Академия наук республики узбекистан

Fig.1.2. General view of the experimental setup for the implementation of the process of 
synthesis of alkoxysilane 
Fig. 1.
3 
illustrates the examples of 
technology operation with ethanol. 
Fig 1.3. The dependence of the reaction rate 
on duration of triethoxysilane synthesis 
process 
In order to compare the results, in the 
first case, 
metallurgical 
silicon (purity 
99%) was ground under normal 
105 

---

conditions, i.e. by a dry method to a size of partic
les up to 30-
100 microns, and 
loaded with a catalyst into the reactor, after reaching a temperature of 240° C of 
the reaction mixture, alcohol 
- absolute ethanol was fed to the reactor with the rate 
of about 600 ml per hour. At the moment of appearance of the liquid product from 
the reactor and then every 30 min the samples were taken. Analysis of the samples 
with a gas chromatograph Agilent® GC 7890A showed that the triethoxysilane 
synthesis reaction of metallurgical silicon and ethyl alcohol began only 150 
minutes after alcohol feeding, and its intensity gradually increases (Fig. 1.
3
, Curve 
1), then the reaction damped
at 500 minutes after the beginning of the alcohol 
feeding. Selectivity in triethoxysilane was 70%. 
The second experiment was carried out u
nder the same conditions as in the 
first one, but with the fundamental difference in environment of reagent 
preparation. 
Milling
of metallurgical silicon was made by a wet method directly in 
the solvent environment - 
THERMINOL® 66. During the 
milling process a 
catalyst was introduced into the suspension. The synthesis reaction started at the 
beginning of the tenth minute after the alcohol feeding into the reactor and the 
reaction rate increased for the first 60 minutes (Fig. 1.
3
, Curve 2). The reaction rate
of triethoxysilane synthesis began to decrease after 180 minutes of the process, and 
the reaction completely damped at 
260 minutes after the start of the alcohol 
feeding. Selectivity in triethoxysilane was 94%. The reason for the reaction 
damping in this case is complete consumption of silicon loaded in the beginning of 
the synthesis process. 
In the third experiment under the same conditions the silicon consumption 
was compensated by silicon feeding into the suspension according to the formula:
m
Si
 = k
1
·m
TES
 + k
2
·m
TEOS, 
 
(1.5
) 
where m
TES
is the mass of triethoxysilane, m
TEOS
is the mass of tetraethoxysilane 
obtained in units of time as a result of the direct reaction k
1
and k
2
are the 
coefficients taking into account consumption of silicon in the synthesis of tri- and 
tetraethoxysilane, respectively. In this case, k
1
= 0.171 and k
2
= 0.135. The 
synthesis reaction started nine minutes after the beginning of the alcohol feeding 
into the reactor, the reaction rate increased for the first 90 minutes and then 
s
tabilized (Fig. 1.
3
, Curve 3) at 400 g/h of triethoxysilane. The reaction was 
deliberately stopped at the 250th minute after the start of alcohol feeding because 
of the appearance of an abundant mass of foam in the reaction products. Feeding of 
additional amounts of solvent in the suspension resulted in increasing the reaction 
mass in the reactor, which caused foam formation. Selectivity in triethoxysilane 
was 93%.
In the fourth case, we have implemented all the proposed technical solutions 
for the alkoxysilane synthesis process: 1) silicon 
milling in a protective 
environment, 2) continuous compensation of silicon consumption, 3) continuously 
removal of excess solvent through the porous ceramic membrane mounted on the 
reactor wall. The synthesis reaction sta
rted ten minutes after the start of the alcohol 
feeding into the reactor, the reaction rate sharply increased for the first 60 minutes, 
106

---

then after the ~120th minute there was a slow growth, and further the rate 
stabilized (Fig. 1.
3
, Curve 4). For 500 minu
tes of continuous feeding of the 
suspension consisting of 600 g of silica and 1200 g of solvent 
- 
THERMINOL® 66 
into the reactor, 3380 g of triethoxysilane and 141 g of tetraethoxysilane were 
obtained. Selectivity in triethoxysilane was 96%.
After filtrati
on and cleaning, the solvent is returned to the technological 
process. Analysis showed that the impurity composition of filtercake solvent 
corresponds to that of initial metallurgical silicon. It can be concluded that the 
impurities in the initial silicon