Thermal wet oxidation of GaP and Al0.4Ga0.6P
J. H. Epple, K. L. Chang, G. W. Pickrell, K. Y. Cheng, and K. C. Hsieh
Citation: Appl. Phys. Lett. 77, 1161 (2000); doi: 10.1063/1.1286871
View online: http://dx.doi.org/10.1063/1.1286871
View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v77/i8
Published by the American Institute of Physics.
Related Articles
Ge redistribution in SiO2/SiGe structures under thermal oxidation: Dynamics and predictions
J. Appl. Phys. 111, 024308 (2012)
Effect of the capping on the local Mn oxidation state in buried (001) and (110) SrTiO3/La2/3Ca1/3MnO3
interfaces
J. Appl. Phys. 110, 103903 (2011)
Stress release phenomena in chromia scales formed on NiCr-30 alloys: Influence of metallurgical parameters
J. Appl. Phys. 110, 093516 (2011)
Electrical properties of thermally oxidized AlInN/AlN/GaN-based metal oxide semiconductor hetero field effect
transistors
J. Appl. Phys. 110, 084501 (2011)
Stress evolution during the oxidation of silicon nanowires in the sub-10 nm diameter regime
Appl. Phys. Lett. 99, 143115 (2011)
Additional information on Appl. Phys. Lett.
Journal Homepage: http://apl.aip.org/
Journal Information: http://apl.aip.org/about/about_the_journal
Top downloads: http://apl.aip.org/features/most_downloaded
Information for Authors: http://apl.aip.org/authors
Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions
APPLIED PHYSICS LETTERS
VOLUME 77, NUMBER 8
21 AUGUST 2000
Thermal wet oxidation of GaP and Al0.4Ga0.6P
J. H. Epple, K. L. Chang, G. W. Pickrell, K. Y. Cheng, and K. C. Hsieha)
Department of Electrical and Computer Engineering and Microelectronics Laboratory,
University of Illinois at Urbana-Champaign, Urbana, Illinois 61820
共Received 24 March 2000; accepted for publication 19 May 2000兲
Thermal wet oxidations of GaP and Al0.4Ga0.6P at 650 °C for various times have been performed.
Comparisons are made on oxidation rates and post oxidation morphology. Transmission electron
microscopy shows that when oxidizing GaP, polycrystalline monoclinic GaPO4•2H2O forms
without noticeable loss of phosphorus. Oxidation for 6 h or more leads to poor morphology resulting
in cracks and detachment. A thickness expansion of about 2.5–3 times is noticed as a result of
oxidation. In contrast, oxidized Al0.4Ga0.6P exhibits much better morphology without cracks or
detachment from the substrate. The oxide has an almost amorphous-like microstructure. The
oxidation process shows typical diffusion-limited reaction at long anneals. Preliminary work on the
oxidation of AlP indicates that the reaction leads to formation of Al2O3 and possible volatile P2O5
diffusing out of the specimen. Thus, from the structural viewpoint, AlGaP forms a better oxide
suitable for device needs. © 2000 American Institute of Physics. 关S0003-6951共00兲02228-2兴
metalorganic chemical vapor deposition system. The material was grown on a 共100兲 GaP substrate miscut 10° towards
the 关111兴 direction. Oxidation was performed in a Lindberg
open tube furnace. Water vapor was supplied to the samples
using nitrogen bubbled through a water reservoir kept at
85 °C. Samples were oxidized for times ranging from 1 to 36
h at 650 °C. After oxidation, the samples were analyzed using cross-sectional transmission electron microscopy 共TEM兲
to determine oxidation rates, volume changes, and microstructure. Compositions were estimated using energy dispersive x-ray spectroscopy 共EDAX兲.
GaP was wet oxidized at 650 °C for varying times up to
24 h. After an oxidation of 6 h the surface of the oxidized
GaP began to get very rough, cracked, and detached as was
reported for thermal wet oxidations of GaP by Kato et al.8
This can be seen in a Nomarski photograph of the GaP oxidized at 650 °C for 24 h shown in Fig. 1共a兲. This oxide has a
thickness of ⬃1.3 ␮m and is obviously very rough and
cracked as shown in the cross-sectional TEM micrograph,
Fig. 1共b兲. Also seen is the detachment between the oxide and
substrate. The oxidized material was identified using TEM
diffraction 关insert in Fig. 1共b兲兴 and EDAX. The x-ray spectroscopy confirmed the presence of oxygen and equal atomic
concentrations of Ga and P. Among various possible GaPOx
compounds listed by the International Center for Diffraction
Data, the hydrous monoclinic GaPO4•2H2O is the only structure that matches the TEM diffraction data.11 TEM also
showed that the oxide was polycrystalline without any preferred crystal orientation correlation between the oxide and
substrate. GaP oxidized for less than 6 h did not have as
severe structural damage or detachment. However, there
were noticeable spots on the surface. It was also noticed that
the oxide was 2.5–3 times the thickness of the original material. This can be attributed to the absorption of O and H2O.
It is indicated by the breakdown after longer oxidation times
that there is likely a critical volume for the highly strained
oxide at which it will crack and detach from the substrate.
For a comparison, published results on the oxidation of
For decades the oxidation of many compound semiconductors has been investigated to yield native oxides for different applications. Most beneficial has been the study of
oxidizing Alx Ga1⫺x As. 1 This technology has been used for
numerous applications such as distributed Bragg reflectors in
vertical cavity surface emitting lasers,2,3 insulating layers in
field effect transistors, and current guiding in heterojunction
bipolar transistors.4–6 Less extensive work has been done to
study the materials grown on GaP. Rubenstein and Kato
et al. have studied the thermal dry oxidation of GaP.7,8 They
reported that the product of the oxidation was mostly GaPO4
with very little Ga2O3. This is in contrast to both GaAs and
GaSb, which when oxidized form primarily Ga2O3 with the
As and Sb forming volatile oxides that diffuse out of the
sample. Dry oxidations for 30 min at temperature of 850 °C
resulted in the GaPO4 being very rough with caverns forming
at the oxide/substrate interface. In contrast, when GaP was
thermally wet oxidized the material was smoother without
the cavern formation but was cracked and separated from the
substrate.8 It was determined that thermally oxidized GaP
could not be used for device applications.
Following this there was much work done on the anodic
oxidation of GaP. Schwartz experimented with annealing anodically oxidized GaP with various P2O5 overpressures to
formulate a Ga–P–O ternary phase diagram.9 The anodic
oxide consisted of P2O5 and Ga2O3. His results indicated that
GaPO4 formed after annealing. Kato et al. also did work annealing anodically oxidized GaP in a nitrogen ambient instead of P2O5 overpressure.10 It was found that when the
anodic oxide was annealed at temperatures above 600 °C, the
surface became very rough with hollow bumps and blisters
believed to be caused by vaporization of P2O5 in the material.
In this letter results are presented on the thermal wet
oxidation of both Al0.4Ga0.6P and GaP. A 1.5 ␮m Al0.4Ga0.6P
layer was grown in an Emcore GS 3000 DFM low-pressure
a兲
Electronic mail: [email protected]
0003-6951/2000/77(8)/1161/3/$17.00
1161
© 2000 American Institute of Physics
Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions
1162
Epple et al.
Appl. Phys. Lett., Vol. 77, No. 8, 21 August 2000
FIG. 2. Oxidation distance vs time for Al0.4Ga0.6P oxidized at 650 °C.
FIG. 1. Nomarski photograph 共a兲 and TEM and diffraction pattern 共b兲 of
GaP oxidized at 650 °C for 24 h. The GaP is cracked and damaged. Note the
detachment at the oxide/substrate interface. The TEM also shows the material is polycrystalline. A typical diffraction of a small grain is shown in the
insert, and diffraction analysis indicates that the material is monoclinic
GaPO4 •2H2O.
GaAs were analyzed. Rubenstein published data on the oxidation of GaAs with his GaP work.7 It is proposed that upon
wet oxidation, GaAs forms Ga2O3 and some volatile material
such as As2O3 that diffuses out of the sample. Also, Oh
et al.12 have noticed that thermal wet oxidation of GaAs
leads to a volume shrinkage of about 5%–10%. Thus, GaP
and GaAs oxidize very differently. GaP will retain its
column-V element and will form GaPO4 whereas GaAs will
not hold on to its column-V element and will form Ga2O3
leaving the As to form volatile compounds. This suggests
that the formation of GaPO4 is more favorable than that of
Ga2O3.
With the recent success in various applications of wet
oxidized Al-bearing arsenides, it is of interest to study the
oxidation of Alx Ga1⫺x P. Two different aluminum contents
have been chosen for the study, i.e., 40% and 100%. To
study the oxidation of AlP, a sample was grown with 1500 Å
of GaP covering a 1000 Å AlP layer. The AlP was then
oxidized from the top down through the GaP cap layer at
650 °C for a sufficiently long time and analyzed by TEM.
The resulting material showed again the presence of oxygen
but with a significant drop in the P content as compared to
aluminum. The P likely forms P2O5 and diffuses out of the
sample. In addition, the loss of P results in a volume shrink-
age of about 10%. Apparently, the oxidation mechanism of
AlP is very different from that of GaP.
Sugg et al. studied the thermal wet oxidation of
Al0.8Ga0.2As. 13 It was found that after oxidation the material
was mostly polycrystalline ␥ -Al2O3 with As being below the
detection level for Auger electron spectroscopy. At the oxidation temperatures used 共⬃425 °C兲 it was found that Ga did
not oxidize. Thus, the resulting products are Al2O3 and elemental Ga. Choquette et al. presented information on the
thermal wet oxidation of AlAs.14 It is stated that when AlAs
reacts with water the products are ␥ -Al2O3, As2O3, and H2.
Thermodynamically, As2O3 can react with H2 to form elemental As, and both As2O3 and As are volatile.15 They can
easily diffuse and escape from the sample. As a result of the
material loss, a volume shrinkage of 12%–13% after oxidation is experimentally observed. Hence, the results of oxidizing AlP in this work are similar to those of oxidizing AlAs.
The Al easily forms Al2O3 leaving the column-V element to
form volatile compounds.
Oxidations of Al0.4Ga0.6P were done for various times up
to 36 h at 650 °C. The oxidation thickness versus time is
shown in Fig. 2. It should be noted that the graph indicates
growth of the oxide and not consumption of the material
being oxidized. As can be seen in Fig. 2, the oxidation rate
starts out at about 700 Å/h but begins to slow as oxidation
times approach 20 h. This roll off is caused by the oxidation
becoming diffusion limited rather than reaction limited. On
the oxidation layer, EDAX showed that the relative amount
of P dropped slightly compared to the as-grown sample. This
indicates there was some loss of P due to the oxidation. It is
believed that some volatile P2O5 has formed and diffused out
of the sample. This phenomenon was not noticed in GaP.
After the oxidation it was found the material swelled to
roughly twice its original volume. Compared to oxidized
GaP, the volume expansion 共increase in thickness兲 of oxidized Al0.4Ga0.6P is smaller, which further supports the notion that AlP shrinks upon wet oxidation probably due to the
loss of volatile P2O5.
The postoxidation morphology of the Al0.4Ga0.6P and
GaP can be used as evidence that the damage to oxidized
material is caused by excessive expansion of a polycrystalline material. A Nomarski photograph of Al0.4Ga0.6P oxidized at 650 °C for 24 h is shown in Fig. 3共a兲. The oxidized
Al0.4Ga0.6P has a thickness of ⬃1.5 ␮m and is very smooth.
Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions
Epple et al.
Appl. Phys. Lett., Vol. 77, No. 8, 21 August 2000
1163
from ellipsometer measurements that the index of refraction
is ⬃1.7–1.8, which is significantly smaller than those of
crystalline AlGaP.
In conclusion, GaP and Al0.4Ga0.6P are oxidized at
650 °C for various times up to 36 h. The result of the GaP
oxidation is the formation of polycrystalline monoclinic
GaPO4 •2H2O. This material tends to have poor structural
quality and detaches from the substrate, probably due to a
large volume expansion. Oxidizing a buried layer of AlP
shows similarities to AlAs in that the column-V element diffuses out of the sample leaving Al2O3 and showing a volume
shrinkage. Al0.4Ga0.6P oxidizes at a faster rate than does GaP
and the resulting oxide has a smooth morphology with no
cracking or detachment even after 36 h of oxidation despite
an expanded volume. The smooth morphology and large
change in the index of refraction makes the oxide attractive
for different applications in passivation and waveguiding.
The authors greatly appreciate the use of the facility at
the Center for Microanalysis of Materials, University of Illinois, which is supported by the U.S. Department of Energy
under Grant No. DEFG02-91-ER45439. This work is further
supported by DARPA DAAG 55-98-1-0303 and F49602-981-10496.
1
FIG. 3. Nomarski photograph 共a兲 and TEM and diffraction pattern 共b兲 of
Al0.4Ga0.6P oxidized at 650 °C for 24 h. The Nomarski photograph shows
the oxidized material is very smooth in contrast to the oxidized GaP. The
TEM diffraction analysis indicates that the material is practically amorphous.
Shown in Fig. 3共b兲 is the TEM micrograph of the Al0.4Ga0.6P
after oxidation. In contrast to GaP, the oxidized Al0.4Ga0.6P
has no noticeable structural problems as seen in the oxidized
GaP. The diffraction pattern in Fig. 3共b兲 shows practically
only amorphous rings, which indicates that the oxidized
Al0.4Ga0.6P either consists of extremely fine polycrystalline
grains or is basically amorphous in structure. This is in contrast to the obvious polycrystalline material after the oxidation of GaP. It is possible that when the Al0.4Ga0.6P is oxidized it forms a glassy phase material that conforms to the
substrate. This material is not subject to the same problems
caused by large expansion the way polycrystalline material
is. Thus, instead of being cracked and detached the material
is smooth and adheres well. This is an important advantage
to the oxidation of Alx Ga1⫺x P. In addition, it was found
J. M. Dallesasse, N. Holonyak, Jr., A. R. Sugg, T. A. Richard, and N.
El-Zein, Appl. Phys. Lett. 57, 2844 共1990兲.
2
K. D. Choquette, K. L. Lear, R. P. Schneider, Jr., and K. M. Geib, Appl.
Phys. Lett. 66, 3413 共1995兲.
3
J. J. Wierer, P. W. Evans, N. Holonyak, Jr., and D. A. Kellogg, Appl.
Phys. Lett. 71, 2286 共1997兲.
4
E. I. Chen, N. Holonyak, Jr., and S. A. Maranowski, Appl. Phys. Lett. 66,
2688 共1995兲.
5
A. Massengale, M. C. Larson, C. Dai, and J. S. Harris, Jr., Electron. Lett.
32, 399 共1996兲.
6
P. A. Parikh, P. M. Chavarkar, and U. K. Mishra, IEEE Electron Device
Lett. 18, 111 共1997兲.
7
M. Rubenstein, J. Electrochem. Soc. 113, 540 共1966兲.
8
Y. Kato, K. M. Geib, R. G. Gann, P. R. Brusenback, and C. W. Wilmsen,
J. Vac. Sci. Technol. A 2, 588 共1984兲.
9
G. P. Schwartz, G. J. Gualtieri, J. E. Griffiths, C. D. Thurmond, and B.
Schwartz, J. Electrochem. Soc. 127, 2488 共1980兲.
10
Y. Kato, K. M. Geib, R. G. Gann, P. R. Brusenback, and C. W. Wilmsen,
J. Vac. Sci. Technol. B 9, 1530 共1991兲.
11
JCPDS Powder Diffraction File, International Centre for Diffraction Data
共1994兲.
12
T. H. Oh, D. L. Huffaker, L. A. Graham, H. Deng, and D. G. Deppe,
Electron. Lett. 32, 2024 共1996兲.
13
A. R. Sugg, E. I. Chen, N. Holonyak, Jr., K. C. Hsieh, J. E. Baker, and N.
Finnegan, J. Appl. Phys. 74, 3880 共1993兲.
14
K. D. Choquette, K. M. Geib, C. I. H. Ashby, R. D. Twesten, O. Blum, H.
Q. Hou, D. M. Follstaedt, B. E. Hammons, D. Mathes, and R. Hull, IEEE
J. Sel. Top. Quantum Electron. 3, 916 共1997兲.
15
C. I. H. Ashby, J. P. Sullivan, P. P. Newcomer, N. A. Missert, H. Q. Huo,
B. E. Hammons, M. J. Hafich, and A. G. Baca, Appl. Phys. Lett. 70, 2443
共1997兲.
Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions