Mechanically activated synthesis of monosilane by the reaction of calcium hydride with silicon tetrafluoride

Article abstract of DOI:10.1134/S1070427210060108

Possibility of performing a mechanically activated synthesis of pure monosilane by the reaction of calcium hydride with silicon tetrafluoride in a rotating flow-through reactor containing milling bodies was analyzed. A number of operation parameters of the grinding reactor were theoretically evaluated and experimentally determined.

Full text of DOI:10.1134/S1070427210060108

ISSN 1070-4272, Russian Journal of Applied Chemistry, 2010, Vol. 83, No. 6, pp. 984–988. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © O.Yu. Troshin, A.D. Bulanov, V.S. Mikheev, A.Yu. Lashkov, 2010, published in Zhurnal Prikladnoi Khimii, 2010, Vol. 83, No. 6,
pp. 933–937.
INORGANIC SYNTHESIS AND INDUSTRIAL
INORGANIC CHEMISTRY
Mechanically Activated Synthesis of Monosilane
by the Reaction of Calcium Hydride with Silicon Tetrafluoride
O. Yu. Troshina, A. D. Bulanovb, V. S. Mikheevb, and A. Yu. Lashkovb
a Lobachevsky State University, State Educational Enterprise for Higher Professional Education, Nizhni Novgorod, Russia
b Institute for Chemistry of Ultrapure Substances, Russian Academy of Sciences, Nizhni Novgorod, Russia
Received September 30, 2009
Abstract—Possibility of performing a mechanically activated synthesis of pure monosilane by the reaction of
calcium hydride with silicon tetrafluoride in a rotating flow-through reactor containing milling bodies was
analyzed. A number of operation parameters of the grinding reactor were theoretically evaluated and
experimentally determined.
DOI: 10.1134/S1070427210060108
High-purity silicon is frequently obtained by using
the hydride method, which consists in synthesis of
monosilane SiH4, followed by purification and thermal
decomposition of this compound [1]. One of
techniques for production of the starting monosilane is
based on the heterogeneous reaction of silicon
trifluoride with calcium hydride CaH2 in a flow-
through reactor:
(>90%) yield of monosilane in terms of SiF4 and CaH2
[6]. Implementation of the method is complicated by
the high synthesis temperature (~400°C) and the
necessity for carrying out additional procedures for
regeneration of the salt melt.
In view of the aforesaid, it is a topical task to
improve methods for synthesis of monosilane and raise
their efficiency.
SiF4 + 2CaH2 → SiH4 + 2CaF2.
(1)
A known way to stimulate heterogeneous chemical
reactions is their mechanical activation [7]. In addition
to activating the reactants, an intense mechanical
treatment causes formation and renewal of the phase
boundary surface, which is necessary for a chemical
interaction to occur [8]. The mechanical activation is
used for synthesis of volatile boron hydrides [9] and
ammonia [10]. A flow-through vibrational installation
for synthesis of monosilane, with counter flows of
lithium hydride and silicon tetrachloride, has been
reported [11] without any data on the efficiency of the
activator of this type.
This method has been successfully used to obtain
isotope-enriched monosilanes [2, 3]. The process is
characterized by a high yield relative to SiF4 (90–
97%), but the conversion of CaH2 in the flow-through
reactor does not exceed 14.5% because of the
formation of a layer of a solid-phase product (CaF2) on
the surface of calcium hydride [4].
To raise the conversion of calcium hydride, the
synthesis of SiH4 by the reaction of SiF4 with CaH2
was performed under the static conditions. By keeping
the reactants for 24 h under a SiF4 pressure of no less
than 30 atm at a temperature of 200°C, the conversion
of CaH2 was raised to 20% [5]. The disadvantages of
the technique consist in its long duration, required
elevated pressure, and need to regenerate or utilize
solid-phase reaction products containing a con-
siderable amount of calcium hydride.
The aim of the present study was to perform a
mechanically activated synthesis of monosilane by the
reaction of silicon tetrafluoride and calcium hydride
and to determined the conversion of calcium hydride.
EXPERIMENTAL
In activators used for gas–solid systems, the
mechanical activation occurs in parallel with grinding
Performing the synthesis of monosilane by reaction
(1) in a salt melt makes it possible to achieve a high
984

MECHANICALLY ACTIVATED SYNTHESIS OF MONOSILANE
985
Table 1. Input data and results of experiments performed for determining the optimal grinding conditions. Grinding duration
60 min
Relative content of indicated fractions, mm, upon grinding, %
n,
φ, %
VGB : VMG
min–1
–1.6+1.0
12.8
0.1
–1.0+0.63
22.1
0.4
–0.63+0.4
14.0
0.5
–0.4+0.2
24.4
5.0
–0.2+0.1
21.9
25.7
2.0
< 0.1
4.8
100
125
150
125
125
125
125
125
125
40
40
40
30
40
50
40
40
40
1:1
1:1
1:1
1:1
1:1
1:1
2:1
1:1
1:2
68.3
1.6
50.7
13.1
0.1
30.1
22.6
0.4
7.0
8.6
8.9
17.8
5.0
6.9
30.7
68.3
55.9
78.5
68.3
30.5
0.5
25.7
29.2
19.9
25.7
22.8
1.1
1.9
2.8
9.1
0
0
0.1
1.5
0.1
0.4
0.5
5.0
2.5
4.7
9.3
30.2
of the solid phase [12]. The following requirements to
an activator for synthesis of monosilane can be
formulated: leak tightness, safety, simplicity of design
and maintenance, and possibility of control over, and
variation of temperature. These requirements are
satisfied by a tumbling or tubular mill in which a flow-
through mode is possible and high output capacity can
be provided. The grinding efficiency in apparatus of
this type depends on its dimensions and design, drum
rotation frequency n, mill-filling coefficient φ,
physicomechanical properties of a material being
ground, and a number of other factors [13].
used as a model material to be ground in these
experiments. The experimental value of E0 for 1-mm
particles of this substance was 3 × 10–3 J, which is
comparable with E0 for CaH2 particles of this size.
As grinding bodies served steel cylinders with a
diameter of 10 mm and height of 10 mm because of
their simpler fabrication and more effective grinding,
compared with balls, via abrasion. The experimental
procedure was as follows. The reactor was charged
with a weighed portion of sodium chloride and
grinding bodies in a certain volume ration; the reactor
was closed, mounted on support bearings, connected to
the gearbox shaft via a pulley, and the required rotation
frequency of the reactor was set. The ground material
was periodically subjected to a sieve analysis; the mass
of fractions of the ground material was determined on
a VL E134 electronic laboratory balance with an
accuracy of ±0.1 g. The experimental data obtained in
a study of how the rotation frequency n, filling
coefficient φ, and volume ratio between the grinding
bodies and the material being ground, VGB : VMG affect
the grinding efficiency are listed in Table 1.
To provide effective grinding and renewal of the
CaH2 surface, a number of operation parameters were
theoretically evaluated and experimentally determined.
The experiments were performed using a flange
reactor made of 12Kh18N10T stainless steel, with an
inner diameter of 66 mm, length of 70 cm, and useful
volume of 2350 cm3. The following calculated values
of the optimal operating parameters were obtained for
the given reactor: rotation frequency n = 124 rpm,
reactor-filling coefficient φ = 39%, minimum grinding
body weight q = 0.015 N. To calculate q, it is
necessary to know the minimum disintegration onset
energy E0 of particles of the material. E0 was
experimentally determined by the method described in
[13]; for CaH2 particles 1 mm in size, E0 was 2 × 10–3 J.
It can be seen in Table 1 that the grinding is more
effective at a rotation frequency of 125 rpm and
reactor-filling coefficient of 40%, in agreement with
the calculated data. It is noteworthy that an increase in
the number of grinding bodies makes the grinding
efficiency higher. However, the amount of calcium
hydride charged into the reactor should be lowered in
this case, which may impair the output capacity of the
technique.
Calcium hydride is a hygroscopic substance whose
hydrolysis yields fine CaO and Ca(OH)2 particles
capable of distorting results obtained in grinding
experiments. Therefore, commercial common salt was
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 83 No. 6 2010

986
TROSHIN et al.
6
5
7
5
6
3
−
−
−
−
−
3
8
3
3
12
3
4
4
−
−
9
1
2
11
10
Fig. 1. Schematic of a setup for mechanically activated synthesis of monosilane. (1) Cylinder with SiF4, (2) cylinder with H2,
(3) vacuum gage, (4) gas flow control valve, (5) junction unit (pivot), (6) supporting bearing, (7) flange reactor, (8) electric motor
with a gearbox, (9) Petryanov filter, (10) cryogenic condenser, (11) receiving cylinder, and (12) rheometer.
The results obtained were used when performing a
mechanochemical synthesis of monosilane by the
reaction of SiF4 with CaH2. The experiments were
carried out on a setup shown schematically in Fig. 1.
The basic units of the setup were cylinders with SiF4
and H2, Bronkhorst Hi Tech gas flow control valves,
above-described flange reactor with a resistive heater,
condenser consisting of three in-series connected traps,
and receiving cylinder. Silicon tetrafluoride was
produced by thermal decomposition of Na2SiF6 [14];
CaF2 was synthesized from shavings of distilled
calcium and purified hydrogen [15]; on being
synthesized, calcium hydride was ground in a box with
dried atmosphere. The –1.0 + 0.63 mm fraction was
used in the experiments.
The experimental procedure included the following
stages: charging of calcium hydride and grinding
bodies into the reactor, evacuation and heating of the
reactor to 180–190°C, and delivery of a mixture of
silicon tetrafluoride and hydrogen (50 : 50 vol %) into
the rotating reactor. Gaseous reaction products were
frozen-out in the condenser cooled with liquid nitrogen
and, after an experiment was complete, were loaded
into the receiving cylinder. The mass of consumed
SiF4 and gaseous reaction products was periodically
determined by weighing cylinders with gases on an
SP-30 electronic balance with an accuracy of ±2 g. The
experiments were continued until gaseous monosilane
ceased to be released, as indicated by gravimetry and
gas chromatography.
Table 2. Results of a laser mass-spectrometric analysis of
calcium hydride samples
CaH2 content, ppm
Element
starting
<0.05
≤1700
≤400
≤56000
61
upon synthesis
В
0.2
≤680
≤13
25000
>20 at %
138
C
N
O
F
Mg
Si
S
130
30
15000
14
9.5
Figures 2 and 3 show how the content of SiH4 in
the gaseous reaction products and the conversion of
CaH2 depend on the duration of mechanical activation.
It can be seen that, after approximately 30% of calcium
hydride is consumed, the content of monosilane in the
gaseous reaction products decreases and that of SiF4
increases. After about 60% of calcium hydride is
consumed, monosilane almost ceases to be formed. A
possible reason why the reactant conversion decreases
Cr
Mn
Fe
Ni
Cu
Sr
17
13
21
22
29
50
3.8
6.5
17
67
93
98
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 83 No. 6 2010

MECHANICALLY ACTIVATED SYNTHESIS OF MONOSILANE
987
c, vol %
100
80
x, %
60
40
70
50
20
30
10
4
8
12
16
τ, h
4
8
12
16
τ, h
Fig. 2. Content c of SiH4 in gaseous reaction products vs.
the duration τ of mechanical activation. m(CaH2) = 500 g;
volumetric flow rate of SiF4, 10 l h–1; φ = 40%; n = 125 ± 2 rpm;
Fig. 3. Conversion x of CaH2 vs. the duration τ of mechanical
V
GB : VMG = 1 : 1; the same for Fig. 3.
activation.
is that the limiting grinding of calcium hydride
particles is reached [16], with the result that the surface
renewal and the reaction of calcium hydride with
silicon tetrafluoride almost cease. This assumption is
confirmed by the presence of a considerable amount of
calcium hydride in the solid phase, determined by its
dissolution in water. The reaction of spent calcium
hydride in water has an induction period (3–4 s),
whereas the starting substance reacts with water almost
instantaneously.
spent CaH2 decreases. Pres-umably, this is due to the
formation of volatile com-pounds of these elements
with hydrogen and washing-out of these compounds
from calcium hydride.
The monosilane sample obtained was analyzed for
the content of C1–C4 hydrocarbon impurities by the
method described in [18]. The results obtained are
listed in Table 3. It can be seen that the total content of
C1–C4 hydrocarbon impurities in monosilane does not
exceed 3 × 10–3 mol %, which is comparable with the
content of these impurities in monosilane produced
from SiF4 and CaH2 by other methods without
mechanical activation [3, 6].
The particle size of the solid reagent was evaluated
upon the synthesis by optical microscopy on a Zeiss
Axioplan 2 imaging instrument. It was found that the
main part of particles or their agglomerates have sizes
of the order of n × 10–6 m.
In conclusion, the following can be noted. The
mechanically activated synthesis of monosilane from
silicon tetrafluoride and calcium hydride does not
require any additional reagents (solvents and salt
melts), which makes unnecessary their regeneration.
The technique is comparatively simple methodolo-
It is noteworthy that the coloration of the solid
phase changes upon synthesis from light gray to
brown. A possible reason for the coloration change is
that amorphous silicon is formed via reduction of
silicon (SiIV → Si0) or thermal decomposition of mono-
silane. This assumption is confirmed by the results of a
laser mass-spectrometric analysis of the starting CaH2
and solid-phase reaction products (Table 2). According
to the data in Table 2, the content of fluorine and
silicon in the solid phase substantially increases. It
should be noted that the reaction of SiF4 with CaF2 to
give CaSiF6 is possible; however, the equilibrium of
this reaction is almost completely shifted to the left at a
temperature of 195°C [17]. The high content of the
oxygen impurity is due to the interaction of CaH2 with
water vapor in sample preparation. The content of gas-
forming elements, such as carbon and nitrogen, in
Table 3. Content of C1–C4 hydrocarbon impurities in mono-
silane according to gas-chromatographic data
Impurity
СН4
Content, mol %
(2.5±0.3)×10–3
(1.3±0.1)×10–4
(2.6±0.3)×10–4
(4.0±0.5)×10–5
(1.8±0.2)×10–5
(3.7±0.7)×10–6
С2Н4
С2Н6
С3Н8
n-С4Н10
iso-С4Н10
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 83 No. 6 2010

988
TROSHIN et al.
3. Devyatykh, G.G., Dianov, E.M., Bulanov, A.D., et al.,
gically and instrumentally and can substantially raise
the conversion of calcium hydride. A promising way to
improve the output capacity is to successively perform
synthesis of calcium hydride, its grinding, and
synthesis of monosilane in the same reactor or to or-
ganize a continuous process. The mechanical treatment
of calcium hydride in synthesis of monosilane yields
fine particles, and, therefore, it is necessary to take
measures to preclude carry-over of these particles from
the reactor and to remove inhomogeneous impurities
from the monosilane obtained.
Dokl. Akad. Nauk, 2003, vol. 391, no. 5, pp. 638–639.
4. Bulanov, A.D., Mikheev, V.S., Troshin, O.Yu., and
Lashkov, A.Yu., Zh. Neorg. Khim., 2008, vol. 53, no. 1,
pp. 11–15.
5. RF Patent 2226501.
6. RF Patent 2077483.
7. Avvakumov, E.G., Mekhanicheskie metody aktivatsii
khimicheskikh protsessov (Mechanical Methods for
Activation of Chemical Processes), Novosibirsk: Nauka,
1986.
8. Smolyakov, V.K., Lapshin, O.V., and Boldyrev, V.V.,
Teor. Osn. Khim. Tekhnol., 2008, vol. 42, no. 1, pp. 57–
62.
CONCLUSIONS
9. Volkov, V.V. and Myakishev, K.G., Khim. Interesakh
Ustoich. Razv., 2002, vol. 10, nos. 1–2, pp. 31–44.
10. Volkov, V.V., Myakishev, K.G., and Drozdova, M.K.,
Khim. Interesakh Ustoich. Razv., 2000, vol. 8, no. 3,
pp. 363–366.
(1)
A
mechanically activated synthesis of
monosilane by the reaction of silicon tetrafluoride with
calcium hydride was performed in a flow-through
reactor. It was found that the conversion of calcium
hydride reaches a value of 60%.
11. RF Patent 2164218.
(2) The conversion of calcium hydride is 30% at a
nearly complete conversion of silicon tetrafluoride in
mechanically activated synthesis.
12. Baranov, D.A., Blinichev, V.N.,Vyaz’min, A.V., et al.,
Protsessy i apparaty khimicheskoi tekhnologii, tom 2.
Mekhanicheskie
i
gidromekhanicheskie protsessy
(Processes and Apparatus of Chemical Technology, vol. 2,
Mechanical and Hydrodynamic Processes), Kutepov, A.M.,
Ed., Moscow: Logos, 2001.
(3) The total content of C1–C4 hydrocarbon (CH4,
C2H4, C2H6, C3H8, n-C4H10, iso-C4H10) impurities in
the resulting monosilane does not exceed 3 × 10–3 mol %
according to gas-chromatographic data.
13. Sidenko, P.M., Izmel’chenie v khimicheskoi promysh-
lennosti (Grinding in Chemical Industry), Moscow:
Khimiya, 1977, 2nd revised ed.
ACKNOWLEDGMENTS
14. Bulanov, A.D., Pryakhin, D.A., and Balabanov, V.V.,
Zh. Prikl. Khim., 2003, vol. 76, no. 9, pp. 1433–1435.
The study was financially supported by the RF
Presidential grant (NSh 4701.2008.3) in the framework
of the Program of State support for leading scientific
schools of the Russian Federation.
15. Bulanov, A.D., Troshin, O.Yu., and Balabanov, V.V.,
Zh. Prikl. Khim., 2004, vol. 77, no. 6, pp. 887–889.
16. Vasil’ev, L.S. and Lomaeva, S.F., Khim. Interesakh
Ustoich. Razv., 2002, vol. 10, nos. 1–2, pp. 13–22.
17. Issledovanie termicheskoi stoikosti ftorboratov shche-
lochnozemel’nykh metallov [Elektronnyi resurs] (Study
of the Thermal Stability of Fluoroborates of Alkaline-
Earth Metals [Electronic Resource]), Seversk. Gos.
REFERENCES
1. Devyatykh, G.G. and Zorin, A.D., Letuchie neor-
ganicheskie gidridy osoboi chistoty (Special-Purity
Volatile Inorganic Hydrides), Moscow: Nauka, 1974.
Tekhnol. Inst.,
URL: http://www. science.ssti.ru/
progress/works/m2.pdf (retrieval date 15.06.2009).
18. Krylov, V.A., Malygina, L.S., Salganskii, Yu.M., et al.,
Zh. Anal. Khim., 2003, vol. 58, no. 8, pp. 841–846.
2. Bulanov, A.D., Balabanov, V.V., Pryakhin, D.A., et al.,
Neorg. Mater., 2002, vol. 38, no. 3, pp. 356–361.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 83 No. 6 2010