In situ passivation of InP surface using H2 S during metal organic vapor phase epitaxy

Article abstract of DOI:10.1063/1.3233935

An in situ surface passivation of InP(100) using H² S during metal organic vapor phase epitaxy has been characterized by x-ray photoemission spectroscopy and photoluminescence. X-ray photoelectron spectra indicate that the H² S -trea

Full text of DOI:10.1063/1.3233935

In situ passivation of InP surface using H 2 S during metal organic vapor phase
epitaxy
Hong-Liang Lu, Yuki Terada, Yukihiro Shimogaki, Yoshiaki Nakano, and Masakazu Sugiyama
Citation: Applied Physics Letters 95, 152103 (2009); doi: 10.1063/1.3233935
View online: http://dx.doi.org/10.1063/1.3233935
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/95/15?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Strong surface passivation of GaAs nanowires with ultrathin InP and GaP capping layers
Appl. Phys. Lett. 105, 033114 (2014); 10.1063/1.4891535
In situ X-ray investigation of changing barrier growth temperatures on InGaN single quantum wells in metal-
organic vapor phase epitaxy
J. Appl. Phys. 115, 094906 (2014); 10.1063/1.4867640
Heteroepitaxy of InP on Si(001) by selective-area metal organic vapor-phase epitaxy in sub-50nm width
trenches: The role of the nucleation layer and the recess engineering
J. Appl. Phys. 115, 023710 (2014); 10.1063/1.4862044
Small valence-band offset of In 0.17 Al 0.83 N / GaN heterostructure grown by metal-organic vapor phase
epitaxy
Appl. Phys. Lett. 96, 132104 (2010); 10.1063/1.3368689
High-resolution microbeam x-ray diffractometry applied to InGaAsP/InP layers grown by narrow-stripe
selective metal-organic vapor phase epitaxy
Appl. Phys. Lett. 77, 1286 (2000); 10.1063/1.1290048
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
140.254.87.149 On: Fri, 19 Dec 2014 01:22:25

APPLIED PHYSICS LETTERS 95, 152103 ͑2009͒
In situ passivation of InP surface using H S during metal organic vapor
2
phase epitaxy
1
,a͒
1
2
1,3
Hong-Liang Lu,
Masakazu Sugiyama
Yuki Terada, Yukihiro Shimogaki, Yoshiaki Nakano, and
1,4
1
Department of Electronic Engineering and Information Systems, School of Engineering,
The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
2
Department of Materials Engineering, School of Engineering, The University of Tokyo, 7-3-1 Hongo,
Bunkyo-ku, Tokyo 113-8656, Japan
3
Research Center for Advanced Science and Technology, The University of Tokyo, 4-6-1 Komaba,
Meguroku, Tokyo 153-8904, Japan
4
Institute of Engineering Innovation, School of Engineering, The University of Tokyo, 7-3-1 Hongo,
Bunkyo-ku, Tokyo 113-8656, Japan
͑
Received 27 July 2009; accepted 31 August 2009; published online 13 October 2009͒
An in situ surface passivation of InP͑100͒ using H S during metal organic vapor phase epitaxy
2
has been characterized by x-ray photoemission spectroscopy and photoluminescence. X-ray
photoelectron spectra indicate that the H S-treated InP at 300 °C is free of P and In oxides even
2
after exposure to air. The enhancement of photoluminescence intensity confirms that H2S
passivation of an InP epilayer can reduce the surface defects. It is shown that H S treatment results
2
in In–S bonds, which dominate the sulfur-passivated InP surface, effectively suppressing interface
oxidation during the subsequent ultrathin Al O dielectric film growth. © 2009 American Institute
2
3
of Physics. ͓doi:10.1063/1.3233935͔
Recently, there has been a renewed research interest in
III-V semiconductor materials as alternative channel material
other than Si in metal oxide semiconductor field effect tran-
sistors ͑MOSFETs͒, and their applications at the 22 nm fea-
ment methods have been proposed to improve InP surface
properties before dielectric deposition. One method employs
aqueous sulfur solutions such as S Cl and ͑NH ͒ S solu-
tions to passivate the substrate surface with a sulfur layer.
It was reported that applying sulfur passivation using
2
2
4 2
1
3
1
–5
ture size and beyond due to intrinsic transport properties.
However, preparations of well-defined III-V surfaces and
high-quality oxide-semiconductor interfaces with low inter-
face state density remain challenging tasks. Some encourag-
ing results have been reported in the refinement of III-V
MOSFETs technology including in situ deposition of Ga O
͑NH ͒ S and postdeposition annealing improve the drive cur-
4
2
1
4
rent and subthreshold swing. While there have been re-
markable improvements in surface recombination velocity
and device performance, some problems still exist such as
poor reproducibility and aging effects. Another method em-
2
3
6
7
͑
Gd O ͒, molecular beam epitaxy of Gd O , and atomic
ploying H S in an ultrahigh vacuum chamber has been dem-
2
3
2
3
2
layer deposition of Al O ͑Refs. 1 and 8͒ and HfO ͑Ref. 3͒
onstrated to produce higher quality device surfaces, than
2
3
2
1
5
thin films on GaAs or InGaAs. Similarly, InP inversion-type
MOSFETs with atomic layer deposition Al O have demon-
those treated with the wet process. Passivation of the elec-
2
3
tronic defect states at the SiN /InP interface has been
x
11
strated the capability of high drive current density, and can
provide a much smaller off-current density compared to
achieved by Kaplia et al. using gaseous H S treatments of
2
the InP surface. Despite considerable experimental and the-
oretical work, the atomic structures of sulfur treated III-V
surfaces are not established and the passivation mechanism
is still unclear.
9
InGaAs. In addition, the insulated gate structure of an InP
MOSFET allows the larger gate-voltage swing in both po-
1
0
larities needed in high power devices. However, InP com-
pounds do not have native oxides suitable for use as insula-
tors in MOSFETs. Thus, an externally applied insulator, and
a method for applying it is required. Compounds of SiO2,
Al O , and Si N have been previously employed to produce
In this work, an in situ passivation of InP͑100͒ surface
with H S during metal organic vapor phase epitaxy
2
͑MOVPE͒ has been chosen to prevent the formation of na-
tive oxides and suppress the interfacial oxide growth during
the dielectric Al O growth. The chemical and structural
2
3
3
4
11
insulated gate layers on InP. More recently, InP-based MOS
2
3
structures with HfO grown by atomic layer deposition have
properties of in situ H S treated InP͑100͒ surfaces, and sur-
2
2
1
2
also been prepared by Kim et al.
faces capped with an ultrathin Al O layer are reported.
2
3
Prior to insulator deposition, the InP surface should be
pretreated and well passivated, as InP rapidly oxidizes in air
to form a native oxide around 1–2 nm thick. In the absence
of special precautions to eliminate native oxide formation,
prior to deposition of a dielectric layer, it is likely that the
electrical properties of the interface will be greatly affected
by the native surface oxide. As a result, many surface treat-
The InP epitaxial layers were grown on a S-doped
n-InP͑100͒ wafer using a horizontal MOVPE reactor ͑AIX-
TRON, AIX200/4͒. During the epitaxial growth process,
trimethy-indium and tertiarybutyl-phosphine were selected
as the sources of In and P, respectively. All specimens were
prepared at 600 °C using a hydrogen carrier flow with a
growth pressure of 10 kPa. After the growth of InP epilayers,
sulfur passivation was performed using hydrogen sulfide
a͒Author to whom correspondence should be addressed. Electronic mail:
hllu@hotaka.t.u-tokyo.ac.jp.
͑H S͒ in the temperature range of 200–350 °C for 30 min in
2
the MOVPE reactor. After surface treatment, a 3 nm Al O
2
3
0
003-6951/2009/95͑15͒/152103/3/$25.00
95, 152103-1
40.254.87.149 On: Fri, 19 Dec 2014 01:22:25
© 2009 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
1

152103-2
Lu et al.
Appl. Phys. Lett. 95, 152103 ͑2009͒
FIG. 1. ͑Color online͒ XPS data of ͑a͒ P 2p and ͑b͒ In 3d spectra for
as-received, epitaxial, H S-treated InP surfaces.
2
film as measured by spectroscopic ellipsometry was depos-
ited by metal-organic chemical vapor deposition at 300 °C
using dimethylaluminumhydride and O as precursors.
2
FIG. 2. ͑Color online͒ O 1s core level spectra for as-received, epitaxial,
2
as-received and epitaxial InP surfaces.
The specimens were then transfered into an ultrahigh
vacuum x-ray photoemission spectroscopy ͑XPS͒ chamber in
less than 15 min. The XPS measurements were performed in
an ULVAC-PHI Model 1600C system equipped with a non-
monochromatic Mg K␣ ͑1253.6 eV͒ line as an x-ray source.
H S-treated InP surfaces. Inset shows the deconvolution of O 1s signal for
3d spectra, it is concluded that the H2S treated InP surface is
terminated by In–S bonds, which can protect the InP surface
from oxidation during exposure to air.
−
7
All experiments were conducted at 2.7ϫ10 Pa. Photoelec-
trons were taken from the surface with 45° angle, and the C
1
s peak at 284.6 eV was taken as a reference for charge
Figure 2 shows the O 1s core level spectra for the as-
correction. Photoluminescence ͑PL͒ data was collected at
room temperature under excitation by an argon ion laser
beam of 514.5 nm.
received, epitaxial, and H S treated samples. The O 1s peaks
2
for as-received and epitaxial samples ͑inset to Fig. 2͒ could
be split into two components, O–P ͑531.7 eV͒ and O–In
͑531.1 eV͒, respectively. It can be observed that the intensity
of O–P signal for epilayer decreases compared to that for
as-received sample. These results are in agreement with the P
The chemical binding states of InP samples prior to the
Al O deposition were first investigated by XPS. P 2p and In
2
3
3
d core levels for the samples with and without H S treat-
2
ments are shown in Fig. 1. The P 2p core level spectrum
attributed to the P–In bond from the substrate is composed of
2
p spectra deconvolution. Moreover, the intensity of O 1s
peak for H S-treated samples is much lower than that for
2
two peaks, P 2p1 ͑129.3 eV͒ and P 2p3/2 ͑128.5 eV͒, as
/2
as-received and epitaxial samples. The minor oxygen in the
sulfide passivated samples is believed from the adsorbed sur-
face CO ͑531.0 eV͒. Based on the above In 3d spectra
shown in Fig. 1͑a͒. A component at higher binding energy
with a peak at 133.1 eV which correlates with a P–O bond
was observed for the as-received sample, conducive with the
2
5/2
information, it suggests that the sulfur can bridge between
the surface In atoms through occupying the P-related vacan-
cies. As a result, the surface would be passivated as there
would be practically no dangling bonds on the sulfur-
passivated InP surface.
1
6
presence of P oxide on the surface of the sample. A small
amount of P oxide also exists in the InP epilayer surface.
However, the peak with a binding energy in the vicinity of
1
33 eV was not observed in the XPS spectra of H S-treated
2
InP surface, suggesting that neither P–O nor P–S bond
A qualitative indication of passivation can be achieved
by comparing the PL intensity, which is related to the elec-
͑
ϳ132.6 eV͒ ͑Ref. 17͒ exists on the H S treated InP surface.
2
When considering the In 3d spectra, only the In 3d5/2
19
trical properties of MOS capacitor. The PL spectra of InP
peaks will be addressed for simplicity, which are presented in
Fig. 1͑b͒. The In 3d5/2 features can be deconvoluted into
three binding energy regions, which are assigned to In–O,
In–S, and In–P species, respectively. The peak arising at
samples before and after H S treatments are shown in Fig. 3.
2
The spectra show a broadband with an intense peak at 1.37
20
eV due to transitions from band to band. This peak is
1
8
slightly above the intrinsic bandgap energy ͑Ego=1.35 eV͒
of n-InP. As can be seen, a clear enhancement of the PL
4
44.3 eV is due to In–P bonding. Although the distinction
between O and S bonding to In is difficult to establish from
XPS spectra information alone, there is a measurable differ-
ence which suggests a distinction. The binding energy of
In–O peak is slightly higher than that for the In–S peak with
peaks at 444.9 and 445.1 eV originating from In–S and In–O
bonds, respectively. It also can be seen that the In–S compo-
intensity is observed for the epitaxial and H S-treated
2
samples. It is reported that the PL intensity varies with the
1
6
amount of oxide formed on InP surface and surface defects.
The increase in PL intensity for the epitaxial InP sample
compared to the as-received one suggests that only a small
amount of native oxide was present on its surface. The fur-
nent increases with increased H S treatment temperature
2
ther increase in PL intensity for the H S treated samples was
from 200 to 300 °C while the In-O component decreases.
2
The In oxides presence for the H S-treated sample at 300 °C
due to not only the decrease in native oxides but also the
reduction in surface recombination velocity caused by the
reduction in surface state density. As a result, the surface
defects acting as interface traps were nearly passivated after
2
was negligible. However, it shall be noted that the amount of
In-S decreases for the sample treated at 350 °C. This phe-
nomenon may be due to partial desorption of S from the
1
6
sample surface. Based on the information from P 2p and In
an in situ H S treatment.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the term 2s at: http://scitation.aip.org/termsconditions. Downloaded to IP:
40.254.87.149 On: Fri, 19 Dec 2014 01:22:25
1

152103-3
Lu et al.
Appl. Phys. Lett. 95, 152103 ͑2009͒
tected in sulfur-passivated samples at 300 °C, indicating that
H S passivates the InP surface and, therefore, inhibiting the
2
interfacial oxide regrowth during Al O deposition. It is also
2
3
noted that a certain amount of indium oxide exists at the top
of InP epitaxial surface, as suggested by the slight shift of
peak position to higher energy ͑ϳ0.2 eV͒. There was no
increase in indium oxides ͓Fig. 4͑b͔͒ after thin Al O films
2
3
grown on sulfur-passivated InP sample, indicating the good
stability of the sulfur-passivated InP sample.
In summary, XPS and PL have been employed to char-
acterize the chemical and structural properties of in situ
S-passivated InP surface using H S. The P and In oxides
2
were not observed for the H S-treated InP epitaxial sample at
2
3
00 °C. The enhancement of photoluminescence intensity
after H S treatment was attributed to the decrease in native
2
oxides and surface defects. The regrowth of In or P oxides on
the sulfur-passivated InP sample during the subsequent
Al O deposition was effectively suppressed. It is concluded
FIG. 3. ͑Color online͒ PL spectra of as-received, epitaxial, H S-treated
2
InP͑001͒ samples. The dashed line identifies the intrinsic bandgap energy
͑
Ego͒ of n-InP.
2
3
that the H S treatment at 300 °C was effective in preventing
2
the formation of native oxides and inhibiting interfacial ox-
ide regrowth.
Figure 4 shows the XPS data of InP͑001͒ surfaces with
and without H S treatment followed by 3 nm deposition of
2
Al O . It can be seen clearly from Fig. 4͑a͒ that the presence
2
3
The authors would acknowledge to JSPS program and
Mr. Michael Breedon for valuable discussion.
of strong P–O bonds for the epitaxial sample. On the con-
trary, for the sulfur-passivated sample at 200 °C, the P–O
peak intensity significantly reduced. P–O bonds were not de-
1
M. Passlack, M. Hong, and J. P. Mannaerts, Appl. Phys. Lett. 68, 1099
2͑
1996͒.
P. D. Ye, J. Vac. Sci. Technol. A 26, 697 ͑2008͒.
3
F. S. Aguirre-Tostado, M. Milojevic, K. J. Choi, H. C. Kim, C. L. Hinkle,
E. M. Vogel, J. Kim, T. Yang, Y. Xuan, P. D. Ye, and R. M. Wallace, Appl.
Phys. Lett. 93, 061907 ͑2008͒.
4
5
6
7
8
9
0
A. Molle, S. Spiga, A. Andreozzi, M. Fanciulli, G. Brammertz, and M.
Meuris, Appl. Phys. Lett. 93, 133504 ͑2008͒.
H. Zhao, D. Shahrjerdi, F. Zhu, H.-S. Kim, I. Ok, M. Zhang, J. H. Yum, S.
K. Banerjee, and J. C. Lee, Appl. Phys. Lett. 92, 253506 ͑2008͒.
J. F. Zheng, W. Tsai, T. D. Lin, Y. J. Lee, C. P. Chen, M. Hong, J. Kwo, S.
Cui, and T. P. Ma, Appl. Phys. Lett. 91, 223502 ͑2007͒.
C. C. Cheng, C. H. Chien, G. L. Luo, C. K. Tseng, H. C. Chiang, C. H.
Yang, and C. Y. Chang, J. Electrochem. Soc. 155, G56 ͑2008͒.
H. L. Lu, L. Sun, S. J. Ding, M. Xu, D. W. Zhang, and L. K. Wang, Appl.
Phys. Lett. 89, 152910 ͑2006͒.
Y. Q. Wu, Y. Xuan, T. Shen, P. D. Ye, Z. Cheng, and A. Lochtefeld, Appl.
Phys. Lett. 91, 022108 ͑2007͒.
H. Zhao, D. Shahrjerdi, F. Zhu, H.-S. Kim, I. Ok, M. Zhang, J. H. Yum, S.
K. Banerjee, and J. C. Lee, Electrochem. Solid-State Lett. 11, H233
1
1͑
2008͒.
1
A. Kapila, X. Si, and V. Malhotra, Appl. Phys. Lett. 62, 2259 ͑1993͒.
H.-S. Kim, I. Ok, M. Zhang, F. Zhu, S. Park, J. Yum, H. Zhao, J. C. Lee,
and P. Majhi, Appl. Phys. Lett. 93, 102906 ͑2008͒.
1
1
1
2
3
4
L. J. Gao, G. W. Anderson, F. Esposto, P. R. Norton, B. F. Mason, Z.-H.
Lu, and M. J. Graham, J. Vac. Sci. Technol. B 13, 2053 ͑1995͒.
H. Zhao, D. Shahrjerdi, F. Zhu, M. Zhang, H. -S. Kim, I. Ok, J. H. Yum,
S. Park, S. K. Banerjee, and J. C. Lee, Appl. Phys. Lett. 92, 233508
5͑
2008͒.
1
1
M. Shimomura, K. Naka, N. Sanada, Y. Suzuki, Y. Fukuda, and P.
J. Moller, J. Appl. Phys. 79, 4193 ͑1996͒.
I. K. Han, E. K. Kim, J. I. Lee, S. H. Kim, K. N. Kang, Y. Kim, H. Lim,
and H. L. Park, J. Appl. Phys. 81, 6986 ͑1997͒.
6
1
1
7
8
F. Maeda, Y. Watanabe, and M. Oshima, Appl. Phys. Lett. 62, 297 ͑1993͒.
G. Chen, S. B. Visbeck, D. C. Law, and R. F. Hicks, J. Appl. Phys. 91,
9
362 ͑2002͒.
1
2
9
W. D. Chen, X. Q. Li, L. H. Duan, X. L. Xie, and Y. D. Cui, Appl. Surf.
Sci. 100, 592 ͑1996͒.
R. M. Sieg and S. A. Ringel, J. Appl. Phys. 80, 448 ͑1996͒.
FIG. 4. ͑Color online͒ XPS data of ͑a͒ P 2p and ͑b͒ In 3d spectra for
0
as-received, epitaxial, H S-treated InP surfaces after 3 nm Al O deposition.
2
2
3
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
40.254.87.149 On: Fri, 19 Dec 2014 01:22:25
1