METALORGANIC CHEMICAL VAPOR DEPOSITION OF ANTIMONY OXIDE FILMS
577
Triethylantimony was synthesized in a similar way, annealing on the film microstructure and properties, the
by reacting SbCl with ethylmagnesium bromide in films deposited in an argon atmosphere were then
3
flowing inert gas.
Antimony butoxide was prepared by reacting SbCl3
with butyl alcohol in the presence of ammonia:
annealed for 30 min at 300, 350, and 400°C in oxygen,
a mixture of oxygen and water vapor, or argon.
Film thicknesses were measured with an E-3 ellip-
someter to an accuracy of ±1 nm. Film composition was
determined by IR spectroscopy (UR-20 instrument),
electron diffraction (EG-100A unit), and electron probe
x-ray microanalysis (Camscan SI scanning electron
microscope equipped with a Link AN10/55 energy-dis-
persive x-ray analysis system). The surface morphol-
ogy of the films was examined using an MII-4 interfer-
ence microscope and ARM-1000 electron microscope.
SbCl + 3C H OH + 3NH = Sb(OC H ) + 3NH Cl.
3
4
9
3
4
9 3
4
To this end, SbCl (228 g) and absolute butyl alcohol
3
(
550 ml) were mixed in a three-neck flask fitted with a
stirrer, condenser, and an inlet tube for ammonia. After
cooling to 0°C, ammonia was bubbled through the mix-
ture. The precipitating ammonium chloride was then
collected on a filter. The filtrate was vacuum-distilled to
separate antimony butoxide.
Antimony tri-β-aminoethoxide was prepared via
transesterification of antimony butoxide with ethyl Cel-
losolve and ethanolamine:
RESULTS AND DISCUSSION
Our results on the kinetics of film deposition via
thermal decomposition of triethylantimony are summa-
rized in Table 1 and Fig. 1. The film thickness was
found to vary linearly with deposition time. Prelimi-
nary experiments showed that the deposition rate
(
n-C H O) Sb + NH C H OH
4 9 3 2 2 4
=
(NH C H O) Sb + n-C H OH.
2 2 4 3 4 9
Antimony butoxide (5 g) and a small excess of ethano- depended strongly on whether the solution was used as-
lamine were introduced into a vacuum-distillation prepared or after storage. The nature of the substrate
flask. After distilling off the butyl alcohol, the resultant had no significant effect on the deposition rate, which is,
antimony tri-β-aminoethoxide was vacuum distilled.
in addition, almost temperature-independent (Table 1).
This may be due to the fact that the MO precursor fully
decomposes even at 130°C, and the deposition rate is
then controlled by the flow rate of the carrier gas [1]. In
the case of the pyrolysis of a triethylantimony solution
in toluene, the growth rate even decreases with increas-
ing temperature. Increasing the triethylantimony con-
centration in butyl alcohol slightly increases the depo-
sition rate (Fig. 1). The rate constants for the pyrolysis
of solutions with Sb(C H ) : C H OH = 1 : 1 and 2 : 1
The substrates used in kinetic studies were polished
wafers of AGP-1 and AGChTs GaAs(100) with a car-
1
6
–3
rier concentration of 10 cm and KEF Si with an elec-
trical resistivity of 7.5 Ω cm. The process was run in a
resistance-heated two-zone furnace. The temperature
was maintained with a stability of ±1°C. The source and
deposition temperatures were chosen using thermo-
gravimetric data (Paulik–Paulik–Erdey system).
2
5 3
4
9
We used single- and two-zone procedures. In the
former case, the MO precursor (antimony butoxide,
tripropylantimony, or antimony tri-β-aminoethoxide)
was placed in a quartz container, which was then cov-
ered with the substrate. In the two-zone procedure, the
container with the precursor (triethylantimony) was sit-
uated in the lower temperature zone, where the precur-
sor vaporized without decomposition, and the substrate
was located in the higher temperature zone, where the
precursor decomposed.
differ insignificantly: 6 ± 0.4 and 7 ± 0.8 nm/min.
Interestingly enough, the refractive index of the
films is essentially independent of the nature of the sol-
vent and substrate and increases with deposition tem-
perature, pointing to changes in film composition
(Table 1). The refractive index varies from 1.67 to 2.06,
3
00
250
00
150
Given that antimony alkyl compounds are unstable
in air and extremely reactive, we used their solutions in
butyl alcohol or toluene. In the case of triethylanti-
4
3
2
mony, we employed solutions with Sb(C H )
:
2
5 3
C H OH = 1 : 1 or 2 : 1 and Sb(C H ) : C H CH = 1 : 1
2
1
4
9
2
5 3
6
5
3
1
00
50
by volume. The solution temperature was 90°C, and the
substrate (Si or GaAs) temperature was 130, 190, or
3
00°C. This range of deposition temperatures was cho-
sen because, below 130°C, no film growth occurred,
while, above 300°C, the deposition rate was too high,
resulting in poor bonding between the film and the sub-
strate. The flow rate of argon carrier gas was 18 l/h in
all deposition runs.
0
5
10
15
20
25
30
35
40
Deposition time, min
Fig. 1. Film thickness as a function of deposition time for
films deposited on GaAs via triethylantimony pyrolysis at
(
1, 3, 4) 190 and (2) 130°C and an Sb(C H ) : C H OH
2 5 3 4 9
The etching behavior of the films was studied using
.2% hydrochloric acid. To study the effect of thermal
ratio of (1–3) 1 : 1 and (4) 2 : 1; (1, 2) 30 min after solution
preparation, (3, 4) freshly prepared solution.
7
INORGANIC MATERIALS Vol. 38 No. 6 2002