Low temperature silicon nitride and silicon dioxide film processing by inductively coupled plasma chemical vapor deposition

Article abstract of DOI:10.1149/1.1393382

High-density plasma technology is becoming increasingly attractive for the deposition of dielectric films such as silicon nitride and silicon dioxide. In particular, inductively coupled plasma chemical vapor deposition (ICPCVD) offers a great advantage for low-temperature processing over plasma-enhanced chemical vapor deposition (PECVD) for a range of devices including compound semiconductors and magnetic heads. In this paper, the development of low temperature (< 200°C) silicon nitride and silicon-dioxide films utilizing ICP technology is discussed. The material properties of these films have been investigated as a function of ICP source power, radio-frequency chuck power, chamber pressure, gas chemistry, and temperature. The ICPCVD films are compared to PECVD films in terms of wet etch rate, stress, and other film characteristics. Two different gas chemistries, SiH₄/N₂/Ar and SiH₄/NH₃/He, were explored for the deposition of ICPCVD silicon nitride. The ICPCVD silicon dioxide films were prepared from SiH₄/O₂/Ar. The wet etch rates of both silicon nitride and silicon dioxide films are significantly lower than films prepared by conventional PECVD. This implies that ICPCVD films prepared at these low temperatures are of higher quality. The advanced ICPCVD technology can also be used for efficient void-free filling of high aspect ratio (3:1) sub-micrometer trenches.

Full text of DOI:10.1149/1.1393382

Journal of The Electrochemical Society, 147 (4) 1481-1486 (2000)
1481
S0013-4651(99)06-103-0 CCC: $7.00 © The Electrochemical Society, Inc.
Low Temperature Silicon Nitride and Silicon Dioxide Film Processing by
Inductively Coupled Plasma Chemical Vapor Deposition
a,
a
a
a
b,
c,
J. W. Lee, * K. D. Mackenzie, D. Johnson, J. N. Sasserath, S. J. Pearton, ** and F. Ren **
a
b
Plasma-Therm Incorporated, St. Petersburg, Florida 33716, USA
Department of Materials Science Engineering and Department of Chemical Engineering, University of Florida, Gainesville,
c
Florida 32611, USA
High-density plasma technology is becoming increasingly attractive for the deposition of dielectric films such as silicon nitride and
silicon dioxide. In particular, inductively coupled plasma chemical vapor deposition (ICPCVD) offers a great advantage for low-
temperature processing over plasma-enhanced chemical vapor deposition (PECVD) for a range of devices including compound
semiconductors and magnetic heads. In this paper, the development of low temperature (< 200ЊC) silicon nitride and silicon diox-
ide films utilizing ICP technology is discussed. The material properties of these films have been investigated as a function of ICP
source power, radio-frequency chuck power, chamber pressure, gas chemistry, and temperature. The ICPCVD films are compared
to PECVD films in terms of wet etch rate, stress, and other film characteristics. Two different gas chemistries, SiH /N /Ar and
4
2
SiH /NH /He, were explored for the deposition of ICPCVD silicon nitride. The ICPCVD silicon dioxide films were prepared from
4
3
SiH /O /Ar. The wet etch rates of both silicon nitride and silicon dioxide films are significantly lower than films prepared by con-
4
2
ventional PECVD. This implies that ICPCVD films prepared at these low temperatures are of higher quality. The advanced ICPCVD
technology can also be used for efficient void-free filling of high aspect ratio (3:1) sub-micrometer trenches.
©
2000 The Electrochemical Society. S0013-4651(99)06-103-X. All rights reserved.
Manuscript submitted June 21, 1999; revised manuscript received December 15, 1999.
There is a growing interest in high-density plasma processing in
both the semiconductor and the magnetic thin film head industry.
process because ion energy is a major factor causing ion damage to
the device. With HDPCVD, the ion energy can be reduced to min-
1-7
22
In particular, much research has been conducted on dry etching with
inductively coupled plasma (ICP) sources because they provides ad-
imize damage to the devices. The process pressures for PECVD and
HDPCVD are quite different. More than 500 mTorr is common for
PECVD. The pressure range of HDPCVD is 1-30 mTorr. We report
8-10
vanced processes for pattern transfer.
A great deal of research has
been reported for dielectric film deposition using remote or high-den-
on the effect of rf chuck power and pressure for SiN films deposit-
ed by HDPCVD with an ICP source.
x
11-14
sity plasmas.
However, relatively little work has been done on
15,16
deposition technology for dielectric materials using ICP.
an ICP source, we explored high-density plasma (>10 cmϪ3) chem-
Using
Experimental
11
Both SiN and SiO films were deposited in a Plasma-Therm
x
2
ical vapor deposition (HDPCVD) of SiN and SiO . In comparing
x
2
HDPCVD Versalock system on 4 in. Si wafers. Figure 1 shows a
schematic of the ICP chamber configuration. The load module of the
HDPCVD technology with conventional plasma-enhanced chemical
vapor deposition (PECVD), some potential advantages are lower hy-
drogen content films, higher quality films at lower process tempera-
tures (<200ЊC), void-free gap filling of high aspect ratio features, and
self-planarization. Low temperature SiNx film deposition of low
hydrogen content by HDPCVD is of special interest for cap and
system can handle up to about 50 wafers at a time. For the SiN depo-
x
sition, two different gas chemistries, SiH /N /Ar and SiH / NH /He,
4
2
4
3
were explored. A gas chemistry of SiH /O /Ar was used for the SiO
4
2
2
film deposition. Electronic mass flow controllers regulated all gas
flows. The SiH was fed through a lower gas ring (gas inlet 2). All
17-21
4
capacitor layers in III-V semiconductor devices.
Due to the rela-
other gases entered the chamber via a showerhead in the ICP source
tively low dissociation efficiency of N , typical process recipes for
2
(gas inlet 1). The process chamber is a hybrid configuration consist-
SiN deposition by PECVD use NH as the source of nitrogen. There-
x
3
ing of an ICP source and an rf powered wafer chuck. An oil recircu-
lating heat exchanger connected to the chuck controls the wafer tem-
perature. Helium back-side cooling of the wafer is used for efficient
heat transfer. For this work, the ICP source power and chuck temper-
fore, some portion of hydrogen incorporation from NH in deposited
3
SiN films is inevitable. However, HDPCVD technology enables us
x
to deposit SiN with a NH -free recipe because high-density plasma
x
3
sources have typically one order of magnitude higher ion dissociation
efficiency (i.e., ϳ0.1% for PECVD and ϳ1% for HDPCVD).
Some advantages of an ICP source over other types of high-den-
sity sources include easier scale up, advanced automatic tuning for
the source, and lower cost of ownership. In addition, with a hybrid
ICP configuration, such as used in this work, it is possible through
the addition of radio-frequency (rf) power to the wafer chuck, to
control ion flux and ion energy independently. This expands the
applications for dielectric film deposition by ICP. For example, gap-
filling techniques require simultaneous high ion bombardment by an
inert gas such as Ar during deposition to prevent void formation. To
achieve high ion bombardment, high ion energy can be induced by
controlling the rf power on the wafer chuck in HDPCVD without
change to the ion density in the source. In PECVD, that approach is
not feasible and alternating sequences of deposition and sputter etch
are usually required. This requires extra steps and more process
time. For damage sensitive devices, such as the high electron mobil-
ity transistor (HEMT), it is essential to use a very low ion energy
*
*
* Electrochemical Society Active Member.
* Electrochemical Society Fellow.
Figure 1. Schematic of the ICP chamber.
Downloaded on 2015-01-16 to IP 141.214.17.222 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

1
482
Journal of The Electrochemical Society, 147 (4) 1481-1486 (2000)
S0013-4651(99)06-103-0 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 2. (a) Ion density and (b) negatively induced dc bias as a function of
rf chuck power in an ICP chamber for Ar plasmas.
Figure 3. Optical emission spectrum of an ICP SiH /N /Ar plasma at 800 W
ICP, 120 W rf, and 10 mTorr.
4
2
ature were fixed at 800 W and 150ЊC, respectively. We found that
lower deposition temperatures lead to poor quality dielectrics with
relatively high wet etch rates. The rf chuck power and chamber pres-
sure were varied from 25 to 150 W and 1 to 20 mTorr, respectively.
Langmuir probe measurements of ICP Ar plasmas were used to
determine the ion density characteristics of the process chamber.
In situ optical emission spectroscopy (OES) was also used to char-
acterize the ICP SiH /N /Ar plasmas.
measurements were done with a Tencor model P-2 profilometer. A
buffered oxide etch (BOE) solution of 7:1 NH : HF was used for the
4
wet-etch rate measurements.
Results and Discussion
ICPCVD plasma characterization.—Figure 2 shows ion density
and negatively induced dc bias as a function of ICP source power
and rf chuck power for Ar plasmas at a fixed pressure of 10 mTorr.
These data were obtained from a design of experiment simulation,
which we developed with experimental data obtained on the ICP
process chamber. As shown in Fig. 2a, the ion density of the Ar plas-
4
2
Deposition rate and uniformity were determined by a NanoSpec
model 4150 metrology system. Refractive index measurements were
made on a Gaertner model L116D-PC ellipsometer. Film stress
Figure 4. Detailed optical emission spectra for an ICP SiH /N /Ar plasma for rf chuck powers of 20 and 120 W. The ICP power and pressure were 800 W ICP
4
2
and 10 mTorr, respectively.
Downloaded on 2015-01-16 to IP 141.214.17.222 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 147 (4) 1481-1486 (2000)
1483
S0013-4651(99)06-103-0 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 7. BOE rate of ICPCVD SiN as a function of rf chuck power.
x
Figure 5. SiN deposition rate as a function of rf chuck power.
x
The typical deposition rate was in the range 400-600 Å/min. This is
comparable to conventional PECVD using the same precursors at
much higher pressures.
Figures 6 and 7 show the variation of refractive index and buffered
oxide etch (BOE) rate with rf chuck power. At 25 W rf chuck power,
a refractive index of 1.95 was achieved. The combination of decrease
of deposition rate and BOE rate, and increase of refractive index with
rf chuck power indicates that deposited films become slightly Si-rich
ma can be controlled by ICP source power and is almost independent
of rf chuck power. The Ar ion density is about 2 ϫ 10 cm with
1
1
Ϫ3
8
1
00 W ICP power, 20 sccm Ar, 10 mTorr chamber pressure, and
00 W rf chuck power. As illustrated in Fig. 2b, the dc bias on the
chuck was a strong function of ICP source power and rf chuck
power. The dc bias increased with rf chuck power and decreased
with ICP source power.
(
Si fraction increased from 0.51 to 0.57). This was confirmed by
Silicon nitride: optical emission spectra.—In Fig. 3, a typical
Auger electron spectroscopy (AES) measurements.
optical emission spectrum (OES) is presented for an ICP SiH /N /Ar
4
2
One possible explanation for the increase of refractive index with
plasma. This spectrum was recorded during a SiN deposition. All
x
rf power is that the dissociation of SiH into reactive neutrals in the
SiH /N /Ar plasma increased relative to that of N with rf chuck
power (in other words, ion energy). Another possible explanation is
that high rf power brings more Si-related species than N to partici-
pate in the film growth. Table I shows the dissociation energy of
4
N , SiH, H , and Ar related peaks were identified. Figure 4 shows
2
2
4
2
2
more detailed OES data taken at different rf powers. Many N relat-
2
ϩ
ed peaks were detected between 300 and 400 nm. A N2 peak is also
noticed at 391.4 nm. A SiH and two H peaks are observed at 414.2,
2
4
86.1, and 656.2 nm, respectively. The peak intensities of both SiH
related diatomic molecules for SiN deposition. Note that the N–N
x
and H increased with increasing rf power.
2
bond has a much higher bond strength (945.33 kJ/mol) than that of
Si–H (Յ299.2 kJ/mol). Dissociation of N2 requires more energy
than that of SiH . This data suggests that more SiH species are
Silicon nitride: film properties.—Results for SiH /N /Ar.—In
4
2
Fig. 5, the dependence of deposition rate on rf chuck power is shown.
The deposition rate decreases with increasing rf chuck power. At
high rf power, it is expected that more ion bombardment during
deposition will occur. High rf power increases the dc bias and con-
4
4
brought to the substrate than N with increase of rf chuck power.
2
It is worthwhile commenting further on the BOE data presented
in Fig. 7. Considering the low deposition temperature of 150ЊC, the
typical BOE rate of 1000 Å/min is very low. At 150ЊC, the BOE rate
of PECVD SiN is greater than 5,000 Å/min. This implies that the
ICPCVD SiN films are denser and contain less hydrogen.
ϩ
sequently the ion energies of the bombarding species, e.g., Ar and
ϩ
N , will increase. This will enhance the sputtering of Si and N reac-
x
2
25
tive neutrals before and during the formation of the SiN film. This
x
x
As indicated in Fig. 8, the stress of all ICPCVD SiN films pre-
accounts for the observed reduction in deposition rate with rf power.
x
pared from SiH /N /Ar was compressive. At 25 W rf chuck power,
4
2
the film stress was as high as 900 MPa. Raising the rf chuck power
lowers the film stress. A stress of 300 MPa is achieved at 120 W.
Film stress of less than 500 MPa is acceptable for many applications.
We speculate that incorporated hydrogen relieves the stress in the
SiN film, although the correlation with rf chuck power is not fully
x
understood. The typical H content in our SiN films ranged from 15-
x
2
0 atom %. compared to 25-30 atom % for PECVD films deposited
at the same temperatures. In PECVD, the film stress increases with
rf power regardless of whether it is compressive or tensile. This is
diametrically opposite to the current observations with ICPCVD.
Table I. Dissociation energies of diatomic molecules.
Bond
Dissociation energy at room temperature (kJ/mol)
Si–H
N–H
N–N
H–H
Յ299.21
Յ339.11
Ն945.33
Ն435.99
Figure 6. Refractive index of ICPCVD SiN as a function of rf chuck power.
x
Downloaded on 2015-01-16 to IP 141.214.17.222 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

1
484
Journal of The Electrochemical Society, 147 (4) 1481-1486 (2000)
S0013-4651(99)06-103-0 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 8. Compressive stress of ICPCVD SiN as a function of rf chuck
power.
Figure 10. Deposition rate of ICPCVD SiN as a function of chamber pressure.
x
x
Figure 9 shows deposition uniformity vs. rf chuck power. The typ-
ical uniformity was <4%. The uniformity becames somewhat worse
as the rf chuck power was increased. It was observed that the center
of the deposited films was thicker than edge of the wafer.
As presented in Fig. 10 and 11, raising the chamber pressure
increased both the deposition rate and refractive index. The deposi-
tion rate increases from 350 Å/min at 1 mTorr to 600 Å/min at
2
0 mTorr. Over this pressure range, the refractive index increases
from 1.92 to 2.32. As shown in Fig. 12, the BOE rate decreases from
200 to 400 Å/min as the pressure is increased. At high pressure, the
residence time of gas species increases by
1
␶
ϭ PV/Q
[1]
where ␶, P, V, and Q are the residence time, pressure, volume, and
throughput, respectively. With longer residence time and higher Ar
ion density at high chamber pressure, more SiH may be dissociated
4
into neutrals, which will be incorporated in the films and cause a Si-
rich composition at high pressure. Increased residence time or higher
Ar ion density may contribute to preferential sputtering of the lighter
N atoms from the film. It is well known that increased refractive
index with Si-rich composition reduces the BOE rate of the films.
As shown in Fig. 13, the stress of the deposited films is constant
at about 400 MPa, compressive over the pressure range of 1-
Figure 11. Refractive index of ICPCVD SiN as a function of chamber pressure.
x
increase in refractive index, i.e., Si-enrichment. This change in film
stoichiometry increases the compressive stress.
Results for SiH /NH /He.—ICPCVD SiN deposition from SiH4/
4
3
x
1
5 mTorr. However, at higher pressure (>15 mTorr), stress increased
NH /He was also investigated. Table II summarizes some of data ob-
3
rapidly and reached 1300 MPa at 20 mTorr. This correlated with the
Figure 9. Deposition uniformity of ICPCVD SiN as a function of rf chuck
power.
x
Figure 12. BOE rate of ICPCVD SiN as a function of chamber pressure.
x
Downloaded on 2015-01-16 to IP 141.214.17.222 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 147 (4) 1481-1486 (2000)
1485
S0013-4651(99)06-103-0 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 14. Deposition rate of ICPCVD SiN as a function of rf chuck power.
Figure 13. Compressive stress of ICPCVD SiN as a function of chamber
x
x
pressure.
tained. The process conditions were 800 W ICP power, 30 W rf
chuck power, 10 mTorr, and 150ЊC. Compared with SiH /N /Ar, the
4
2
deposition rates were slightly lower, and the stress was higher. In
general, Ar is a superior choice compared with He, due to lower
stress in the deposited films.
Deposition rate was not a strong function of rf chuck power
(Fig. 14) while stress was controlled by rf chuck power (Fig. 15).
The film stress was compressive at low rf chuck power and it became
tensile with more rf chuck power.
Silicon dioxide: gap filling.—Processing of some electronic devices,
for example, III-V semiconductors and magnetic heads have a limit
on the temperature. Conventional high-temperature PECVD silicon
dioxide processing is not suitable for those materials. The applica-
tion of ICPCVD SiO for void-free filling of 1 ␮m gap features was
2
investigated. The deposition temperature was 100ЊC. Table III sum-
marizes some of the main properties of these films.
Figure 16 shows an SEM micrograph of a gap-filled Si trench.
The aspect ratio of the trench was about 3:1. Note that there is no
void after gap filling is completed. A similar technique was applied
on an electroplated metal which had an irregular shaped sidewall.
Figure 17 shows self-planarization after gap filling. Note that there
were small voids after the process was completed. The origin of the
void is from the irregularity of sidewall at the bottom of the trench.
As illustrated in Fig. 18, with modified process conditions to com-
pensate for the sidewall problem of the control samples, the trench
Figure 15. Compressive stress of ICPCVD SiN as a function of rf chuck
power.
x
pared to PECVD SiN deposited at the same low temperature. This
x
implies that the ICPCVD SiN films are denser and contain less
x
hydrogen. An ICPCVD SiN process for adjustable film stress was
x
demonstrated using SiH /NH /He.
4
3
ICPCVD SiO films were prepared from SiH /O /Ar. Void-free
2
4
2
can be filled with ICPCVD SiO without any void. Further process-
2
gap filling of high aspect ratio trench was demonstrated with
ICPCVD SiO2.
In conclusion, high-density ICPCVD technology clearly has many
advantages compared to PECVD for dielectric film deposition in a
broad range of advanced processes for electronic device fabrication,
especially at low temperatures.
ing (either deposition, or alternatively, chemical mechanical polish-
ing) would bring self-planarization after gap filling.
Conclusions
We investigated low-temperature deposition (150ЊC) of ICPCVD
SiN prepared from SiH /N /Ar and SiH /NH /He. For SiN films
x
4
2
4
3
x
prepared from SiH /N /Ar, rf chuck power increased refractive in-
Acknowledgments
4
2
dex and deposition uniformity while it decreased deposition rate,
stress, and BOE rate. Increasing rf chuck power may bring more Si
species than N species to the wafer surface leading to Si-rich films.
Raising the chamber pressure resulted in higher deposition rate, re-
The authors appreciate Dr. J. Donohue, R. Westerman, and Wei
Pen for fruitful discussions and assistance. We also thank R. McAfee
and L. Heckerd for their technical support. The work at the Univer-
sity of Florida is partially supported by a DOD MURI, monitored by
AFOSR (H. C. DeLong), contract no. F49620-96-1-0026.
fractive index, and film stress. The BOE rate of ICPCVD SiN , pre-
x
pared from either SiH /N /Ar or SiH /NH /He, is much lower com-
4
2
4
3
Table III. Characteristics of ICPCVD SiO film.
2
Table II. Characteristics of ICPCVD SiN with SiH /NH /He.
x
4
3
Gas chemistry
Deposition rate
Film stress
SiH /O /Ar
4 2
Refractive index
Deposition rate
Film stress
2.0
300 Å/min
200 MPa, (compressive)
BOE Rate
Uniformity
1200 Å/min
Ϯ2.5%
ϳ1000 Å/min
300 MPa, (compressive)
1.46
Refractive index
Downloaded on 2015-01-16 to IP 141.214.17.222 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

1
486
Journal of The Electrochemical Society, 147 (4) 1481-1486 (2000)
S0013-4651(99)06-103-0 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 18. Partially gap-filled metal coils by ICPCVD SiO . No voids are
2
Figure 16. SEM micrograph showing a Si trench filled with ICPCVD SiO2.
evident. The SiO2 film is planarized.
6
.
.
K. B. Jung, J. Hong, H. Cho, S. Onishi, D. Johnson, Y. D. Park, J. R. Childress, and
S. J. Pearton, J. Vac. Sci. Technol., A, 17, 535 (1999).
R. J. Shul, G. B. McClellan, R. D. Briggs, D. J. Rieger, S. J. Pearton, C. R. Aber-
nathy, J. W. Lee, C. Constantine, and C. Barratt, J. Vac. Sci. Technol., A, 15, 633
7
(1997).
8
.
.
J. W. Lee, E. S. Lambers, C. R. Abernathy, S. J. Pearton, R. J. Shul, F. Ren, W. S.
Hobson, and C. Constantine, Solid-State Electron., 42, A65 (1998).
R. J. Shul, C. G. Wilson, M. M. Bridges, J. Han, J. W. Lee, S. J. Pearton, C. R.
Abernathy, J. D. MacKenzie, L. Zhang, and L. F. Lester, J. Vac. Sci. Technol., A,
9
1
6, 1621 (1998).
1
0. F. Ren, J. W. Lee, C. R. Abernathy, R. J. Shul, C. Constantine, and C. Barratt, Semi-
cond. Sci. Technol., 12, 1154 (1997).
1
1. G. Lucovsky, P. D. Richard, D. V. Tsu, S. Y. Lin, and R. J. Markunas, J. Vac. Sci.
Technol., A, 4, 681 (1986).
1
1
2. C. B. Labelle, S. J. Limb, and K. K. Gleason, J. Appl. Phys., 82, 1784 (1997).
3. S. J. Limb, C. B. Labelle, K. K. Gleason, D. J. Edell, and E. F. Gleason, Appl. Phys.
Lett., 68, 2810 (996).
1
1
1
1
4. M. Lapeyrade, M. P. Besland, C. Meva’a, A. Sibai, and G. Hollinger, J. Vac. Sci.
Technol., A, 17, 433 (1999).
5. Y. B. Hahn, J. W. Lee, K. D. Mackenzie, D. Johnson, S. J. Pearton, and F. Ren,
Solid-State Electron., 42, 2017 (1998).
6. Y. B. Hahn, J. W. Lee, K. D. Mackenzie, D. Johnson, D. Hays, C. R. Abernathy,
and S. J. Pearton, Electrochem. Solid-State Lett., 1, 230 (1998).
7. J. W. Lee, K. D. Mackenzie, D. Johnson, R. J. Shul, S. J. Pearton, C. R. Abernathy,
and F. Ren, Solid-State Electron., 42, 1031 (1998).
Figure 17. Gap-filled and self-planarized SiO on metal coils by ICPCVD.
18. J. W. Lee, C. R. Abernathy, S. J. Pearton, F. Ren, R. J. Shul, C. Constantine, and C.
2
Barratt, Solid-State Electron., 41, 829 (1997).
1
9. F. Ren, J. W. Lee, C. R. Abernathy; S. J. Pearton, R. J. Shul, C. Constantine, and
C. Barratt, Semicond. Sci. Technol., 12, 1154 (1997).
Plasma-Therm assisted in meeting the publication costs of this article.
2
0. A. Kapila and V. Malhotra, in Twenty-Seventh State-of-the-Art Program on Com-
pound Semiconductors (SOTAPOCS XXVII), S. N. G. Chu, D. N. Buckley, K.
Wada, J. P. Vilcot, W. Pletschen, D. DeCoster, P. Van Daele, and H. Q. Hou, Edi-
tors, PV 97-21, p. 322, The Electrochemical Society Proceedings Series, Penning-
ton, NJ (1997).
References
1.
2.
3.
4.
5.
C. Constantine, C. Barratt, S. J. Pearton, F. Ren, and J. R. Lothian, Appl. Phys.
Lett., 61, 2899 (1992).
J. W. Lee, H. Cho, C. Hays, C. R. Abernathy, S. J. Pearton, R. J. Shul, G. A. Vawter,
and H. Han, IEEE J. Selected Top. Quantum Electron., 4, 557 (1998).
F. Ren, J. W. Lee, C. R. Abernathy, S. J. Pearton, C. Constantine, C. Barratt, and
R. J. Shul, J. Vac. Sci. Technol., B, 15, 983 (1997).
S. J. Pearton, J. W. Lee, E. S. Lambers, J. R. Mileham, C. R. Abernathy, W. S. Hob-
son, and F. Ren, J. Vac. Sci. Technol., B, 14, 118 (1996).
21. R. E. Sah, V. Weiman, H. Baumann, J. Wagner, R. Kiefer, and S. Muller, in Twen-
ty-Seventh State-of-the-Art Program on Compound Semiconductors, D. N. Buck-
ley, S. N. G. Chu, H. Q. Hou, R. E. Sah, J. P. Vilcot, and M. J. Deen, Editors, PV 97-
1, p. 179, The Electrochemical Society Proceedings Series, Pennington, NJ (1997).
22. J. W. Lee, K. D. Mackenzie, D. Johnson, R. J. Shul, S. J. Pearton, C. R. Abernathy,
and F. Ren, Solid-State Electron., 42, 1021 (1998).
S. Thomas III, K. K. Ko, and S. W. Pang, J. Vac. Sci. Technol., A, 13, 894 (1994).
23. J. Kanicki, Mater. Res. Soc. Symp. Proc., 118, 671 (1988).
Downloaded on 2015-01-16 to IP 141.214.17.222 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).