
Ni
3 x 1016
1 x 1015
2 x 1015
7 x 1015
Cu
9 x 1017
2 x 1018
2 x 1018
Zn
4 x 1017
6 x 1017
6 x 1017
3 x 1018
Ag
8 x 1016
2 x 1017
2 x 1017
2 x 1017
W
2 x 1018
2 x 1018
3 x 1017
2 x 1018
Hg
4 x 1021
6 x
5 x 1019
2 x
1021
1018
83
The defect was present
tainable.
These
samples
in all samples
are listed
in Table
and intensity
of each of the three PL peaks
Table
XI data
(with corrected
Here,
the data from two Ga-rich
nearly
lattice-matched
with varying
amounts
from this figure:
approached
sample
is highest
for the In-rich
increasing
linear
trend
parasitic
is assumed
scribed
sample,
Several
and become
and a
features
samples);
sample,
reduction
can be identified
is
for the In-rich
in lattice-matched
and 3) a linear
emerge
as lattice-matching
one peak
to be due to variation
0.69-0.70
at
and
of the bandgap
(variation
from a
in the amount
of
the
is not as broad
Iron
impurity-related
Of the 16 InGaAs
substrates.
54 had very
of the emission
defects
samples
InGaAs
defect"
as in those
Of these,
The most
eV has been observed in InGaAs before, de-
as an "intrinsic
here
32.
in Fig. 32.
reaction).
only
intensity,
is given
of the
to show how the PL changed
peaks merge
resolvable
ob-
the energy
A distillation
for the lattice-matched
sample;
[TMIn]/[TMGa]
Emission
InGaAs
[TMIn]/[TMGa].
are sometimes
2) the PL intensity
with
of input
side,
seen.
one In-rich
are reproduced
PL data were
XI along with
intensities)
samples,
1) the two defect
(two peaks
lowest
PL peak
sample
from the Ga-rich
from which
reports,
While
in Table
XI, 3 were
65 had much
53 had slightly
low emission
in each of these
above
larger
average
intensity.
samples
that can be said regarding
the PL observed
the defect may be related.
have also been observed
listed
InGaAs
[108,109].
in InGaAs
grown
on Fe doped
than average
emission
The spectral
correlated
with
the Fe doped
[109].
emission
intensity,
and
distribution
that of Fig.
substrate
as a
84
TABLE XI.
InGaAs
PL Peak Energies for OMVPE In
Peak fI1
[TMln]
[TMGa] Energy
Intensity
(eV)
(a)
Sample
/I
l -x
Peak
Energy
(eV)
Ga As.
x
Peak
/12
Intensity
(a)
Energy
(eV)
/13
Intensity
(a)
39
1.60
0.816
63
0.775
177
0.756
115
40
1.70
0.820
25
0.731
263
0.710
300
41
1.70
0.813
50
0.732
431
0.708
269
42
1.60
0.821
53
0.794
44
0.775
28
43
1.60
0.827
13
0.719
438
0.695
363
44
1.65
0.816
44
0.754
311
0.730
255
46
1.63
0.818
59
0.786
56
48
1.72
0.820
80
0.738
55
0.718
70
49
1.72
0.816
11
0.734
24
0.718
16
53
1.72
0.810
63
0.750
313
0.735
288
54
1.74
0.810
14
0.780
18
55
1.75
. . .
. . .
0.750
119
58
2.05
0.795
56
0.737
1000
59
2.09
0.777
18
0.734
27
O.704
15
62
2.19
0.787
19
0.734
110
65
2.07
0.787
325
0.742
4350
0.729
2500
alntensity values are relative to: 1.5 mm slit width, 5000 gain,
10 mV lock-in, 26.5 A laser, Si filter, 10 V full scale on X-Yo
In most cases, the FWHM was not resolvable.
-5
Growth conditions were: [9.23 x 10 ] of TMln, [V]/[III]
55 except
for InGaAs 39 and 40 in which 30 was used, T = 650°C for InGaAs 41-
=
57, T
=
630°C
for InGaAs
39-40,
T
=
600°C
for InGaAs
58-66,
no InP
buffer was grown on InGaAs 47 & 49, a 2 pm buffer was grown on InGaAs
53, one micron buffers were grown on all others, 4 l/min. H2 flow was
used in the rectangular growth tube.
85
-.
~
o
OMVPE
Inl_XGoXAs
77K
~
to(/)
Z
ILl
tZ
ILl
U
Z
ILl
U
(/)
ILl
Z
:E
::)
I
o
t-
O
::L
Cl.
1.6
1.7
1.8
1.9 2.0
2.1 2.2
INPUT
FIG. 32.
Variation of
In
! -x
[T MIn]
[TMGo]
Ga As PL
x
withx.
86
contributor
diffusion
grown
Peak
activated
an emission
peak heights
#2 was always
significantly
40.
This
than InGaAs
increase
in peak
#2 intensity
due to recombination
condition
in Table
XI.
level
defect
the low energy
There
deviation
because
not
#3, with
lower temperature
[V]/[III]
ratio
were otherwise
significantly
data
to variability
the same.
a large
instead
technique
light
in surface
in the data
were not available
enough
(630
of the laser
Variation
layers were not thick
of
(30:1 versus
40 may
surface-sensitive
thickness.
absorption
consistent.
the exception
and the fact that most
was some suspicion
that the ternary
to InP.
is given by Figs.
MBE sample,
might
However,
peaks were
for this
to measure
a in
regime.
LPE material
shown, which
0.41.
is that Fe
in samples
41, or InGaAs
PL is a very
contributed
from two InO.53GaO.47As
dard"
layers
ratio may have caused
for InGaAs
from lattice-matching
the case
denotes
[V]/[III]
mechanisms
Corroborative
lower
conditions
in less than one micron
may have
than
at a slightly
a somewhat
and/or
as an anomaly.
is absorbed
higher
41, but growth
temperature
be regarded
of the two defect
sample was grown
650 °C) and with
The higher
not
InGaAs
that was also present
The relative
°c versus
deep
PL seen in these
on this type of substrate.
InGaAs
55:1)
to the deep-level
33 and 34.
samples
reported
Striking
evidence
that this
Figure
33 shows
77 K PL data
not grown by OMVPE.
by Pearsall
OSU 1-29-7-86.
Also,
be taken as indicating
the coincident
defect was due to a
[56], while
Of these,
300 K PL data in Fig.
(a)
(b) is the "stan-
data from InGaAs65
an In-rich
is
are
composition
of x
34 for the same
=
87
:J
21
InO.53Go0.47 As
I
77K
en
z
w
z-
-1r-
w
u
z
w
u
en
w
-z
::>
....J
0
0
:I:
a..
1500
1600
WAVELENGTH
FIG. 33.
1700
1800
(nm)
PL for InO.53GaO.47As at 77 K. a) data from
Pearsall [56], b) OSU 1-29-7-86, c) InGaAs 65.
88
::3
I InO.53GaO.47As
300 K
>-1
(f)
z
w
z
-1r-
w
u
z
w
u
(f)
w
z
-
::>
-1
0
0
J:
a..
1500
1600
WAVELENGTH
1700
1800
(nm)
FIG. 34. PL for InO.53GaO.47As
at 300 K. a) data from
Pearsall [56], b) OSU 1-29-7-86,c) InGaAs 65.
89
three
samples
support
the conclusion
that all of these
layers
are
lattice-matched.
As a further
65 was analyzed
Within
test for lattice-matching,
by AES.
the experimental
same composition
derive
We must
the OMVPE
InGaAs/InP
t ch
samples
study was made
tion by optical
[TMIn]/[TMGa]
is of the
that was used to
The accumulated
that structural
which
evidence
to x and hence
defects
thus
to
are present
are not due to impurities
of composition
absorption.
were measured
in a log-linear
bandgap
then conclude
of
in Fig. 35.
(2%), this sample
(i.e. the sample
in Fig. 35).
of InGaAs
in
or lattice
alone.
Further
samples
scale
are presented
of the method
Fig. 32 for the scaling
bandgap.
misma
error
results
as OSU 1-29-7-86
the vertical
confirms
These
the composition
plot
The absorption
at 10 K.
bandgap
coefficients
Data from four of these
in Fig. 36.
was obtained
through
The steepest
for lattice-matched
determina-
of eleven
are presented
drop in a below
samples.
The value
the
of a rose
-1
to about
6000 em
near
The bandgaps
37, along with
the bandgap,
determined
the bandgap
was band-to-band
are shifted
20 meV higher
rect
for the increase
77 K to 10 K [56].
are underlined
Absorption
by absorption
found
PL emission
regardless
by PL
of composition.
at 10 K are plotted
(assuming
the highest
emission).
In this figure,
than the actual
77 K measured
in bandgap
Five samples
expected
were
when
tested
InGaAs
in Fig.
energy
the PL data
value
to cor-
is cooled
by both methods,
from
and these
in the figure.
gave
somewhat
more
reliable
data for these
samples.
90
100
90
I
OMVPE
InGaAs 65
BOr
70
r-
z
UJ
u
a:
UJ
DU
60
50
H
c
.....
40
et
30
20
10
DEPTH FROMSURFACE (A)
FIG.
35.
AES Profile
for
InGaAs
65.
91
20,000
10,000
I
I
OMVPE
In'_XGaxAs
IOK
9000
8000
7000
-
-
6000
E 5000
I
-u
4000
tS
3000
.
.
.
I
I
/J
J
I
1000
0.70
.
I
,
2000
I
.
I
,6InGoAs
InGoAs 65
X=0.47
0.75
InGoAs 53
X=0.51
0.80
57
X = 0.50
0.85
Energy
FIG. 36.
0.90
0.95
(eV)
Absorption Coefficientfor OMVPE In 1 Ga As at 10 K.
-x
x
1.00
92
0.90
-
OMVPE
InGaAs Sample Given
~
Opt ical Absorption at 10 K
6 Highest PL Energy at
77K + 20 meV
-
.
>CD
o
_ 0.85
c
CL
<I
C)
o
Z
<I
m
o
LaJ 0.80
r<I
-r:E
28
CJ)
LaJ
0.75
2.3
2.2
2.0
2.1
1.9
1.8
1.7
1.6
[TMIn] INPUT RATIO
[TMGa]
FIG. 37.
Bandgap Estimates
Ratio.
for OMVPE InGaAs vs.
[TMIn]/[TMGa]
93
This
is particularly
the a data are quite
line of best
2.0 < [TMIn]/[TMGa]
seen in the range
orderly,
fit is drawn
yet the PL data are scattered.
through
the absorption
the two methods
suggest
are not lattice-matched
moved
from their
mismatch,
while
different
substrates,
PL samples
are still
was obtained
-0.948).
The a samples
attached
that
are re-
due to lattice
to their
substrates.
of this effect.)
for samples
with E (10 K)
g
=
0.81 eV
which corresponded to an input [TMIn]/[TMGa] ratio of 2.0
[56,59],
The flattening
values
of the absorption
may be evidence
tion of the parasitic
quantify
data curve
of increased
at high
depletion
reaction--but
[TMIn]/[TMGa]
of [TMln]--an
there are not enough
acce1era-
data to
this.
The growth
conditions
Fig. 37 are given
fully
=
a straight
for compositions
and so have no strain
(See Kuo et a1 [110] for a discussion
Lattice-matching
bandgaps
is not unreasonable.
A curved
data, while
least-squares fit line seems adequate for the PL data (r
That
< 2.1, where
removed
these were
in Table
from InGaAs
taken with
and PL peak energies
XI.
Note
for the samples
that the substrate
27 and InGaAs
could not be
50, so the absorption
the InP substrate,
in
data for
and its effect was subtracted
out.
The InGaAs/lnP
matched
samples,
epi1ayers
spots.
InGaAs
good morphology.
such as InGaAs 65, were highly
grey haze was present
ticu1ar1y
had mostly
on samples
28, 57, and 60.
which
were not
InGaAs
The 1attice-
reflective.
A light
lattice-matched,
27 and 56 had large
par-
dark grey
94
A comparison
InGaAs
can be made
65 and the "standard"
of the 10 K absorption
InO.53GaO.47As
As can be seen in Fig. 38, these
86.
dicating
ature
lattice-matching
for
MBE sample, OSU 1-29-7-
data are quite
for the OMVPE
InGaAs a data are given
coefficient
sample.
in the figure
close,
again
Comparable
also
[69].
in-
low temper-
The near-band-
-1
gap peak absorptionof 5800 cm
for InGaAs 65 is the same as the
(For the sample in [69],
Zielinski et a1 value.
mated
to be 0.82 eV rather
ditional
range
reports;
data
Figure
[69], labelled
[Ill],
(b), and of Burkhard
appear
to be too high,
error.
Above
in either
and a(lO
[112] or InP
splitting
was not observed
[113].
as those
et al room
of Humphreys
The Burkhard
of a thickness
Above
et al
et al data
measurement
in InGaAs
as it does
1.3 eV, the curves a(300
(=
at (E
0.35
Enhanced
g+
eV
0.1 mm or the temperature
of the unknown
is not explored
either with
lowered
~),
for
where
~
K)
occurred
of Humphreys
here.
[114])
at 0.1 eV
et aI, but
No strong
the monochromator
to 7 K, perhaps
denotes
InO.53GaO.47As
absorption
for both our data and those
was observed
Absorption
spectrum
parameter
the cause of this absorption
presence
the result
in this study.
the bandgap
absorption
(a), as well
a does not rise as fast
in the absorption
spin-orbit
above
the Zielinski
K) merge.
A step
the
39 shows
are two ad-
65 is over a broader
et al [61], (c).
perhaps
the bandgap,
GaAs
At 300 K, there
but the data from InGaAs
than any of those.
temperature
than 0.81 eV.)
the bandgap was esti-
exciton
slits narrowed
because
to
of the
defect.
data at 300 K for InGaAs
samples
of varying
composition
95
1nO.53 GOO.47As
10,000 r
t;,.
-I
-
l-
/:).
I
.
E
()
I
t1
1000
/:).
"
. ..
L_-Zielinski
et 01 [69].
. InGoAs65,
0.75
0.80
0.85
Energy
FIG.
38.
Near-bandgap
Absorption
Coefficient
10K
1-29-7-86,
/:). OSU
100
0.70
5K
0.90
0.95
(eV)
for
InO.53GaO.47As.
10K
1.00
96
100,000
InO.53 G00.41 As
300K
~
10,000
I
E
u
1000
.
100
0.7
0.8 0.9
1.0
1.1
Energy
FIG.
39.
Absorption
Coefficient
a) data from Zielinski
Humphreys et al [111],
[61],
data points
are
I.2
I.3
1.4
I.5
(eV)
for InO.53GaO.47As
at 300 K.
et al [69], b) data from
c) data from Burkhard
et a1
InGaAs 65 from this
study.
97
are shown
in Fig. 40.
a, regardless
fortunate
parameter
large
which
lattice
that the energy
scale
InGaAs/InP
that the absorption
could be measured
samples
were
coefficient
reliably
the same
on this fig-
energy
range.
ambiguous,
it was
turned
on samples
out to be a
with
relatively
mismatch.
Absorption
Fig.
Note
had nearly
than that of Fig. 36, to show the higher
the PL data for OMVPE
indeed
1.0 eV, all samples
of composition.
ure is broader
Since
Above
data were
41, for sample
also taken at 77 K.
InGaAs
K data for all samples,
and 2) less structure
59.
except
above
An example
Data at 77 K were
appears
coincident
for 1) a 20 meV shift near
the bandgap
in
with
10
the bandgap,
(e.g. no threshold
at E
g + 0.1
eV).
The Hall mobility
grown
was measured
on semi-insulating
InP substrates.
measurement
for all such InGaAs
and carrier
concentration
and the OGe samples.
light
of the brief
mobility,
but still
The best InGaAs
which
compare
sample
this sample,
a log-log
in Fig. 42.
The growth
those
for InGaAs
samples.)
Table
The OGe results
time available
Samples
(Poor contacts
for the best samples
ture,
techniques.
for some of the InGaAs
XII lists Hall mobility
reported
obtained
65 (as listed
in
InGaAs
have much
InP and GaAs
vs. temperature
for this sample were
in Table
II), except
growth
reduced
at 300 K.
in this study was InGaAs
plot of Hall mobility
conditions
in the litera-
to develop
are not lattice-matched
with
prevented
are encouraging
(six months)
favorably
epilayers.
66.
For
is given
the same as
that at the end
98
100,000
OMVPE
In l-xGoxAs
300K
InGaAs62
X = 0.46
InGa
10,000
As 59
X=0.47
-
I
E
u
InGaAs 53
X=0.51
1000
InGaAs 60
X= 0.475
100
0.7
0.8
0.9
1.0
Energy
FIG.
40.
1.1
I.2
I.3
1.4
I.5
(eV)
Absorption Coefficient for OMVPE In l -x Gax As at
300 K.
99
100.000
OMVPE
loGo As 59
77K
10.000
-
I
-E
,I
u
tS
L
I
I
I
I
I
I
I
I
1000r
100
0.7
0.8
0.9
1.0
Eneroy
FIG. 41.
1.1
1.2
1.3
1.4
1.5
(eV)
Absorption Coefficient for InO.S3GaO.47As at
77 K.
100
TABLE XII.
Hall Mobility and Carrier Concentration for In1_xGAs/lnP.
Sample/Reference
Hall Mobility
300 K
Lattice-matched (x
=
(cm2/Vs)
n
-3
(cm )
77K
0.47):
Pearsall et aI,
Thomson/CSF
[115]
10,000
26,500
1 x 1016
Goetz et aI, Thomson/
CSF [59]
10,400a
43,400a
4 x 1015
79,000
8 x 1014
64,000
5 x 1014
Saxena
Carey
et aI, HP
et aI, HP
[116] 10,600
[117]
. . .
OGC:
InGaAs 65
6500
7600
2 x 1015
InGaAs 66
9000
26,500
6 x 1015
Not lattice-matched (x
=
0.49):
OG:
InGaAs 47
3000
4000
1 x 1016
InGaAs 50
5300
10,600
5 x 1015
a
average
of four samples.
101
40,000
OMVPE
30,000
In GoAs 66
-. 20,000
U)
>
.......
N
-
E
u
:I:
:1..
10,000
8000
L
40
60
80 100
TEMPERATURE
FIG. 42.
200
300
400
(K)
Hall Mobility for OMVPE InO.53GaO.47As.
102
of the buffer
layer
begin
growth,
ternary
600 °c within
temperature
reduce
fell below
phosphorus
=
was reduced
to the final
growth
to 450°C
to
temperature
of
The PH3 flow was maintained
500°C.
evaporation
The purpose
epilayers
to the resistivities
0.315 Q em, and for InGaAs
until
of this procedure
from the buffer
for the InGaAs/lnP
at 300 K, similar
InGaAs 65, P
the temperature
then raised
two minutes.
Resistivity
0.4Qcm
growth,
the
was to
layer.
was typically
obtained
0.1 to
for InP.
66, 0.113 Q em.
For
103
CHAPTER VII
CONCLUSIONS
growth of the III-V compounds InP, GaAs, and InGaAs
Epitaxial
using
OMVPE
each.
Factors
provements
influencing
of GaAs,
based
sources
Magnesium
InP in addition
depends
though
Also,
InP crystals
reflected
for measuring
the absorption
for energies
10
larger
The final
of
imof
series
of material
of this work has been
trimethy1-
showed new and interesting
several
shallow
to 10
mechanisms
on composition,
coefficient
defect
acceptor.
complexes
Mg doping
in
was ac-
cm
absorption
of optical
edge clearly
of Mg incorpora-
19-3
the optical
The strength
the absorption
the quality
growth with
at least two deep level
to the expected
strongly
With
the quality
thrust
of OMVPE
16
developed.
which
formed
reports
discussed.
purity,
approached
50 samples
few.
over the range
Methods
were
which
in at least
improved.
The primary
for which
of Mg doped
photoluminescence
.
comp11shed
progressively
InP, and InGaAs
have been
A series
tion.
and source material
in the literature.
the ternary,
resulting
the growth have been
of each compound
of growths
reported
accomplished,
in technique
epi1ayers
with
has been
coefficient
absorption
of InGaAs
is not a parameter
as seen in the data here
shifts with
becomes
composition
independent
(a1-
change).
of temperature
than 1.3 eVe
Optical absorptionproved useful in demonstratingthe growth of
104
lattice-matched
InGaAs/lnP.
This was supplemented
scence
and AES characterization.
InGaAs
layers
did not depend
and electrical
InGaAs
when
This
properties
grown
study has shown
by OhVPE.
deficiencies
strongly
retaining
should
the economies
of
the morphology
improved
epitaxial
device
The data presented
provide
properties
on composition,
that InGaAs
and optical
in the OMVPE
photolumine-
dramatically
for
was achieved.
process;
MBE-, LPE-, and VPE-grown material.
techniques
the optical
did; Hall mobility
lattice-matching
for both electrical
While
with
material
layers
applications
of good quality
can be readily
here does not indicate
favorable
comparisons
inherent
are made
to
Further refinement of growth
of even higher
perfection
while
of scale and the relative
simplicity
of
OMVPE.
105
REFERENCES
[1] V. Narayanamurti,
"Artificially
structured thin-film
and interfaces," Science 235, 1023 (1987).
materials
[2] F. Capasso, K. Mohammed, and A. Y. Cho, "Tunable barrier heights
and band discontinuities
via doping interface dipoles: an
interface engineering technique and its device applications,"
J. Vac. Sci. Technol. B 1, 1245 (1985).
[3] P. D. Dapkus, "Metalorganic
chemical vapor
Rev. Mater. Sci. 1l, 243-69 (1982).
deposition,"
Ann.
[4] Proceedings
of the Third International
Conference on Metalorganic
Vapor Phase Epitaxy, edited by G. B. Stringfellow,
J. Cryst.
Growth 12 (1-3)(1986).
[5] M.J. Ludowise, "Metalorganic
semiconductors,"
J. Appl.
chemical vapor deposition
Phys. 58, R3l (1985).
of III-V
[6] G. B. Stringfellow,
"OMVPE growth of III-V semiconductors,"
in
Semiconductors
and Semimetals, Vol. 22A, edited by W. T.
Tsang (Academic, New York, 1985) pp. 209-60.
[7] S. D. Hersee and J. P. Duchemin,
deposition," Ann. Rev. Mater.
"Low-pressure
chemical vapor
Sci. 1l, 65-80 (1982).
[8] G. H. Olsen and T. J. Zamerowski, "Vapor-phase growth of
(In,Ga)(As,P) quaternary alloys," IEEE J. Quantum Electron.
QE-17, 128 (1981).
[9] R. H. Moss and J. S. Evans, "A new approach to MOCVD
GaInAs," J. Cryst. Growth 55, 129 (1981).
of InP and
[10] H. M. Cox, "Vapor levitation
crystal growth," J. Cryst.
in epitaxial
epitaxy: a new concept
Growth 69, 641 (1984).
[11] R. Solanki, C. A. Moore, and G. J. Col1ins, "Laser-induced
chemical vapor deposition,"
Solid State Technol. 28, 220
(1985).
[12] J. J. Yang, R. P. Ruth, and H. M. Manasevit, "Electrical properties of epitaxial indium phosphide films grown by meta1organic chemical vapor deposition," J. App1. Phys. 52, 6279
(1981)
--
106
[13] H. M. Manasevit and W. 1. Simpson, "The use of metalorganics
preparation
of semiconductor materials,"
J. Electrochem.
120, 135 (1973).
in the
Soc.
[14] Proceedings of the First International Conference on Metalorganic
Chemical Vapor Deposition, edited by J. F. Bonfils, S. J.
Irvine, and J. B. Mullin, J. Cryst. Growth 55 (1981).
[15] Proceedings
of the Second International
Conference on Metalorganic
Vapour Phase Epitaxy, edited by J. B. Mullin, J. Cryst. Growth
68 (1)(1984).
[16] B. J. Baliga and S. K. Ghandhi, "Growth and properties of heteroepitaxial GaInAs alloys on GaAs substrates using trimethylgallium, triethylindium,
and arsine," J. Electrochem.
Soc. 122,
683 (1975).
[17] B. J. Baliga, R. Bhat, and S. K. Ghandhi, "Composition dependence
of energy gap in InGaAs alloys," J. Appl. Phys. 46, 4608 (1975).
[18] C. B. Cooper, M. J. Ludowise, V. Aebi, and R. L. Moon, "The organometallic VPE growth of GaAs l Sb using trimethylantimony
and
-y y
Ga
l-xIn xAs using
trimethylarsenic,"
J. Electron.
Mater.
-9,
209
(1980).
[19] J. P. Noad and A. J. Springthorpe,
by organometallic
pyrolysis
Mater. ~, 601 (1980).
"The preparation
for homojunction
of Ga l-xInxAs
LEDs,"
[20] R. Dupuis, R. Lynch, C. Thurmond, and W. Bonner,
by MOCVD," Proc. SPIE 323, 131 (1982).
J. Electron.
"Growth
[21] T. Fukui and Y. Horikoshi, "Properties of InP films grown
OMVPE method," Jpn. J. Appl. Phys. 19, L395 (1980).
[22] M. Ogura,
MESFET
of InP
by
K. Inoue, Y. Ban, T. Uno, M. Morisaki, and N. Hase, "InP
grown by MOCVD," Jpn. J. Appl. Phys. 21, L548 (1982).
[23] M. Ogura, M. Mizuta, N. Hase, and H. Kukimoto, "Deep levels
grown by MOCVD," Jpn. J. Appl. Phys. 22, 658 (1983).
[24] E. P. Menu, D. Moroni, J. N. Patillon, T. Ngo,
"High mobility of two-dimensional
electrons
heterostructures
Growth~,
grown by atmospheric
920 (1986).
in InP
and J. P. Andre,
in Ga l-xIn xAs/InP
pressure
MOVPE,"
J. Cryst.
107
[25]
J.
S.
Whiteley
and S. K. Ghandhi, "Factors influencing the growth
of GaO.47InO.53As
process,"
[26]
on InP substrates
J. Electrochem.
using
the metalorganic
Soc. 129, 383 (1982);
characterization
of GaO.47InO.53As
triethylgallium,
triethylindium,
"Growth
and
films on InP substrates
and arsine,"
130, 1191
using
(1983).
K. Carey, "OMVPE growth and characterization of high purity
GalnAs
on InP," Appl.
Phys.
Lett.
46, 89 (1985).
[27] K. T. Chan, L. D. Zhu, and J. M. Ballantyne, "Growth of high
quality GaInAs on InP buffer layers by MOCVD," Appl. Phys.
Lett. 47, 44 (1985).
[28]
A. Mircea, R. Azoulay, L. Dugrand, R. Mellet, K. Rao, and M.
on InP epitaxy
Sacilotti, "Metalorganic InP and In Ga l As PI
x
at atmospheric
pressure,"
J. Electron.
-x y
-y
Mater.
13, 603
(1984).
[29]
B. I. Miller, E. F. Schubert, U. Koren, A. Ourmazd, A. H. Dayem,
and R. J. Capik, "High quality narrow GaInAs/lnP quantum wells
grown by atmospheric organometallic
vapor phase epitaxy,"
Appl. Phys. Lett. ~,
1384 (1986).
[30]
M. Sacilotte, A. Mircea, and R. Azoulay, "InP growth by OMVPE,"
J. Cryst. Growth 63, 111 (1983).
[31] M. D. Scott, A. G. Norman, and R. R. Bradley, "The characterisation
of Ga
l-xIn xAs, Al l-xIn xAs and InP epitaxial layers prepared by
metal organic chemicalvapour deposition,"J. Cryst. Growth 68,
-
319 (1984).
[32] D. Wake, A. W. Nelson,
S. Cole, S. Wong, I. D. Henning, and E. G.
Scott, "InGaAs/InP junction field-effect transistors with high
transconductance
made using metal organic vapor phase epitaxy,"
IEEE Electron Device Lett. EDL-6, 626 (1985).
[33]
K. Fry, C. Kuo, R. Cohen, and G. Stringfellow, "Photoluminescence
of OMVPE GaInAs," Appl. Phys. Lett. 46, 955 (1985).
[34] C. Hsu, R. Cohen, and G. Stringfellow, "OMVPE growth of InP using
TMI," J. Cryst. Growth 63, 8 (1983).
[35] C. Hsu, R. Cohen, and G. Stringfellow, "OMVPE growth of InP using
TMI," Proc..Electrochem. Soc. 83-13,193
(1983).
[36] C. Hsu, R. Cohen, and G. Stringfellow, "OMVPE growth of InP using
TMI," J. Cryst. Growth 62, 648 (1983).
108
[37] C. Kuo, J. Yuan, R. Cohen, J. Dunn, and G. Stringfellow,
"OMVPE
growth of high-purity
GaInAs using TMI," Appl. Phys. Lett. 44,
550 (1984).
-[38] C. Kuo, R. Cohen, and G. Stringfellow,
J. Cryst. Growth 64, 461 (1983).
"OMVPEgrowth of GaInAs,"
[39] C. P. Kuo, R. M. Cohen, K. L. Fry, and G. B. Stringfellow,
"Characterization
of GaxIn l -xAs grown with TMIn," J. Electron.
Mater.
~,
231 (1985).
[40] S. Bass, C. Pickering,
and M. Young, "MOVPEof InP,"
Growth 64, 68 (1983).
[41]
S. Bass and M. Young,
alloys
grown using
reactor,"
J. Cryst.
J.
Cryst.
"High quality
epitaxial
InP and indium
TMI and phosphine
in an atmospheric
pressure
Growth 68, 311 (1984).
[42] S. J. Bass, S. J. Barnett, G. T. BrownN. G. Chew, A. G. Cullis,
A. D. Pitt,
and M. S. Skolnick,
"Effect of growth temperature
on the optical,
electrical
and crystallographic
properties
of
epitaxial
indium gallium arsenide
grown by MOCVDin an atmospheric pressure
reactor,"
J. Cryst. Growth~,
378 (1986).
[43] S. J. Bass, M. S. Skolnick,
H. Chudzynska, and L. Smith, "MOCVD
of indium phosphide and indium gallium arsenide
methylindium-trimethylamine
adducts,"
J. Cryst.
221 (1986).
using triGrowth 75
[44] R. F. C. Farrow, "The evaporation
of InP under Knudsen (equilibrium) and Langmuir (free) evaporation
conditions,"
J. Phys.
D: 2, 2436 (1974).
[45] T. S. Moss, Optical Properties
York, 1959), p. 13.
of Semiconductors,
(Academic,
New
[46] T. P. Pearsall,
L. Eaves, and J. C. Portal, "Photoluminescence
and impurity concentration
in Ga In l As PI
alloys
latticex
-x y -y
matched to InP," J. Appl. Phys. 54,1037 (1983).
[47] Handbook of Optics,
edited
York, 1978), p. 14-40.
[48] E. D. Towe and T. J.
by W. G. Driscoll
Zamerowski,
"Properties
(McGraw-Hill,
New
of Zn-doped p-type
InO.53GaO.47As grown by vapor phase epitaxy
(VPE) on InP
substrates,"
J. Electron.
Mater. 11, 957 (1982).
109
[49]
B. D. Joyce and E. W. Williams,
"The preparation and photoluminescent properties of high purity vapour grown indium phosphide
layers," Inst. Phys. Conf. Ser. No.9, 57 (1971).
[50] E. W. Williams, W. Elder, M. G. Ast1es, M. Webb, J. B. Mullin, B.
Straughan, and P. J. Tufton, "Indium phosphide 1. a photoluminescence materials study," J. Electrochem. Soc. 120,1741 (1973).
[51]
C. Pickering, P.R. Tapster, P.J. Dean, L. L. Taylor, P. L. Giles,
and P. Davies, "Optical characterisation of acceptors in doped
and undoped VPE InP," J. Cryst. Growth 64, 142 (1983).
[52]
L.
J.
van der Pauw, "A method of measuring specific resistivity and
Hall
effect
1 (1958).
of
discs
of
arbitrary
shape,"
Philips
Res.
Rep.
13,
[53] G. E. Stillman,
C. M. Wolfe, and J. o. Dimmock, "Hall coefficient
factor for polar mode scattering in n-type GaAs," J. Phys.
Chern. Solids 31, 1199 (1970).
[54] M. Benzaquen, D. Walsh, and K. Mazuruk, "Hall factor of doped
n-type GaAs and n-type InP," Phys. Rev. B: 34, 8947 (1986).
[55] T. Y. Wuand G. L. Pearson,
band structure
of In Ga
(1972).
"Phase diagram,
l As," J. Phys.
x
crystal
growth, end
Chern. Solids
-x
33,
409
-
[56] T. P. Pearsall,
"GaO 47InO 53As: a ternary semiconductorfor
photodetectorappl1cat10ns,"IEEE J. Quantum Electron. QE-16,
709 (1980).
[57] H. C. Casey and M. B. Panish, Heterostructure Lasers Part B,
(Academic, New York, 1978).
[58] R. E. Nahory, M. A. Pollack, W. D. Johnston, and R. L. Barns,
"Band gap versus composition and demonstration of Vegard's
law for In
Ga As P
matched to InP," Appl. Phys.
l -x x y I -y lattice
Lett. 33, 659 (1978).
[59] K. -H.
Goetz,
D. Bimberg,
G. F.
H. Jurgensen,
J. Selders,
A. V. Solomonov,
Glinskii,and M. Razeghi, "Opticaland crystallographic
propertiesand impurity incorporationof Ga In] As (0.44 < x
< 0.49) grown by liquid phase epitaxy, vapo~ pna~e epitaxy,
and metal
organic
chemical
vapor
deposition,"
J. Appl.
Phys.
54, 4543 (1983).
[60] P. Chandra, L. A.
from Ga In
x
Coldren,'
As P
l -x
Y I -y
and K. E. Strege, "Refractiveindex data
films," Electron. Lett. 17,
-
6 (1981).
110
[61]
H. Burkhard, H. W. Dinges, and E. Kuphal, "Optical properties of
In
Ga P
l
I As , InP, GaAs, and GaP determined by ellipsometry,"
-x x
J. Appl.
-y Y
Phys. 2l 655 (1982).
[62] E. Rabinowicz, An Introduction to Experimentation,
Reading, Mass., 1970), p.32.
(Addison-Wesley
[63] M. Born and E. Wolf, Principles
Oxford, 1970), p. 324.
(Pergamon,
of Optics,
4th ed.
[64] Handbook of Mathematical
Functions, edited by M. Abramowitz
Stegun (NBS, Washington,
1972), p. 81.
and I.
[65] M. M. Tashima, L. W. Cook, and G. E. Stillman, "The application of
x-ray diffraction measurements
in the growth of LPE InGaAs/InP,"
J. Cryst. Growth~,
132 (1981).
[66] R. E. Enstrom,
P. J. Zanzucchi, and J. R. Appert,
ties of vapor-grown
In xGa l-xAs epitaxial films
InxGal_xP substrates," J. App1. Phys. 45,
[67]
300 (1974).
Y. Takeda, A. Sasaki, Y. Imamura, and T. Takagi, "Electron mobility
and energy gap of In 53GaO 47As on InP substrate," J. Appl.
Phys.
[68]
"Optical properon GaAs and
.
!!2, 5405 (1976 9 :
R. J. Elliott, "Intensity of optical absorption by excitons,"
Phys. Rev. 108, 1384 (1957).
[69] E. Zielinski, H. Schweizer, K. Streubel,H. Eisele, and G. Weimann,
"Excitonic transitions and exciton damping processes in InGaAs/
InP,"
[70]
J. App1.
Phys.
22.,2196
(1986).
E. V. K. Rao, A. Sibille,
and N. Duhamel, "Investigationof processinduced defects in InP," Physica (Utrecht)ll6B, 449 (1983).
[71] H. C. Casey and E. Buehler, "Evidence for low surface recombination velocity on n-type InP," App1. Phys. Lett. 30,247 (1977).
[72] D. J. Ashen, P. J. Dean, D. T. J. Hurle, J. B. Mullin, A. M. White,
and P. D. Greene, "The incorporation and characterisationof
acceptors in epitaxial GaAs," J. Phys. Chem. Solids 36, 1041
(1975).
[73] O. Roder, U. Heim, and M. H. Pilkuhn,
"Electronmobilities and
photoluminescence of solution grown indium phosphide single
crystals,"J. Phys. Chem. Solids 31, 2625 (1970).
[74] P. W. Yu, "A model for the 1.10 eV emission band
State Commun. 34,
183
(1980).
in InP,"
Solid
111
[75]
J.
B. Mullin,
A. Royle,
B. W. Straughan, P. J. Tufton, and E. W.
Williams, "Crystal growth and properties of Group IV doped
indium phosphide," J. Cryst. Growth 13/14, 640 (1972).
[76] L. J. Giling, "Gas flow patterns in horizontal
cells observed by interference holography,"
Soc. 129, 634 (1982).
[77]
epitaxial reactor
J. Electrochem.
D. H. Reep and S. K. Ghandi, "Deposition of GaAs epitaxial layers
by organometallic CVD," J. Electrochem. Soc. 130, 675 (1983).
[78] G. E. Stillman, C. M. Wolfe, and J. o. Dimmock, "Hall coefficient
factor for polar mode scattering in n-type GaAs," J. Phys. Chem.
Solids 1!, 1199 (1970).
[79] L. D. Zhu, K. T. Chan, and J. M. Ballantyne, "Very high mobility
InP grown by low pressure metalorganic vapor phase epitaxy using
solid trimethylindium
source," Appl. Phys. Lett. ~,
47 (1985).
[80] A. H. Moore, M. D. Scott, J. I. Davies, D. C. Bradley, M. M.
Faktor, and H. Chudzynska, "High mobility InP epitaxial layers
prepared by atmospheric pressure MOVPE using trimethylindium
dissociated
from an adduct with 1,2-bis(diphenyl
phosphino)
ethane," J. Cryst. Growth lI, 19 (1986).
[81] G. E. Stillman, L. W. Cook, T. J. Roth, T. S. Low, and B. J.
Skromme, "High purity material," in GaInAsP Alloy Semiconductors,
edited by T. P. Pearsall (John Wiley & Sons, New York, 1982),
p. 147.
[82] W. Walukiewicz,
J. Lagowski, L. Jastrzebski, P. Rava, H. C. Gatos,
M. Lichtensteiger,
and C. H. Gatos, "Electron mobility and
free-carrier absorption in InP; determination
of the compensation
ratio," J. Appl. Phys. 51, 2659 (1980).
[83] A. C. Jones, P. R. Jacobs, R. Cafferty, M. D. Scott, A. H. Moore,
and P. J. Wright, "Analysis of high purity metalorganics
by
ICP emission spectrometry,"
J. Cryst. Growth lI, 47 (1986).
[84] D. H. Reep and S. K. Ghandi, "Electrical properties
metallic chemical vapor deposited GaAs epitaxial
Electrochem.
Soc. 131, 2697 (1984).
of organolayers," J.
[85] J. S. Blakemore, "Semiconducting and other major properties of
gallium arsenide," J. Appl. Phys. 53, R123 (1982).
[86] K. Hess, N. Stath,
J. Electrochem.
and K. W. Benz, "Liquid
Soc. 121, 1208 (1974).
phase
epitaxy
of InP,"
112
[87]
J. C. Paris, M. Gauneau, H. L'Haridon, and G. Pe1ous, "Effects of
annealing on unintentionally doped LEC InP," J. Cryst. Growth
64, 137 (1983).
[88] T. Kamijoh, H. Takano, and M. Sakuta, "Heat treatment of semiinsulating InP:Fe with phosphosi1icate
glass encapsulation,"
J. App1. Phys. 55, 3756 (1984).
[89]
R. E. Hollingsworth, and J. R. Sites, "Photoluminescence dead
layer in p-type InP," J. Appl. Phys. 53,5357 (1982).
[90] H. Nagai, and Y. Noguchi, "Surface-treatment effect on photoluminescence of InP," J. Appl. Phys. 50, 544 (1979).
[91]
G. S. Pomrenke, "Evidence of amphoteric behavior of Si in VPE InP,"
J. Cryst. Growth 64, 158 (1983).
[92]
N. Duhamel,
"Silicon
E. V. K. Rao, M. Gauneau, H. Thibierge, and A. Mircea,
implantation in semi-insulating
bulk InP; electrical
and photoluminescence measurements," J. Cryst. Growth 64, 186
---
(1983).
[93]
K. J. Bachmann, "Properties, preparation, and device applications
of indium phosphide," Ann. Rev. Mater. Sci. 11, 441 (1981).
[94]
Growth and Characterization of InP (First NATO workshop on InP),
edited by J. K. Kennedy, J. Cryst. Growth 54 (1)(1981).
[95]
Materials Aspects of InP (Second NATO workshop on InP), edited by
B. Cackayne,
J.
Cryst.
Growth
64 (1) (1983).
[96]
F. R. Bacher and W. B. Leigh, "Photoluminescence study of magnesium
doped MOVPE indium phosphide," J. Cryst. Growth 80 456 (1987).
[97]
A. W. Nelson and L. D. Westbrook, "A study of p-type dopants for
InP grown by adduct MOVPE," J. Cryst. Growth 68, 102 (1984).
[98] C. R. Lewis, W. T. Dietze and M. J. Ludowise, "The growth of
magnesium-doped
Mater. 12, 507
GaAs by the OM-VPE
(1983).
[99] R. R.
process,"
J. Electron.
Bradley, R. M. Ash, N. W. Forbes, R. J. M. Griffiths, D. P.
Jebb, and H. E. Shephard, "Meta1organic
chemical vapour
deposition of junction isolated GaAsAs/GaAs LED structures,"
J.
Cryst. Growth12,
629 (1986).
113
[100] K. J. Bachmann, E. Buehler, B. I. Miller, J. H. McFee, and F. A.
Thiel,
"The current
status
of the preparation
of single
crystals,
bicrystals,
and epitaxial
layers
of p-InP and of polycyrstalline p-InP films
for photovoltaic
applications,"
J. Cryst.
Growth 39, 137 (1977).
[101] D. Barthruff
and H. Haspeklo, "Excited states
in InP," J. Lumin. 24/25, 181 (1981).
of shallow
acceptors
[102] E. Kubota, Y. Ohmori, and K. Sugii, "Electrical
and optical
properties
of Mg-, Ca-, and Zn-doped InP crystals
grown by
the synthesis,
solute diffusion
technique,"
J. Appl. Phys.
55, 3779 (1984).
[103] G. S. Pomrenke, Y. S. Park, and R. L. Hengehold, "Photoluminescence
from Mg-implanted,
epitaxial,
and semi-insulating
InP," J. Appl.
Phys. 52, 969 (1981).
[104]
S. J. Bass and P. E. Oliver,
"Controlled
doping of gallium arsenide
produced by vapour epitaxy,
using trimethylgallium
and arsine,"
Inst.
Phys. Conf. Sere No. 33b, 1 (1977).
[105] V. Swaminathan,
study
V. M. Donnelly,
of Cd-related
centers
and J. Long, "A photoluminescence
in InP,"
J.
Appl.
Phys.
(1985).
58,
--
4565
[106] H. Burkhard, E. Kuphal, F. Kuchar, and R. Meisels, "Donoracceptor pair recombination from donor ground and first excited
states in InP," Inst. Phys. Conf. Sere No. 56, 659 (1981).
[107] E. Zacks and A. Halperin,
pair-photoluminescence
Rev. B~, 3072 (1972).
[108] T. Yagi,
escence
Phys.
"Dependence of the peak energy of the
band on excitation
intensity,"
Phys.
Y. Fujiwara,
T. Nishino,
and Y. Hamakama, "Photoluminmismatch in InGaAs/lnP,"Jpn. J. Appl.
L467 (1983).
and
11,
lattice
[109] K.-H. Goetz, D. Bimberg, K.-A. Brauchle,
H. Jurgensen,
J. Selders,
M. Razeghi, and E. Kuphal, "Deep Fe and intrinsic
defect levels
in GaO.47InO.53As/InP,"
Appl. Phys. Lett. 46, 277 (1985).
[110] C. P. Kuo, S. K. Vong, R. M. Cohen, and G. B. Stringfellow,
"Effect
of mismatch
J. Appl. Phys. ~,
strain
on band
gap in
[111] D. A. Humphreys, R. J. King, D. Jenkins,
"Measurement
III-V
semiconductors,"
5428 (1985).
of absorption
coefficients
~m,"
over the wavelength range 1.0-1.7
1187 (1985).
and A. J. Moseley,
of GaO 47InO
53As
Electron. Lett. 21,
--
114
[112] D. D. Sell and H. C. Casey, "Optical absorption and photoluminescence studies of thin GaAs layers in GaAs-Al xGa l-xAs double
heterostructures,"
J. Appl. Phys. ~,
800 (1974).
[113] ,~. J. Turner, W. E. Reese, and G. D. Pettit, "Exciton absorption
and emission in InP," Phys. Rev. 136, A1467 (1964).
[114] Y. Yamazoe, T. Nishino, and Y. Hamakawa, "Electroreflectance
study of InGaAsP quaternary alloys lattice matched to InP,"
IEEE J. Quantum Electron. QE-17, 139 (1981).
[115] T. P. Pearsall,
and A. Roizes,
G. Beuchet,
"Electron
J. P. Hirtz,
N. Visentin,
and hole mobilities
Inst. Phys. Conf. Ser. No. 56, 639 (1981).
M. Bonnet,
in GaO 47InO
53As,"
..
[116] R. Saxena, V. Sardi, J. Oberstar, L. Hodge, M. Keever, G. Trott,
K. L. Chen, and R. Moon, "OMVPE growth of InGaAsP materials for
long wavelength detectors and emitters," J. Cryst. Growth 77,
591 (1986).
--
[117] K. W. Carey, S. Y. Wang, R. Hull, J. E. Turner, D. Oertel, R.
Bauer, and D. Bimberg, "Characterization
of InP/GalnAs/lnP
heterostructures
grown by organometallic
vapor phase epitaxy
for high-speed p-i-n photodiodes,"
J. Cryst. Growth 77, 558
(1986).
--
115
APPENDIX
OHMIC
For n-type
as follows:
anneal
at 390°C
tance
4) anneal
necessary,
except
For evaporation,
from the sample,
or cryogenic
res is-
2 Ohms;
operation
was found
separated
as above was followed,
For p-type
samples,
that AuZn alloy was used
The AuZn material
at 900°C
was
to make
and a shutter
ampul
by 2.5 rom in a square pattern.
to exposure
procedure
of AuGe
(75% Au,
for 3 hours.
The source was 4 cm
the sample until
by a mask with
that
at OGC by annealing
were used.
blocked
except
the above
instead
was made
in a quartz
35 mg of metal
The sample was covered
to cool prior
The contact
a thin
poor contacts.
(70% In, 30% Ga by weight)
gold and zinc together
holes
connection
the same procedure
25% Zn by weight).
began.
to bridge
2)
5) epoxy
thus formed was typically
used
was
3) rub In dots onto
for 30 seconds,
for 30 minutes.
joint
contacts
(88% Au, 12% Ge by weight),
in an Ar atmosphere,
1) and 2) were omitted.
followed,
for making
to GaAs.
For InGaAs,
steps
rigid wire
InGa alloy
good contact
alloy
in air at 150°C
paste was occasionally
If neither
the method
in Ar at 200°C
of the wire-to-indium
silver
was
AuGe
for 2 minutes
6) bake
Cu wire,
PREPARATION
GaAs or InP samples,
1) evaporate
the contacts,
CONTACT
A
evaporation
four 0.75 rom diameter
The sample was allowed
to air.
All contacts were verified
to have a linear current/voltage
rela-
116
tionship continuously through positive and negative current values
in the region
mA),
using
of measurement
a Tektronix
for Hall mobility
576 curve tracer.
(usually
less than
1
117
APPENDIX B
EPILAYER THICKNESS MEASUREMENT
Several
layers.
All
The etchant
methods
were
methods
involved
selective
GaAs and
InP
used
K3Fe(CN)6
for
used
to measure
substrate
For InGaAs/lnP,
at
etching
epilayers
+ 8 g KOH in 100 ml H20.
insulating
thickness
of epitaxial
of the
substrate.
was ferricyanide:
Ferricyanide
etches
12 g
a semi-
rate from the n-type epilayer.
a different
40% HCl was used
to etch InP, leaving
the InGaAs
layer undisturbed.
The procedure
sample
mount
cleanly
for delineating
along
a plane
3) etch for
to the growth
the
surface,
2)
in H20,S)
An eyepiece
calibrated
with
scale
to equal
mount
accompanies
a variety
half-millimeter
termined
(GE #7031
varnish was used for HCl etch-
60 seconds under a microscope illuminator(15 s for
InGaAs), 4) rinse
with
perpendicular
1) cleave
the sample on a small piece of microscope slide using clay, with
the freshly cleaved edge up
es),
the layers was:
under
this microscope.
of millimeter
markings.
the Zeiss light microscope.
scale
Thus "one"
rules,
The scale was
including
on the Zeiss
scale was de-
0.061 mm at l28X and 0.0244 mm at 320X.
in this measurementwas estimated to be + 0.01
one
The error
"Zeiss units", or +
0.25 microns at 320X.
The epilayer
separating
could
the layer
be visually
delineatedby noting a dark line
from the substrate.
InGaAs
layers
could be seen
118
quite
easily,
pIes,
cracks
begining
GaAs and InP with more
or etch lines
at the interface
An independent
of multiple
For energies
below
removed
easily
Then
be observed
successive
III.
=
maxima
and leading
coherent
in their
transmission
m A/2n
light
using
minima
refractive
of several
the appropriate
value
angles.
utilized
the
light.
the InP substrate
and maxima
versus
could
wavelength
plots.
with m the order
index
maxima
sambe seen
of monochromatic
t for normal incidence,
and n t the InGaAs
occasionally
thickness
layers with
etalons:
To find m, the wavelengths
are compared,
InGaAs
as Fabry-Perot
=
t
could
reflections
InGaAs
On InGaAs/lnP
away at random
for measuring
the bandgap,
functioned
thickness
in the substrate
method
interference
difficulty.
found
in Chapter
(m, m + 1, etc.)
of nt for each.
of
119
APPENDIX
TABLE XIII.
Reference
77 K FWHM versus Net Carrier Concentration:
From the Literature.
Material
I
[49]
[49]
[50]
C
InP:Cr
InP:Sn
InP:Zn
ND-NAI
(cm-3)
PL FWHM
Raw Data
(meV)
1 x 1016
11
1 x 1018
32
4 x 1017
19
1. 3 x 1018
33
2.2 x 1018
31
6.5 x 1018
38
2 x 1019
40
2.7 x 1019
53
3.7 x 1019
62
1.2 x 1016
11
2.5 x 1016
16
3 x 1016
13
5 x 1016
18
7 x 1016
15
1.4 x 1017
19
5 x 1017
35
9 x 1017
35
1.5 x 1018
50
1.2 x 1016
16
1. 9 x 1016
17
6 x 1016
21
7 x 1016
19
1. 6 x 1017
17
2.6 x 1017
20
3.7 x 1017
25
8 x 1017
21,
1.4 x 1018
30
2.1 x 1018
30
2.4 x 1018
5 x 1018
33, 35
65
5 x 1018
65
26,
29
120
APPENDIX
D
RAW SIMS DATA FOR InGaAs/lnP
The data
1-29-7-86
in Figs.
are summarized
by Charles
each sample,
performed
in Table
43-6 are concentration
as provided
discussed
from SIMS analysis
Evans
The horizontal
thickness
determined
absorption
testing.
(in atoms/cc)
& Assoc.
most
There
accurate.
VI.
65 and OSU
Presented
versus
depth
here
profiles
are two profiles
for
(The SIMS conditions
were
III.)
scale on these
by interference
At least
The lower curve
figures was calibrated
effects
two isotopes
for S, Si, Cu, Zn, and Hg to check
interference.
on InGaAs
X of Chapter
one Cs beam and one 0 beam.
in Chapter
LAYERS
observed
per element
for possible
for each element
using
the
in optical
were monitored
molecular
ion
is presumed
to be the
121
Cu
s
s
.w.
L;
104 en
CD
e
!0
..0
t.)
z
0
t-I
.!!
Ag
Z
0
>cr
0
t-I
c
c
c
z
0
t.)
i
w
1cr en
0
t.)
1
MBE
InO.S3GaO.47As!InP
OSU 1-29-7-86
1
2
3
4
5
DEPTH(microns)
FIG.
43.
Beam SI~ Data for MBE InO.53GaO.47As.
scale app11es to As only.
Cs
Right
122
MBE
Ino.S3GaO. 47As!InP
OSU 1-29-7-86
U
I.J
......
en
Cu
en
&1
CI
5
0
u
Hg
.....
z
0
H;J
z
....
W
C[
CU
u
z
0
u
Al
0
....
>a::
C[
0
z
0
u
UJ
en
1rr
AA.
Fe
AsNi
Cr1if
M9
Na
1
2
3
DEPTH
FIG.
44.
4
5
(microns)
0 Beam SIMS Data for MBE InO.S3GaO.47As.
scale applies to As only.
Right
123
OMVPE
1
InO. 53GaO. 47As /InP
InGaAs 65
(InP buffer)
1
Cu
u...,
.......
0
...
104 tn
Zn
10
0
to)
:z
0
0
Ag
:z
.0
>Ir
<
c
:z
<
i
:z
0
to)
0
to)
10
Hr
U!
tn
Cl
Cl
10'
AS~
2
1
DEPTH
FIG.
45.
3
(microns)
Cs Beam SI~
Data for OMVPE InO.53GaO.47As.
scale
app1~es to As only.
Right
124
OMVPE
InGaAs65
Ino. 53GaO. 47As/InP
(InP buffer)
- ..v
to>
to>
.....
en
E
0
.....
104
en
.....
:z
0
Co)
.!!!
:z
0
.....
-<
Cu
:z
0
>cr
-<
CU
f=
ffi
Co)
:z
0
Co)
0
:z
0
Co)
w
en
W
Fe
1er
Al
1
2
3
DEPTHCmicrons)
FIG. 46.
o Beam SI¥ Data for OMVPE InO.53GaO.47As.
scale app11es to As only.
Right
125
APPENDIX
TABLE
XIV.
List
of chemicals
Item
E
and equipment.
Manufacturer
& model
#
Comments
PHOTOLUMINESCENCE:
275 mm monochromator
Jarrell
Ash, Mark X 82462
PbS detector
Hammamatsu
Ar+ ion laser
Coherent
cryostat
MMR TechnologiesR2l05
temperature
Radiation
indic. MMR Technologies
Labs
52
f/3.85, 6 nm/mm
dispersion,
600/mm
514.5 nm, 150 mT
1 mm sapphire
amplifier
x-y recorder
Houston
lock-in amplifier
EG&G Princeton Applied Res. 5105
vacuum
VacTorr
pump
N2 pump
Thomas
Si filter
262C
glass
Bausch
filter
AM 502
2000
current
source
Industries
107CA18
1. 5 mm,
& Lomb
780 nm
Keithly
619
Keithly
220
scanner
Keithly
705
6" electromagnet
Alpha
Scientific
6002
temp.
MMR Technologies
K-20
cryostat
MMR Technologies
R2105
IEEE bus control
Apple
vacuum
Edwards
controller
pump
gaussmeter
Bell
810 Hz
100
HALL MOBILITY:
voltmeter
window
K-77
Tektronix
Instrument
lIe
2 stage
Inc. 615
slit
grating
polished
126
TABLE
XIV.
OMVPE
CRYSTAL
vacuum
(cont.)
GROWTH:
leak checker
Alcatel
HZ purifier
Johnson
rf generator
Lepel
(glove box)
NZ filter
heat
tape
-8
He to 10
T
ASM 10
Matthy
HP-Z5
T-5-3-KC-A-B
Millipore
7.5 kW, 100 kHz
O.Z micron
N66
Thermolyne
heat tape controller
YSI 7Z
NZ (glove box)
Air Products pre-purified
HZ
Air Products UHP grade
Z ppm 0Z' 3.5 ppm HZO
NZ
Air Products
1 ppm 0Z' 1 ppm HZO
rated + 0.01 °c
temperature
mass
baths
flow control
rf/oven
controller
Neslab
Unit
UPC grade
RTE 9DD
Instruments
Watlow
series
gas regulators
Veriflow
gas filters
David
trimethylindium
Alfa Products
trimethylgallium
AlfaProducts
UFC-lOOO
TDR 450
J. Tripp Assoc.
TEM
87958
6.6 ppm Si
Strem Chemical
arsine
Phoenix
phosphine
Phoenix Research NA 1953
valves
Nupro SS-4BK-TW
HZO
rated + O.Z% of full scl.
859
CpZMg
deionized
Z-3 ppm 0Z' 10-5 ppm HZO
Research
10%
in HZ
10%
in HZ
Continental Water Sys. 3Zl8 O.Z micron, 17 megOhm
30% HZOZ
General Chemical 109-00350Z
96% HZS04
Baker 9673-1
37% HCl
Van Waters & Rogers UN 1789
microprocess grade
70% HN03
99.5% TCE
Van Waters & Rogers UN Z03l
microprocess grade
American Burdick & Jackson MS80897
99.5%
Van Waters
& Rogers
Van Waters
& Rogers
[CH3]ZCO
99.9% CH30H
clean bench
Envirco
low particle grade
trace metal analysis grade
UNl090
VLSI grade
microprocess
grade
microprocess
grade
class 10 tested 3/86
127
TABLE
XIV.
OPTICAL
(cont.)
ABSORPTION:
275 mrn monochromator
Jarrell
Ash, Mark
PbS detector
Hammamatsu
lock-in
EG&G Princeton
amplifier
amplifier
Tektronix
Si filter
262C
glass
Bausch
filter
X 82487
Applied
Res. 5101
810 Hz
AM 502
1. 5 mrn, polished
& Lomb
780 nm
cryostat
Air Products
x-y recorder
Hewlett Packard 7045B
Displex
temp.
.
cal.
zeroed
~
with
covered
tungsten
epoxy
OHMIC
8V
Oriel
lamp
Hardman
transparent
24 hour
CONTACTS:
Mo evaporationboat Balzers
diffusionpump
CVC CVE-15
Ag paste
DuPont 7941
conductive
curve
epoxy
tracer
THICKNESS
Epoxy
Technology
Tektronix
576
MEASUREMENT:
microscope
(see optical
Zeiss
absorption
64702
apparatus)
H3lD
one-part
cure
1
°c
detector
128
APPENDIX
COMPUTER
Two programs
Table
XV gives
measurement.
The program
the full program
the characterization
for resistivity
was controlled
and the electromagnet,
with
and magnetic
field,
any magnetic
actual
sample/contact
and Hall mobility
were manually
the field on.
ID, set temperature,
The output
temperature,
uniformity
chart
is easily
if desired.
then Hall
is a printout
assumed
thickness
(F), resistivity,
Figure
Hall
The Hall
47 gives
the flow-
for this program.
A second
cients
program
for InGaAs.
wavelength
shows
incorporated
except
operated.
field,
mobility, carrier concentration, and conductivity type.
factor
effort.
by the computer,
which
is to be run once without
are remeasured
of sample
PROGR&~S
used to speed
This apparatus
for N2 cooling
voltages
were
F
and temperature
interference
effects
coefficient
estimates
transmission.
to this value
to calculate
The refractive
the flowchart,
absorption
was used
The operator
and re-enter
coefficient
into the program.
and Table XVI gives
which
the program
the operator
the computer
uses
can then compare
estimates
is obtained.
absorption
coeffi-
index of InGaAs as a function
are written
are included,
optical
until
Figure
lines.
inputs
of
48
When
absorption
to calculate
the measured
the best value
a predicted
transmission
of the
129
IJSE
PRE-SET
PARAMETERS
PARAMETERS
DEFINE
}
VARIABLES
INCREMENTS
TEMPERATURE
RANGE,
CURRENT TO SAMPLE
INPUT
SAMPLE
}
MAGNETIC
THICKNESS
FIELD
NAME INSTRUMENT ADDRESSES
DIMENSION STRING ARRAYS
ACTIVATE IEEE
REMOTE
BUS
ENABLE INSTRUMENTS
-
NEXT
VOLTAGE
SET RELAYS
(SIX TIMES,
NEXT X LOOP)
MEASUREMENT
SUBROUTINE
TEMPERATURE
POINT
LOOP
CALCULATE
RESISTIVITY
STORE fJ( T)
HALL VOLTAGE
STORE
DATA
ON DISC
FIG.
47.
,
CALCULATE
MOBILITY,
CARRIER CONC.
OUTPUT
DATA
TO PRINTER
Flowchart for Hall Mobility Measurement Program.
130
TABLE
Line
10
XV.
Hall Mobility
BASIC
/I
Measurement
and Calculation
Program.
(DOS 3.3) Program Statement
CALL 1002
= CHR$ (4)
PRINT D$;"PR#3"
PRINT D$;"IN#0"
PRINT"
van der Pauw Resistivity/Hall Mobility measurement.....
PRINT "NOW' With SAMPLEFIX'
(Rev. 3/26/87)"
PRINT
PRINT "Enter sample identification (up to 20 characters, no commas).
20 D$
30
40
50
60
70
80
90 INPUT 1$
100 IF U
"TEMPCON1" THEN GoTo 210
110 PRIN' "Enter thickness
(in microns)."
120 INPUT T
122 PRINT "Enter settle time (1000 = 1 sec.)"
124 INPUT TT
130 PPIIH "Use Preset (PR) or Mc.nual(MA) current
es?"
PF:IN'
PRItH
ts
are
"Presets:
tco.ken
I
and temperature valu
.0005A, starting temp. is 78K, and 7 data poin
30K apart up to 258K, assuming a 5000 Gauss
R.:o.!;Inetic field."
14~, PRINT
1S0
PRIN' "'ype PR or MA."
160
INPUT Tt.
1-/0
IF
1$
= "PR" THEN
180
190
200
20~
GoTo 189(2)
IF T$ = "MA" THEN
GoTo
PRINl "TRY AGAIN"
GO TO 130
PRltH D$; "DE.LETE PATCH"
210 Z$ = CHR$ (26)
A$ =
CHF.:$:(38)
230
240
250
260
270
28C
290
::'.00
B$ =
CHR$
(44)
C$
1"$
CHR$
(49)
CHR$
(97)
=
=
(;1$ =
J$. =
E$ =
1780
CHR$ (42)
CHR$ (74)
CHR$ (98)
F$ = CHR$
(70)
IF I~
"TEMF-CONT" THEN
GOTo 2620
311ZJ
PRINT "Is the magnet on?
(Type Y or N; if yes, only Hall data will
be taken.)"
320 IF T$ = "PR" THEN
PRINT "MAGNETIC FIELD ASSUMED TO BE 5000 GAUSS."
330
34~
350
:~\60
370
375
380
=
INPUT S$
IF S$. =
"N" THEN GoTo 400
IF T$ = "PR" THEN GOTO 380
PRINT "Enter magnetic field (in Gauss)."
INPUT C
IF LEF1$ (1$,3) = "INP" THEN GoTo 394
PRINT "Use Hall Factor correction? Type Y or N (for N, Hall Factor
= 1.0)"
39121
392
394
40121
41121
42121
430
440
450
INPUT ZZ$
GOlD 400
ZZ$ = "N"
PRINT "Use temperature controller?
INPUT U$
IF U$.= "N" THEN GOTo 440
GOTO 2620
FOR Y = 1 TO J
CALL
11211212
(Type Y or N)"
131
TABLE XV. (cont.)
.-60
.-70
480
490
500
510
520
530
540
550
560
570
580
590
600
610
620
630
640
650
660
670
PRINT
PRINT
PRINT
IF U$
ONERR
PRINT
PRINT
INPUT
PRINT
HOME
PRINT
PRINT
D$; "PR#l"
D$;" IN#l"
"LF1"
= "N" THEN GOTO 710
GOTO 520
"WT";Q$;Z$;"SK ";K
"RD";J$;Z$;
TA$
D$;"PR#3"
TA$
"Set temperature is ";1<;",and this is measurement number ";Y
;" of u;.J;II.n
PRINT D$;"PRtll"
PRINT D$;"IN#1"
PRINT "LF1"
FOR B = 1 TO 30000: NEXT
ONERR
GOTO 640
PRINT "WT";Q$;Z$;"TE"
PRINT "RD";J$;Z$;
INPUT TE$
F'RINT D$; "PRtl3"
PRINT "Actual temperature
;Y;II
of
B
is ";TE$;"
(set point
";K;",
tI;J;..).n
680 PRINT D$;"PR#1"
690 PRINT D$;" INtH"
700 PRINT "LF1"
710 PRINT "RM";B$;Z$
720 PRINT "RM";C$;Z$
730 PRINT "RM";A$;P$;Z$
740 PRINT "LL"
750 PRINT "WT";A$;P$;Z$;"F0R1T4P1C1Z0X"
760 PRINT "WT";A$;P$;Z$;"Z1X"
770 FOR B = 1 TO 1000: NEXT B
780 IF S$ = "V" THEN GOTO 1200
790 FOR X = 1 TO 2
800 PRINT "CL";C$;Z$
810 IF X = 1 THEN PRINT "WT";C$;Z$;"B1C2C5CI0CI3CI8X"
820 IF X = 2 THEN PRINT "WT";C$;Z$;"B3C3C6CI0CI4C15X"
830 PRINT "WT";B$;Z$;"F1B1L1D0V100X";"I";I;"X"
840 R = 0
850 GOTO 1000
860 PRINT "WT";A$;P$;Z$;"F0R1T4C0Z0X"
870 FOR B = 1 TO TT: NEXT B
880 PRINT "RD";F$;P$;Z$;
890 INPUT" ";M$
900 FOR B
1 TO 400: NEXT B
910 PRINT "UT"
920 FOR B = 1 TO 400: NEXT B
930 PRINT "WT";B$;Z$;"F0X"
940 M = VAL ( HID$ (H$,5,16»
950 H.= ABS (H)
960 R = H I I + R
970 PRINT "WT";A$;P$;Z$;"F0R1C1X"
980 FOR B = 1 TO 500: NEXT B
990 RETURN
1000 GOSUB 2360
1010 PRINT "WT";B$;Z$;"F1X";"I-";I;"X"
1020 GOSUB 2360
1030 PRINT "CL";C$;Z$
1040 IF X ...1 THEN PRINT "WT";C$; Z$:;"B2C4C7CI0CllCI6X"
1050 IF X ...
2 THEN PRINT "WT";C$;Z$;"B4C1C8C10C12CI7X"
=
data
point"
132
TABLE XV. (cont.)
1060
PRINT "WT",B$,Z$;"FIX"
1070 GOSUB 23611I
1080
PRINT .WT".B$,Z$,"FIX"."I". I."X"
SOSUB 2360
IF X .. 1 THEN S .. R I 4
IF X - 2THENR-R/4
lC119C11
1100
1110
1120
NEXT X
S) I (R + S).
.34657 * «R
F - 1
S»
4
1140 R - .000226622 * F * (R + ~) * T
1130
,..
2 - .092358 * «R
- S) I (R +
,..
1200
1210
1220
1230
1240
1250
1260
1270
1280
1290
1300
1310
PRINT
PRINT
PRINT
SOSUB
"eL",C$,Z$
"WT",C$;Z$;"B5C3C5CIC11CI2C1BX"
"WT",B$,Z$,"F1B1LID0VI00X","I".I,"X"
2440
ZA$ .. W$
PRINT "WT",B$;Z$;"F1X"."I-";I,"X"
SOSUB 2440
ZB$ .. W$
PRINT "CL";C$;Z$
PRINT "WT";C$;Z$;"B6C4C6CI0Cl1CI7X"
PRINT "WT";B$;Z$;"F1X"
SOSUS 2440
1320 ZC$
..
W$
1330 PRINT "WT",B$;Z$;"FIX","I",I."X"
1340 SOSUB 2440
1350 ZD$ - W$
1370 ONERR SOTO 1390
1380 IF V
..
J THEN PRINT "WT",Q$;Z_;"SK000.00"
1390 PRINT "RD",J$;Z$;
1400 INPUT TA$
1415 VV - 100 + V
1420 PRINT
. "WT".C$'Z$'''R0X''
",
1430 IF V
J THEN PRINT "LA"
..
1440 PRINT
1450
1460
1470
D:t.; "PR4t3"
PRINT !J$;"INtt0"
(;ALL 1002
PRINT D$; "OPEN fY.T;;H,LlfZ;(..;.'
149111
15010
1510
15:tJ~
15:.~
I:'Ai!l
I:5S!.::
PF-lt,:IJJ-; "Wio:l1E PATCH,R";'t
f"f:an
K
PFOUl f.'
F'R 1NT ZA:t
PRJln
ZBt:
PRJNT ZC$
PRINT ZD$
1~.6/L1
PRINT D$; "CLOSE PATCH"
1480 IF S$
Co
".,."
THEU GOTD 17113
1570
IF U$ = "N" THEN
GOTO 16135
1:5E!~.' PRINT D:t; "PR*4"
1:581
IF Y > 1 THEN GOTO 1600
15B:! IF Sf: :: "N" THEN
PRINT "Wit.h IIIAgnetoff,
1~B4
If. S$ :: "Y" THEN
PRINT
"M.gnet on:"
actuAl
temps
_reI"
Ib0~ PRINT TE$
16'215
K .. K + A
1610
N~q V
Ib20 PRINT D$;"PRtt3"
165'" PRINT "MeAsurement complete."
1640 PRINT "To rerun type RE, to calculate type CA, to end type EN."
1650 INPUT 0$
133
TABLE XV. (cent.)
16b0
1670
1680
1690
1700
1710
1720
1730
1740
1750
1760
1770
1780
1790
1800
1810
1830
1840
1850
1860
IF
0$
= "RE" THEN
IF 0$ = "CA" THEN
IF
0$
"" "EN"
THEN
GOT a 2370
GOTO 1950
END
PRINT "TRY AGAIN"
GOTO 1630
PRINT D$;"WRITE PATCH,R";YY
PRINT ZA$
PRINT ZS$
PRINT ZC$
PRINT ZD$
PRINT D$;"CLOSE PATCH"
GOTO 1570
PRINT -Enter current (in Aeps, e.l A ..~)."
INPUT I
PRINT "Enter starting temperature <in Kelvins)."
INPUT K
PRINT "Enter number Df data pDints desoired."
INPUT J
IF J = 1 THEN GOTa 205
PRINT "Enter temperature increments <in Kelvins, XX.XX < 50.00K)."
1870
INPUT A
1880
GOTO 205
1890 I .. .0005
78.01
1900 K
1910 A .. 30.00
1920 C
5000
1930 J
7
1940
GOT a 205
=
=
=
1950
FOR P "" 1 TO J
.. 100 + P
ONERR GOTO 2390
CALL 1002
PRINT D$;"OPEN PATCH,L100"
PRINT D$;"READ PATCH,R";P
INPUT AA,AS,AC,AD,AE,AF
PRINT D$;"READ PATCH,R";PP
INPUT AG,AH,AI,AJ
PRINT D$;"CLOSE PATCH"
IF AG > AC AND AI > AE AND AH < AD AND AJ < AF THEN ZN$ = "p-type
1955 PP
1960
1970
1980
1990
2000
2002
2004
2010
2016
sample"
2018
IF AH
> AD AND
AJ > AF AND AG < AC AND
sample"
2019 IF ZN$ < > "n-type sample" AND ZN$ <
"Indeterminate"
2030 V = ( ABS (AC - AG) + ABS (AD - AH) +
AJ » I 4
2040 V = (T * 10000 * V) I (C * AB)
2050 ZA = (1.6022E - 19 * V * AS) A-I
2055 AZ = 1.0
2060 IF ZZ$ = "N" THEN GOTO 2190
THEN
AZ
= AZ
ASS (AE - AI) +
IF AA
2070
2075
IF AA <
250 AND AA > = 150 THEN AZ = AZ * 1.155
IF AA < 150 THEN AZ
AZ * (1.265
(18.61 I AA»
""
""
ZA
=
ZA * AZ
2190 PRINT D$;"PR*3"
2200 PRINT D$;"PR#4"
*
(.9
+
(60.4
-
ZN$ = "n-type
> "p-type sample" THEN ZN$
2065
2077
> 250
AI < AE THEN
I AA»
ASS (AF -
TABLE xv.
134
(cont.)
=
2210
IF P
1 THEN PRINT "SAMPLE ";1$;"
";"Fieldwas ";C;" Gauss, t
hickness ";T;" microns."
2212 IF P .. 1 THEN PRINT""
2213 IF P = 1 THEN PRINT "I = ";1
2214 IF P = 1 THEN PRINT "For the last data point:"
2215 IF P = 1 THEN PRINT"F = ";F;"
";"Hall Factor = ";AZ
2220 IF P = 1 THEN PRINT
2230 IF P = 1 THEN PRINT "TEMP";"
";"RESISTIVITY";"
";"MOBILITY"
;"
";"CARRIER CONCENTRATION"
2240 IF P = 1 THEN PRINT" K
Ohmcm
cm 2/Vs
cm"-3
2250
2260
2270
2280
2290
2300
2310
2312
2313
2314
2315
2316
2320
2330
2340
2350
2360
237~
2380
2390
2400
2410
2420
2430
2440
2450
2460
2470
2480
2490
2500
2510
2520
2530
2540
2550
2590
2600
2610
2620
2630
2635
2640
2650
2660
2670
2680
2690
2700
IF P = 1 THEN PRINT
PRINT AA;"
";AB;"
IF P = 5 THEN PRINT
IF P = 10 THEN PRINT
IF P = 15 THEN PRINT
IF P = 20 THEN PRINT
NEXT P
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT
";V;"
"
'.; ZA;
II
";ZN$
D$;"PR#3"
" "
"end."
END
GOTO 860
K = K - (A * J)
GOTO 310
PRINT D$;"PR#3"
PRINT "END OF DATA...CHECK DISC"
CALL 1002
PRINT D$;"CLOSE PATCH"
GOTO 2350
PRINT "WT";A$;P$;Z$;"F0R1T4C0Z0X"
FOR B = 1 TO TT: NEXT B
PRINT "RD";F$;P$;Z$;
INPUT
" .";M$
FOR B = 1 TO 400: NEXT B
PRINT "UT"
FOR B = 1 TO 400: NEXT B
PRINT "WT";B$;Z$;"F0X"
M = VAL ( MID$ CM$,5,16»
W = M I I
W$ = STR$ (W)
IF LEN (W$) = 1 THEN W$ ..W$ + ".000"
PRINT "WT";A$;P$;Z$;"F0R1C1X"
FOR B = 1 TO 500: NEXT B
RETURN
HOME
CALL 1002
PRINT
PRINT
Temperature controller subroutine"
n
PRINT
PRINT "Enter Command:"
PRINT
TE = Current temperature?"
PRINT
WK ..Current set temperature?"
SK NNN.NN = Sets temp. to NNN.NN"
PRINT
VA .. Current
vacuum?"
PRINT
TABLExv.
271121
272121
273121
274121
2750
2760
2770
278121
279121
28CJI2I
281121
2820
283121
284121
285121
286121
135
(cant.)
PRINT "
EX
PRINT "
EN
INPUT CO$
IF CDS = "EN" THEN
IF CO$ = "EX" THEN
oNERR GoTo 281121
PRINT D$:"PR#1"
PPINT D$:"IN#1"
Pf;:INT
"LF1"
PRINT "WT":QS:Z$:Co$
PRINT "RD":J$: Z$:
INPUT Cl$
PRINT D$:"PR#3"
PRINT D$;"IN#I2I"
PRINT CU
GoTO 2635
Exits from subroutine"
Ends program"
END
GoTO 440
136
-Sample ID
-Thickness
INPUT
SAMPLE
-Temperature
-Substrate thickness
INFORMATION
SELECT
ENERGY
(0.7-1.
5
& doping
(if any)
eV)
28 values
NEXT N
LOOP
.
E >
...-m
0.9 eV?
NO
NO
INCLUDE INTERFERENCE?
DATA DESIRED
FOR THIS
ENERGY?
NO
=
(Select at E
0.7 eV, no subs)
I
YES
NOI
NEXT
N
INPUT %T,
DATA DESIRED
FOR THIS
ENERGY?
CALCULATE
ex
I
(With substrate
LOOP
lNEXT
LOOPN
effects)
I
YES
.
I
PRINT
PRINT
INPUT
%T
IE,
I
%T,
ex
SAMPLE
,
Ci t
INFORMATION
OK
INPUT ex
ESTIMATE,
CALCULATE
I
J
COMP
ON ARE
SCREEN
%T
%T
NOT OK
FIG. 48. Flowchart for Absorption Coefficient Calculation Program.
137
TABLE XVI.
Line
Optical
BASIC
/I
1121
2121
PRtt,5
INtt 0
30
PRINT
Absorption
(DOS 3.3) Program
40
"Enter- sample
am rev. 2/5/87)."
INPUT U
50
PRINT
). II
60
70
80
8~.
90
10121
110
12121
130
140
144
] 4.:>
160
17':)
180
190
200
210
2:20
~~30
::'4b
2S~
.,r.-....
':'.J
Calculation
Program.
Statement
ID for- absor-ption
coefficient
calculation
(progr-
"Enler measurement temperatur-e in Kelvins (10, 77, or 300 only
INPUT K
PRINT "Was ther-e an InP substr-ate for- this sample?
(Y or- N)"
INPUT S:t
IF S:t = "N" THEN GOTa 130
IF St. = "Y" THEN PRINT "Enter- InP substr-ate thickness in micr-ons."
I NF'UT
TT
PRINT
INHiT
"EnterT$
substr-ate
type
(N or- P)."
PRINl "Was the front sUI-face InP (I) or- InGc.As (G)? (Input M for d
iffer-ential sample
calculation; implies R = 121.)"
INPUT
R:t
PRINT
%T scaling factor--enter1 if no sCeling is desir-ed."
INPUT "EnterS
PRINT "Enter-sample thickness in micr-ons."
INPUT T
P~:tt 4
.....
PRIIH "SA~1F'LE ";1;1.; "
Temper oOt.tL'.r e:
..; K; .. .".
PRINT
PEINl
"Er,~rgy(eV)
PRtt 3
FOF- I~ ... 1 10 28
H
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
302
IF
304
IF
31121 IF
312
IF
314
IF
320
IF
322
IF
324
IF
330
IF
332 IF
334 IF
340 IF
342 IF
344 IF
350 IF
254
:::6(1
26.2
:'64
270
27:'
274
280
282
284
29121
292
294
30e1
Coefficient
I.E.~H'
(U..3)
N '"' 1 AIm !<
a
(l/cm)
at"
= "INP" THEN GOTa 3000
10 THEN E = .7:RA
= .574:RB
= .0632:NA
= 3.47
N = 1 ANI.' I..
77 THEN E = .7:RA
=
.572:RB
=
.064:NA
=
3.485
N = 1 AND K
300 THEN E = .7:RA
=
.567:RB
=
.12I657:NA = 3.515
N = 2 AN[' ~:
10 THEN E = .71:RA
= .573:RB = .0635:NA = 3.475
N = 2 AND K
77 THEN E = .71:RA
= .571:RB
= .0643:NA
= 3.49
N = 2 AND K
300 THEN E = .71:RA
=
.566:RB
=
.0659:NA
= 3.52
N = 3 AND I<
10 THEN E = .72:RA
= .572:RB
= .064:NA
= 3.485
N = ::; AND K
77 THEN E = .72:RA
= .569:RB = .0648:NA = 3.5
N = 3 AND K
300
THEN E = .72:RA
= .563:RB
= .0668:NA
= 3.535
N = 4 AND I(
= .571:RB
= .0643:NA
= 3.49
10 THEN E = .73:RA
N = 4 AND K
= .567:RB
= .0654:NA
= 3.51
77 THEN E = .73:RA
N = 4 Ar~D I<
300
THEN E = .73:RA
=
.561:RB
=
.0676:NA
= 3.55
N = 5 AND K
10 THEN E = .74:RA
= .569:RB = .0648:NA = 3.5
N = 5 AND K
77 THENE = .74:RA = .566:RB = .0659:NA = 3.52
N = 5 AND I(
30121THEN E = .74:RA = .559:RB = .0682:NA = 3.56
N = 6 AND I( = 10 THEN E = .75:RA
= .568:RB
= .0651:NA
= 3.505
N = 6 AND K = 77 THEN E = .75:RA
= .564:RB = .0665:NA = 3.53
N = 6 AND I(
300 THENE = .75:RA = .558:RB = .0684:NA = 3.565
N = 7 AND I(
10 THENE = .76:RA = .567:RB = .0657:NA = 3.515
N = 7 AND K
77 THENE = .76:RA = .563:RB = .067:NA = 3.54
N = 7 AND K
300 THENE = .76:RA = .558:RB = .0684:NA = 3.565
N = 8 AND K
10 THEN E = .77:RA = .565:RB = .0662:NA
3.525
N = 8 AND K
77 THEN E = .77:RA
.561:RB
= .0676:NA = 3.55
N = 8 AND K
300 THEN E = .77:RA
= .559:RB = .0682:NA = 3.56
N = 9 AND I<
10 THEN E = .78:RA = .563:RB = .067:NA = 3.54
N = 9 AND K
77 THEN E = .78:RA = .559:RB = .0682:NA
= 3.56
N = 9 AND I<
300 THEN E = .78:RA
= .561:RB = .0676:NA = 3.55
N = 10 AND K
10 THEN E = .79:RA = .561:RB = .0676:NA
= 3.55
N = 10 AND K
77 THEN E = .79:RA
=
.558:RB
=
.0684:NA
= 3.565
N = 10 AND K = 300 THEN E = .79:RA
= .562:RB = .0673:NA
3.545
N = 11 AND K
10 THEN E = .8:RA
= .558:RB = .0684:NA = 3.565
=
=
=
138
TABLE XVI.
352
354
360
362
364
370
372
374
380
382
384
390
392
394
400
402
404
410
412
414
420
4
424
430
432
434
440
442
444
450
452
454
460
462
464
470
472
474
480
482
484
490
492
494
500
502
504
510
512
514
520
522
524
530
'"")'"")
......
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
IF
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
(cont.)
= 11
= 11
= 12
= 12
= 12
= 13
=
=
=
=
=
=
=
=
=
=
AND K
AND K
AND
AND
AND
AND
AND
AND
AND
AND
AND
AND
AND
K
K
K
K
K
K
K
K
K
K
K
= 77 THEN E = .8:RA = .558:RB = .0684:NA = 3.565
= 300 THEN E = .8:RA = .562:RB = .0673:NA = 3.545
= 10 THEN E = .81:RA = .558:RB = .0684:NA = 3.565
= 77 THEN E = .81:RA = .5S8:RB = .0684:NA = 3.565
= 300 THEN E = .81:RA = .562:RB = .0673:NA = 3.545
= 10 THEN E = .82:RA - .558:RB = .0684:NA = 3.565
77 THEN E = .82:RA = .559:RB = .0682:NA = 3.56
300 THEN E = .82:RA = .562:RB = .0673:NA = 3.545
10 THEN E = .83:RA = .558:RB = .0684:NA = 3.565
77 THEN E = .83:RA = .S61:RB = .0676:NA = 3.55
300 THEN E = .83:RA = .S61:RB = .0676:NA = 3.55
10 THEN E = .84:RA = .561:RB = .0676:NA = 3.S5
77 THEN E = .84:RA = .562:RB = .0673:NA = 3.545
300 THEN E = .84:RA = .S61:RB = .0676:NA = 3.55
10 THEN E = .85:RA = .562:RB = .0673:NA = 3.545
77 THEN E = .85:RA = .S62:RB = .0673:NA = 3.S45
IF N = 16 AND K = 300 THEN E = .85:RA = .56:RB = .0679:NA
= 3.555
IF N
17 AND K
10 THEN E
.86:RA
.562:RB
.0673:NA
3.545
IF N
17 AND K
77 THEN E
.86:RA
.562:RB
.0673:NA
3.545
IF N = 17 AND K
300 THEN E
.86:RA
.559:RB
.0682:NA
3.56
18 AND K = 10 THEN E
.87:RA
.562:RB
.0673:NA
3.545
IF N
IF N
18 AND K
77 THEN E
.87:RA
.561:RB
.0676:NA
3.55
IF N
18 AND K
300 THEN E
.87:RA
.S58:RB
.0684:NA
3.565
IF N
19 AND K
10 THEN E
.88:RA
.S61:RB
.0676:NA
3.55
IF N
19 AND K
77 THEN E
.88:RA
.S6:RB
.0679:NA
3.555
IF N = 19 AND K
300 THEN E
.88:RA
.558:RB
.0687:NA
3.57
20 AND K
10 THEN E
.89:RA
.561:RB
.0676:NA
3.55
IF N
IF N
20 AND K
77 THEN E
.89:RA
.SS9:RB
.0682:NA
3.56
IF N
20 AND K
300 THEN E
.89:RA
.558:RB
.0687:NA
3.57
IF N = 21 AND K = 10 THEN E
.9:RA = .561:RB
.0676:NA
3.55
IF N
21 AND K
77 THEN E
.9:RA
.559:RB
.0682:NA
3.56
IF N
21 AND K
300 THEN E
.9:RA
.557:RB
.069:NA
3.S75
IF N
22 AND K
10 THEN E
.95:RA
.558:RB
.0687:NA
3.57
77 THEN E
.95:RA
.556:RB
.0693:NA
3.S8
IF N = 22 AND K
IF N
22 AND K
300 THEN E
.95:RA
.554:RB
.0701:NA
3.595
IF N
23 AND K
10 THEN E
l:RA
.554:RB
.0701:NA
3.595
IF N
23 AND K
77 THEN E = l:RA
.553:RB
.0704:NA
3.6
IF N
23 AND K
300 THEN E
l:RA
.551:RB
.0712:NA
3.615
IF N = 24 AND K = 10 THEN E = 1.1:RA
.S49:RB = .0718:NA = 3.625
24 AND K
77 THEN E
1.1:RA
.548:RB
.0721:NA
3.63
IF N
IF N
24
AND
1<
300 THEN E
1.1:RA
.547:RB
.0724:NA
3.635
IF N = 25 AND K
10 THEN E
1.2:RA
.546:RB
.0729:NA
3.645
IF N
25 AND K = 77 THEN E
1.2:RA
.546:RB
.0729:NA
3.645
IF N
25 AND K
300 THEN E
1.2:RA
.545:RB
.0732:NA
3.65
IF N
26 AND K
10 THEN E
1.3:RA
.545:RB
.0732:NA
3.65
IF N = 26 AND K = 77 THEN E = 1.3:RA
= .S45:RB = .0732:NA = 3.65
IF N
26 AND K
300 THEN E
1.3:RA
.543:RB
.0738:NA
3.66
IF N
27 AND K
10 THEN E
1.4:RA = .543:RB
.074:NA
3.665
IF N
27 AND K
77 THEN E
1.4:RA
.542:RB
.0743:NA
3.67
IF N
27 AND K
300 THEN E
1.4:RA
.S4:RB
.0749:NA
3.68
28 AND K
10 THEN E = 1.5:RA
.538:RB
.0758:NA
3.695
IF N
IF N = 28 AND K = 77 THEN E = 1.5:RA
= .S37:RB = .076:NA = 3.7
IF N = 28 AND K = 300 THEN E
1.S:RA
.535:RB
.0769:NA = 3.715
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
13
13
14
14
14
15
15
15
16
16
AND K
AND K
AND K
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
PRINT "For ";E;" eV energy, enter 7.T (if no data for this energy, e
nter 0; for special
value, enter 1, noting that InGaAs r
eflectance will be for T < 77 K)."
540 INPUT PT
545 IF N = 28 AND PT = 0 THEN GOTO 770
550 IF PT = 0 THEN NEXT N
TABLE XVI.
139
(cont.)
=
557
IF PT
1 THEN
GOTO 1300
IF Sot = "N" THEN GOTO 62121
560
IF E < =.8
THEN AA = (
1121 * E)
+ 13.25
570
IF E :> .8 THEN AA
(10
* E)
3
580
2 * AA
IF K > 10121 THEN AA
590
IF T$ = "P" THEN AA = AA / 10
600
611Z1 GOTO 64121
AA
0
620
TT
121
630
"M" THEN GOTO 664
635
IF
R$
"G" THEN GOTO 693
640
IF
R$
3.17
.9 THEN NN
650
IF E <
(. 591 * E) + 2.625
IF E :> .9 THEN NN
660
GOTO
67121
662
NN
1
664
2 / (NN + 1) A 2
R
(NN
1>
670
- 2
RI
(1
R)
680
PT
685
PT * S
L
RI * PT
690
GOTO 700
692
PT
PT * S
693
IF E < =.9 THEN GOTO 101121
694
2 / (4 * PT
2» +
695
L
(RA /
(2 * PT»
+ «RA
LOG (L)
7121121B
- B
IF R$ = "I" THEN C
71121
B
(AA * TT * .0001)
IF R$ = "G" THEN C
715
F = C * (11211211210
/ T)
720
PI':tI 4
73121
740
PRINT E TAB( 10)PT
TAB( 24)F;"
II;C
IF N = 5 THEN
PRINT"
"
742
IF N = 1121THEN PRINT
744
745
IF N
15 THEN PRINT
IF N
2121 THEN
PRINT
746
IF N
25 THEN
PRINT
748
PRtI 3
75121
NEXT N
760
77121
PRtI 4
780
PRINT
790
PRINT "Sample
thickness
of ";T;"
microns
used."
80121
IF S$ = "¥" THEN PRINT "InP
";T$;"-type
substrate
-
=
=
-
=
=
=
=
=
=
=
=
-
=
=
=
=
RB)
A
.5
-
=
=
=
microns
t
";TT;"
hick
correctior,
included."
810
IF R$ = "G" THEN PRINT "Reflectance
for
InGaAs
used."
820
IF R:t. = "I"
THEN
PRINT "Reflectance
for
InP
used."
825
PRINT "XT scaling
factor
= ";S;"."
830
FOR X = 1 TO 10
PRINT
840
850
NEXT X
860
PRtI 3
870
PRINT
880
PRINT "end"
890
END
1010
IF N
1 THEN PRINT "Include interference effects? (Type ¥ or N)"
1020 IF N = 1 THEN INPUT G$
1030 IF G$ = "N" THEN GOTO 695
1040 LA = E / 1.23977
112150PRINT "Input absorption coefficient estimate, in cmA-1 (Type 0 if
ready for next energy)"
TABLE XVI.
140
(cont.)
1060 INPUT A
1065 IF A = 0 THEN BOTO 730
10700000)
XX =) (12.5664* NA * T * LA) + (A I (LA * 3.1416 * (NA
1080
1090
1095
1100
A * T * .0001
TB = EXP (C) + (RB I
BOTO 2000
IF N = 11 THEN E
=
125121
126121
127121
THEN E
IF N =
THEN E
IF N = 14 THEN E
IF N = 15 THEN E
IF N = 16 THEN E
IF N = 17 THEN E
IF N = 18 THEN E
IF N = 19 THEN E
IF N =
THEN E
IF N = 21 THEN E
IF N = 22 THEN E
IF N = 23 THEN E
IF N = 24 THEN E
IF N = 25 THEN E
IF N = 26 THEN E
IF N = 27 THEN E
IF N = 28 THEN E
128121
BOTO
1110
IF N
1120
1 13121
114121
1150
116121
117121
1180
119121
1200
121121
122121
123121
124121
* 1
EXP
-
(2 * RB
A
.5 *
cos (XX»
53121
1330
INPUT
PT
1332
energy
IF E < .8 THEN
(in
=
RA
IF E > .86 AND E <
<'0188
*
E)
1338
IF E > 1.1 THEN RA
134121
BOTO
56121
= .78 THEN
136121 IF E > .78 THEN NN
135121 IF E <
.668
BOTO
67121
= RA I TB
21211121PRINT
PT
21212121PRINT
Z
A
31213121
IF
N
31214121
IF
N
31215121 IF
N
31216121 IF
N
31217121 IF
N
31218121 IF
N
31219121 IF N
310121 BOTO
1050
=
1 THEN
2 THEN
E
E
3
4
5
6
7
E
E
E
E
E
THEN
THEN
THEN
THEN
THEN
= .7
= .8
= .9
= 1
=
1.1
= 1.2
= 1.25
8 THEN E = 1. 26
= 9 THEN E = 1.27
10 THEN E = 1.28
=
=
1100
(.136
* E):RB
= 1.1 THEN RA = .606
= .571 - (.02 * E):RB
NN = 3.56
= (.222 * E) + 3.375
137121
BOTO
IF N
IF N
IF N
-
=
.121285
+
(.1215 *
= .86 THEN RA = .477 + (.1 * E):RB
21210121Z
206121
300121
301121
302121
eV)."
XT."
1334(.121167
IF E *> E)=.8 AND E <
=
(C»
1.31
1.32
1.33
1.34
1.35
1.36
1.37
1.38
1.39
1.4
1.41
1.42
1.43
1.44
1.45
1.5
2121
"Enter
E
"Enter
21213121 F
2 - 1)
1.29
1.3
12
13
130121 PRINT
131121 INPUT
132121 PRINT
1336
A
=
C
( . 052
* E): RB
E)
= .121818
.0514 +
.121647+ (.00667 * E)
141
VITA
The author
After
graduating
Massachusetts
music,
Galileo
prevailed
In 1979,
overpowering,
engineer
at Tektronix,
Master
of Electronic
study began
because
A strong
interest
Subsequently,
in
from
he worked
at
as a research
of his home
a career
to
led to a conservatory;
his B.S. Physics
attraction
23, 1955.
in 1973, he travelled
Massachusetts,
to begin
on May
state proved
as a CRT manufacturing
Inc.
studying
on a part-time
in 1977.
the magnetic
born
organ, nearly
in Sturbridge,
and he returned
He began
School
and he received
Institute
Electro-Optics
Oregonian,
education.
for the baroque
Polytechnic
engineer.
High
for his undergraduate
science
Worcester
year.
from Beaverton
particularly
however,
1980,
is a third generation
at the Oregon
basis.
While
Science
degree
at OGC in January,
of the author's
still
1985.
marriage
Graduate
Center
at Tektronix,
at OGC in April,
in September
of
he completed
1984.
The present
This time was also significant
to Corinne
A. Abel,
the
in May of that