Silicon epilayers grown by chemical vapor deposition from high-purity silane

Article abstract of DOI:10.1134/S0020168506080012

Structurally perfect, high-purity silicon epilayers up to 20 μm in thickness, with a carrier concentration n = 10¹² cm^-3 are grown through thermal decomposition of silane. Experimental evidence is presented that the concentration of electrically active impurities in high-purity silane can be evaluated from the electrical parameters of Si epilayers grown from it. Pleiades Publishing, Inc., 2006.

Full text of DOI:10.1134/S0020168506080012

ISSN 0020-1685, Inorganic Materials, 2006, Vol. 42, No. 8, pp. 815–818. © Pleiades Publishing, Inc., 2006.
Original Russian Text © A.V. Gusev, O.F. Udalov, V.S. Krasil’nikov, G.A. Maksimov, A.V. Kornaukhov, O.A. Kuznetsov, E.A. Uskova, V.B. Shmagin, 2006, published in Neorga-
nicheskie Materialy, 2006, Vol. 42, No. 8, pp. 903–907.
Silicon Epilayers Grown by Chemical Vapor Deposition
from High-Purity Silane
a
a
a
b
A. V. Gusev , O. F. Udalov , V. S. Krasil’nikov , G. A. Maksimov ,
b
b
b
c
A. V. Kornaukhov , O. A. Kuznetsov , E. A. Uskova , and V. B. Shmagin
a
Institute of Chemistry of High-Purity Substances, Russian Academy of Sciences,
ul. Tropinina 49, Nizhni Novgorod, 603950 Russia
b
Research Physicotechnical Institute, Lobachevski State University,
pr. Gagarina 23/3, Nizhni Novgorod, 603950 Russia
c
Institute for Physics of Microstructures, Russian Academy of Sciences,
GSP-105, Nizhni Novgorod, 603590 Russia
e-mail: gusev@ihps.nnov.ru
Received February 15, 2006
Abstract—Structurally perfect, high-purity silicon epilayers up to 20 µm in thickness, with a carrier concen-
tration n = 1012 cm–3 are grown through thermal decomposition of silane. Experimental evidence is presented
that the concentration of “electrically active” impurities in high-purity silane can be evaluated from the electri-
cal parameters of Si epilayers grown from it.
DOI: 10.1134/S0020168506080012
INTRODUCTION
EXPERIMENTAL
Silicon epilayers were grown via silane pyrolysis at
a pressure of 105 Pa in a water-cooled horizontal metal-
lic reactor fitted with a graphite heater (Fig. 1) [10]. As
substrates, we used n- and p-type silicon wafers 30 mm
in diameter, doped to various levels, which were
mounted on the heater in a zone with a temperature dif-
ference of 5–10°C. During the epitaxial process, the
substrate temperature was maintained at 1050°C, the
flow rate of silane was about 10 ml/min, and the growth
rate of the epilayers was about 0.3 µm. The epilayer
thickness was varied from 5 to 20 µm.
In a number of device applications, high-purity epil-
ayers of silicon, a basic semiconductor material, are
needed. In particular, the fabrication of microwave
diodes, photodetectors, quantum computers, Si–Ge
superlattices, and nuclear detectors requires high-purity
Si films with a carrier concentration below 1013 cm–3
and high structural perfection [1–7].
One attractive precursor for the growth of high-
purity silicon films is silane. As the impurity content of
silane decreases to below 0.01 ppma, this compound
becomes increasingly difficult to analyze by conven-
tional techniques, such as gas chromatography, IR
spectroscopy, and atomic absorption spectroscopy,
since even the most sensitive techniques have detection
limits for impurities at a level of 0.01 to 0.1 ppma [8].
Varied experimentation showed that the growth of
high-purity silicon epilayers with a carrier concentra-
tion at a level of 1013 cm–3 involved serious problems
due to contamination originating from the graphite
The purity of silane can be assessed with high sen- heater. Before being mounted in the reactor, the heater
sitivity by measuring the concentration of electrically
active impurities in silicon single crystals grown from
it [9]. This approach, however, requires a large amount
of silane. In connection with this, it is of interest to
develop procedures for evaluating the purity of silane
by determining electrical parameters (conductivity type
and carrier concentration) of single-crystal silicon epil-
ayers grown from it.
was annealed at 1600°C in vacuum (residual pressure
below 10–4 Pa) for 10 h. High-vacuum annealing, how-
ever, did not rule out background contamination, and
the concentration of uncompensated charge carriers
(holes) in the epilayers was as high as ~1016 cm–3. To
reduce the contaminating effect of the heater, it was
coated with a silicon layer ~5–10 µm thick, which was
deposited through decomposition of high-purity silane
before epilayer growth and was then heat-treated. Next,
the substrates were mounted on the heater, and the epi-
taxial process was run. With this procedure, the con-
tamination from the heater was insignificant. The
In this paper, we describe a process for the growth
of structurally perfect, high-purity silicon epilayers
through silane pyrolysis and compare their electrical
properties with those of silicon single crystals.
815

816
GUSEV et al.
files of the dopant. Because it has two measuring cells
Vacuum
differing in area (2.6 and 10 mm in diameter), the ECV
profiler enables determination of p- and n-type impu-
rity concentrations in a broad range, from ~1013 to
1019 cm–3. Throughout this concentration range, the
error of determination is within 15% for layer thick-
nesses no less than the width of the space charge region
in electrolyte/semiconductor barriers.
Substrates
The depth profiles of impurities in epilayers were
obtained by CV measurements at a fixed low voltage
(150–300 mV) in conjunction with successive electro-
chemical etching in the same cell. This profiling proce-
dure is faster than standard CV measurements on
metal/silicon barriers in combination with chemical
etching of Si steps. Moreover, owing to some inherent
features of the Si–electrolyte barrier potential, ECV
profiling can be performed on thinner epilayers in com-
parison with Schottky barriers [11]. Consequently,
smaller amounts of silane are needed to grow epilayers
for assessing its purity.
Heater
Water
Silane for epilayer growth was prepared through
trichlorosilane disproportionation on an ion-exchange
catalyst [13]. In our experiments, we used two different
batches of silane, prepared by simple distillation and by
fine purification via fractional distillation. The silane
was analyzed by gas chromatography as described by
Devyatykh and Krylov [14]. The analytical results indi-
cated that the impurity concentration in the silane sam-
ples purified by distillation was below the detection
limit (1 × 10–4 to 1 × 10–6 wt % for hydrocarbon impu-
rities and 5 × 10–4 to 1 × 10–5 wt % for Group III to V
hydrides).
The structural perfection of the epilayers and sub-
strates was assessed using double-crystal x-ray diffrac-
tion [15]. The structural perfection was quantified by
the full width at half maximum (FWHM) W of the rock-
ing curves measured on the surface of the epilayer and
SiH4
HCl
H2
Fig. 1. Schematic of the apparatus for the growth of silicon
epilayers.
grown epilayers were n-type, with a carrier concentra-
tion within 1012 cm–3.
The electrical properties of the epilayers were
assessed by determining the carrier concentration N
due to electrically active impurities using capacitance–
voltage (CV) measurements on an electrolyte/silicon
barrier (electrochemical CV (ECV) measurements)
[11] and, for comparison, on an Au/Si Schottky barrier.
In addition, Hall effect measurements in a van der Pauw
geometry were used as an independent technique. The
presence of impurities producing deep centers in the
band gap of silicon was ascertained by deep-level tran-
sient spectroscopy (DLTS) measurements [12] on a
Schottky barrier.
the backside of the substrate using CuKα radiation and
1
the (001) crystallographic plane of the crystal mono-
chromator and the crystal being studied. As the crystal
monochromator, we used a structurally perfect germa-
nium single crystal set to the symmetric 004 reflection.
The rocking curves of the samples were measured
using the symmetric 004 reflection. The x-ray penetra-
tion depth for this reflection was 1.5 µm, so that, in
measurements on the epilayer surface, we recorded the
beam reflected only from the epilayer since its thick-
ness notably exceeded the x-ray penetration depth. The
uncertainty in W was 0.5″, and that in the angle between
the substrate surface and (001) plane (σ) was 0.05°. Our
results on the structural perfection of the epilayers are
summarized in the table.
ECV measurements provide more accurate carrier
concentration depth profiles in comparison with stan-
dard CV measurements [11] and offer the possibility of
measuring N in epilayers that have intricate depth pro-
Parameters of silicon epilayers
W, seconds of arc
Thickness,
σ, deg
µm
substrate
film
RESULTS AND DISCUSSION
4.35
4.30
4.90
1.60
1.60
1.50
13.5
13.5
13.7
13.9
13.7
14.1
Determining the orientation of the substrate surface
by rotating the sample about the normal to its surface
[15–17], we found that the angle between the substrate
INORGANIC MATERIALS Vol. 42 No. 8 2006

SILICON EPILAYERS GROWN BY CHEMICAL VAPOR DEPOSITION
817
N, cm–3
1020
crystals (7.3″) [18]. It can be seen from the table that, in
all of the samples, the W of the epilayer (0.2″–0.4″)
exceeds that of the substrate only slightly.
All of the Si epilayers were n-type, independent of
the silane used. The carrier concentration in the epilay-
ers depended on the purity of the silane and varied from
8 × 1014 cm–3 in the case of the silane purified by simple
distillation to ≤1 × 1013 cm–3 in the case of the silane
purified with the use of fractional distillation.
1018
1016
1014
1
3
2
Figure 2 shows the impurity profiles in four silicon
epilayers grown using the two batches of silane—after
the first (simple distillation) (curves 1, 2) and second
(fractional distillation) (curves 3, 4) purification cycles.
The carrier concentration is 4 × 1014 cm–3 in the case of
the silane purified by simple distillation and 8 ×
1012 cm–3 after fractional distillation. The results in
Fig. 2 were obtained by measurements on electro-
lyte/silicon contacts (curves 1, 3) and Au/Si barriers
(curves 2, 4). The samples differed in thickness. Com-
parison of curves 3 and 4 clearly demonstrates one
advantage of measurements in an electrolyte for the
epilayers grown from the high-purity silane, purified by
fractional distillation: for measurements on the Schot-
tky barrier, the layer thickness must exceed 10 µm
since, at zero bias, the space charge region in the Schot-
tky barrier is ~10 µm in thickness. It is also seen in
Fig. 2 that the carrier concentration is constant through-
out the thickness of the epilayers, which was the case in
most of our samples.
4
1012
0
4
8
12
16
x, µm
Fig. 2. Impurity profiles in silicon epilayers grown using
two batches of silane: (1, 2) first purification (simple distil-
lation), (3, 4) second purification (fractional distillation);
measurements on (1, 3) electrolyte/silicon and (2, 4) Au/Si
contacts.
surface and the (001) crystallographic plane was σ =
1.5°–1.6°. This small misorientation of the substrate
surface had no significant effect on the width of the dif-
fraction peaks from the substrate and epilayer. The
grown silicon epilayers were single-crystalline, and
their structural perfection (W ~ 14″) was close to that of
the substrates (table).
In addition to layers on heavily doped, n+ substrates
(KEM-0.05 silicon), we grew epilayers on high-resis-
tivity p-type substrates (KDB-10 silicon) for CV and
The rocking curves of the silicon substrates used in
epilayer growth had FWHM = 13.5″, which exceeds the
intrinsic broadening predicted theoretically for single Hall measurements. Hall measurements are, however,
0.23 eV
5 × 1013 cm–3
0.20
0.15
0.10
0.05
25
20
15
10
2
1
0.28 eV
0.46 eV
2 × 1012 cm–3
2 × 1012 cm–3
0
0
5
50
100
150
200
250
300
T, K
Fig. 3. (1) DLTS spectrum and (2) temperature-dependent capacitance of the Schottky contact for silicon epilayers grown from the
first silane batch; the numbers at the peaks specify the concentrations and ionization energies of the corresponding levels; the carrier
14
–3
concentration in the epilayers at T = 300 K, V
= 4 V, and f = 400 kHz is n = 4 × 10 cm .
sample
INORGANIC MATERIALS Vol. 42 No. 8 2006

818
GUSEV et al.
ments and Prospects, XII Vsesoyuznaya nauchno-tekh-
nicheskaya konferentsiya po tverdotel’noi elektronike
SVCh (XII All-Union Sci. and Technol. Conf. on Micro-
wave Solid-State Electronics), Kiev, 1990, pp. 286–287.
inapplicable in the case of epilayers with carrier con-
centrations of ~1013 cm–3 and below. For comparison
with the electrical properties of the films, the purity of
the silane purified by fractional distillation was deter-
mined by a well-known method, from the parameters of
a reference silicon single crystal [9]. Hall measure-
ments were performed on plates cut from a reference
single crystal grown using the same batch of purified
silane as the epilayers for CV measurements. The mea-
surement results for the epilayers and reference single
crystal were in good agreement. The carrier concentra-
tion determined from the Hall data for the reference sin-
gle crystal was 9 × 1012 cm–3.
3. Greene, J.E., Gressmann, H.-J., and Schubert, E.F., Delta
Doping in Silicon, Solid State Mater. Sci., 2005,
vol. 152, no. 6, pp. G448–G451.
4. Ager, J.W., Beeman, J.W., Hansen, W.L., et al., High-
Purity, Isotopically Enriched Bulk Silicon, J. Electro-
chem. Soc., 1993, vol. 18, no. 1, pp. 1–67.
5. Kizilyalli, I.C., Safar, H., Herbsommer, J., et al., Power
Transistor Fabricated Using Isotopically Purified Silicon
(28Si), IEEE Electron. Device Lett., 2005, vol. 26, no. 6,
pp. 404–406.
Figure 3 shows the DLTS spectrum and the temper-
ature-dependent capacitance of the Au/Si contact for
the epilayers grown from the silane purified by simple
distillation. The impurity profile in these epilayers is
represented by curve 2 in Fig. 2. Analysis of the DLTS
spectrum indicates that the carrier concentration in the
epilayers is determined primarily by electrically active
impurities with shallow levels in the band gap of sili-
con. The concentration of deep levels with ionization
energies above 200 meV is 2 × 1012 to 5 × 1013 cm–3 (see
Fig. 3), which is one order of magnitude lower than the
concentration of shallow impurities. The ionization
energies of the deep levels suggest that these originate
from impurity–defect complexes forming during the
epitaxial process [9].
6. Ladd, T.D., Goldmann, J.R., Yamaguchi, F., and Yama-
moto, Y., All-Silicon Quantum Computer, Phys. Rev.
Lett., 2002, vol. 89, no. 1, p. 017901 (1–4).
7. Sabinin, V.E., Nuzhin, V.S., and Solks, S.V., Application
of Silicon and Germanium in Optical Devices: Require-
ments on Quality and Characterization Techniques,
Soveshchanie “Kremnii-2004” (Silicon-2004 Conf.,
Irkutsk, 2004), Irkutsk: Inst. Geografii Sib. Otd. Ross.
Akad. Nauk, 2004, p. 240.
8. Taylor, P.A., Purification Techniques and Analytical
Methods for Gaseous and Metallic Impurities in High-
Purity Silane, J. Cryst. Growth, 1988, vol. 89, no. 1,
pp. 28–38.
9. Salli, I.V. and Fal’kevich, E.S., Proizvodstvo polupro-
vodnikovogo kremniya (Production of Semiconductor-
Grade Silicon), Moscow: Metallurgiya, 1970.
In the epilayers grown from the silane purified by
fractional distillation, the concentration of electrically
active impurities (Fig. 2, curves 3, 4) was almost two
orders of magnitude lower, and the concentration of
deep impurities was below 1011 cm–3. The good agree-
ment between the electrical parameters of Si epilayers
and reference single crystals suggests that the tech-
niques we used are sufficiently accurate at impurity
concentrations in epilayers on the order of 1013 cm–3.
Therefore, the purity of silane can be evaluated from
the electrical parameters of Si epilayers grown from it.
10. Zeveke, T.A., Tolomasov, V.A., Kornev, L.N., et al.,
Direct-Resistance-Heating Metallic Reactor for the
Growth of Germanium Layers by the Hydride Tech-
nique, Elektron. Tekh., Ser. 6: Mater., 1973, no. 6,
pp. 116–118.
11. Blood, P., Capacitance–Voltage Profiling and the Char-
acterization of III–V Semiconductors Using Electrolyte
Barriers, Semicond. Sci. Technol., 1986, vol. 1, pp. 7–27.
12. Berman, L.S., Emkostnye metody issledovaniya polupro-
vodnikov (Capacitance Techniques for Characterization
of Semiconductors), Moscow: Nauka, 1972, ch. 3.
13. Grishnova, N.D., Gusev, A.V., Moiseev, A.V., et al., Cat-
alytic Activity of Anion-Exchange Resins for Trichlo-
rosilane Disproportionation, Zh. Prikl. Khim. (S.-Peter-
burg), 1999, vol. 72, no. 10, pp. 1667–1673.
14. Devyatykh, G.G. and Krylov, V.A., Analysis of High-
Purity Volatile Inorganic Substances by Gas Chromatog-
raphy, Vysokochist. Veshchestva, 1987, no. 3, pp. 35–48.
15. Krasil’nikov, V.S., Bublik, V.T., Lebedev, S.N., et al.,
Defect Structure of Epitaxial Films by X-ray Diffraction
Profiling, Elektron. Tekh., Ser. 8, 1985, no. 2 (113),
pp. 26–31.
CONCLUSIONS
High-purity silicon epilayers were grown through
silane pyrolysis. The layers offer high structural perfec-
tion (W ~ 14″) and carrier concentrations within
1013 cm–3.
The present results demonstrate that the purity of
silane can be evaluated from the electrical parameters
of Si epilayers grown from it.
16. Krasil’nikov, V.S., USSR Inventor’s Certificate
no. 890176, Byull. Izobret., 1981, no. 46, p. 219.
REFERENCES
17. Andreev, A.Yu., Gudkova, N.V., Kuznetsov, V.P., et al.,
Phosphorus Doping of Si Layers during Molecular
Beam Epitaxial Growth on (001) Si, Izv. Akad. Nauk
SSSR, Neorg. Mater., 1988, vol. 24, no. 9, pp. 1423–1426.
18. Tanner, B.K., X-ray Topography and Precision Diffrac-
tometry of Semiconducting Materials, J. Electrochem.
Soc., 1979, vol. 136, no. 11, pp. 3438–3443.
1. Aleshkin, V.Ya. and Kozlov, D.V., Blocked Impurity
Conductivity GeSi Photodetector, Materialy soveshcha-
niya “Nanosfera” (Proc. Nanosphere Conf.), Nizhni
Novgorod, 2001, pp. 217–219.
2. Grimal’skii, V.V., Kishchenko,Ya.I., and Koshevaya, S.V.,
Si and GaAs Surface-Oriented PIN Devices: Develop-
INORGANIC MATERIALS Vol. 42 No. 8 2006