Experimental study and simulation of kinetics of acetophenone hydrosilylation with diphenylsilane in the presence of rhodium complexes in a microreactor

Article abstract of DOI:10.1134/S1070363210110071

The hydrosilylation of acetophenone with diphenylsilane in a microreactor in the presence of complexes [Rh(cod)Cl]₂ and [Rh(CO) ₂(μ-Cl)]₂ and (R)-(-)-cis-mirtanyl- and (R)-(+)-bornylamine in situ was studied, the kinetics simulation of the process was performed, and the multicriteria optimization of the process was carried out. The influence of the micro-mixing effect on the reaction rate was revealed. Best results in the microreactor were obtained for the [Rh(cod)Cl] ₂-BornylNH₂ catalytic system. It was established that the formation of 1-phenylethanol and related enol silyl ethers are simultaneous competing reactions. Pleiades Publishing, Ltd., 2010.

Full text of DOI:10.1134/S1070363210110071

ISSN 1070-3632, Russian Journal of General Chemistry, 2010, Vol. 80, No. 11, pp. 2263–2273. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © V.M. Uvarov, E.S. Borovinskaya, D.A. de Vekki, V.P. Reshetilovskii, 2010, published in Zhurnal Obshchei Khimii, 2010, Vol. 80,
No. 11, pp. 1799–1809.
Experimental Study and Simulation of Kinetics
of Acetophenone Hydrosilylation with Diphenylsilane
in the Presence of Rhodium Complexes in a Microreactor
V. M. Uvarova, E. S. Borovinskayaa, D. A. de Vekkia, and V. P. Reshetilovskiib
a St. Petersburg State Institute of Technology, Moscovskii pr. 26, St. Petersburg, 190013 Russia
e-mail: hydrosilation@newmail.ru
b Technical University of Dresden, 01062, Dresden, Germany
Received January 19, 2010
Abstract―The hydrosilylation of acetophenone with diphenylsilane in a microreactor in the presence of
complexes [Rh(cod)Cl]2 and [Rh(CO)2(μ-Cl)]2 and (R)-(–)-cis-mirtanyl- and (R)-(+)-bornylamine in situ was
studied, the kinetics simulation of the process was performed, and the multicriteria optimization of the process
was carried out. The influence of the micro-mixing effect on the reaction rate was revealed. Best results in the
microreactor were obtained for the [Rh(cod)Cl]2–BornylNH2 catalytic system. It was established that the
formation of 1-phenylethanol and related enol silyl ethers are simultaneous competing reactions.
DOI: 10.1134/S1070363210110071
Microstructural reactors (microreactors) compose
an innovative class of reaction equipment with the
internal size generally not exceeding 2 mm. They have
several advantages over conventional reactor, like
higher safety, better control over the flow parameters,
short time of diffusion of reactants, high specific
surface area of interaction and efficiency of heat
transfer [1, 2]. In addition, the research based on the
microreactor technique allows the convergence of the
conditions of the laboratory experiments to industrial
processes and provides more opportunities for the
study of chemical reactions and improve their
efficiency [2]. These benefits contribute to high
potential of microstructural reactors for use in the field
of analytical and synthetic chemistry [2–9].
The high chemo- and stereoselectivity of hydro-
silylation of C=O group is achieved using catalytic
systems based on the rhodium(I) complexes with
various nitrogen-containing chiral ligands.
A
disadvantage of the ligands of this type is relative
complexity of their synthesis and the low rate of the
hydrosilylation on such catalysts [12, 13].
An approach to solving this problem is to use the
nitrogen-containing ligands based on available natural
monoterpenes [14–17]. At the same time, hydro-
silylation using microreactor technique may result in
increased rate and selectivity of the process. Todayit is
known that the hydrosilylation of alkenes and alkynes
can be successfully carried out in the microstructured
reactor [18].
A highly effective method for the synthesis of
organosilicon compounds is the catalytic hydro-
silylation. At present, importance of this reaction in-
creases due to new fields of application of organo-
silicon compounds like the synthesis of biologically
active substances, flavorings, immobilized catalysts,
insec-ticides, etc. Of particular interest is the hydro-
silylation of ketones, an alternative to catalytic hydro-
genation and reduction of C=O group by metal
hydrides [10, 11].
In this regard, we considered it important and
interesting to study the reaction of liquid-phase hydro-
silylation in the presence of rhodium complexes and
chiral nitrogen-containing monoterpenes as ligands in
a flow microreactor.
Description of the reaction system. The hydro-
silylation of acetophenone with diphenylsilane is used
often as a model reaction for estimating effectiveness
of rhodium catalysts and studying the effect of various
2263

2264
UVAROV et al.
conditions on the hydrosilylation of ketones [13, 19,
20]. The reaction products are 1-phenylethanol and
acetophenone enol silyl ethers. The subsequent
hydrolysis of the mixture of products leads to the
formation of 1-phenylethanol (S)- and (R)-isomers and
regeneration of the parent ketone.
H2O, H+
H3C
H
H3C
H
H3C
O
H2C
O
O
OH
SiHPh2
+
SiHPh2
H2O, H+
[Rh]
+ Ph2SiH2
0−5°C
(S), (R) isomers
(S), (R) isomers
As a pre-catalysts of the process the complexes
[Rh(cod)Cl]2 and [Rh(CO)2(μ-Cl)]2 were selected,
which have high catalytic activity in hydrosilylation of
ketones and are used traditionally as pre-catalysts for
the reactions of this type [13, 15]. The ligands used in
this work were the industrially available nitrogen-
containing derivatives of mono- and bicyclic terpenes,
(R)-(–)-cis-mirtanylamine (MyrtNH2, I) and (R)-(+)-
bornylamine (BornylNH2, II) that differ by the
distance between the coordinating nitrogen atom and
asymmetric center.
The formation of the catalytic system was per-
formed by adding respective amine in situ, since the
interaction of rhodium complex with the amine
proceeded rapidly enough [15, 21], and the complex
formed was not inferior by its catalytic properties to
one synthesized in advance. Technologically this is
significantly simpler because it allows avoiding the
laborious stage of the synthesis of the complex proper
[15]. In our experiments we used the amine in excess
from 2.1 to 10-fold with respect to the complex, since
it is well known [22–25] that the introduction of 2–13-
fold molar excess of chiral ligand in situ can increase
asymmetric induction. The acetophenone conversion in
a certain time period was taken as a measure of the
catalytic activity of the system.
NH2
H
NH2
CH3
I
II
Research installation. The microreactor setup for
the hydrosilylation comprises individual modules
placed on the assembling plate and having standard
interfaces to provide the flexibility of the reactor at its
use in the desired reaction conditions (Fig. 1) [3].
Acetophenone, rhodium(I)
complex, and ligand
The microreactor was filled with anhydrous THF
and cooled to 0°C using micro heat exchangers and a
thermostat. The reagents are fed into the micro heat
exchangers for cooling to 0°C, and then passed in a
micromixer, in which each stream is divided into many
layers passing the flat channels of 70 μm thickness.
Then the layers are directed to mixing. The resulting
reaction mixture enters the microreactor, which is a
temperature-controlled thin tube of 80 μm inner
diameter, the typical residence time of the reaction
mixture is 15 min. At the reactor output the product
was sampled to determine the acetophenone
conversion and the reaction chemoselectivity by means
1-Phenyletanol and enol
silyl ethers and 1,1,3,3-
tetraphenyldisiloxane
Diphenylsilane
1
Fig. 1. Scheme of the installation for hydrosilylation of
acetophenone with diphenylsilane in a microstructured reactor.
of H NMR spectroscopy. Then the reaction mixture
enters the refrigerated container, where the hydrolysis
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

EXPERIMENTAL STUDY AND SIMULATION OF KINETICS
2265
Hydrosilylation of acetophenone with diphenylsilane in the microreactor and the mixing reactor
Conversion of acetophenone,
Yield,b %
Selectivity, %
%
Yield,a %
mixing reactor,
Complex
Ligand (L) L:Rh
mixing reactor, min
mixing reactor, min
min
15
65
–
30
84
–
60
95
92
76
92
83
74
36
81
74
70
22
85
80
15
15
86
–
30
81
–
60
80
70
68
71
62
69
78
70
66
34
57
27
33
56
15 30 60
[Rh(cod)Cl]2
–
–
2.1
5
68 77
6
82
65
61
53
50
56
40
100
60
12
12с
41
9
82
73
78
71
67
60
90
90
15
73
75с
21
50
67
47
48
38
33
34
36
90
9
56
–
BornylNH2
–
–
–
–
–
–
65
52
66
52
51
28
13
13
13
12
12
16
6
–
–
–
–
–
10
2.1
5
–
–
–
–
–
MyrtNH2
–
–
42
18
–
68
72
–
29
12
–
–
–
10
–
–
–
[Rh(CO)2(μ-Cl)]2
–
80
–
73
–
58 57
80
–
73
–
58
–
BornylNH2
2.1
5
–
–
–
–
–
–
49
23
13
23
26
8
20
22
4
–
–
–
–
9
–
10
2.1
5
–
–
–
–
9с
–
MyrtNH2
–
–
–
–
9
–
30
5
–
–
–
–
5
–
10
8
–
–
62
–
–
5
–
5
a
1
b
Yield of 1-phenylethanol silyl ether according to H NMR spectrum. Yield of the products of dehydrocondensation (tetraphenyldi-
siloxane and other siloxanes), according to 1H NMR spectroscopy. c Reaction duration 35 min.
Yield of tetraphenyldisiloxane in the microreactor
depends considerably on the applied catalytic system
(see the table). Use of [Rh(cod)Cl]2 or [Rh(CO)2(μ-Cl)]2
in the absence of amines gives disiloxane in only 6%
yield, whereas the catalytic system based on [Rh(cod)Cl]2
and amines added in situ leads to a two-fold increase in
the yield. Unlike cyclooctadiene catalytic system, the
yield of tetraphenyldisiloxane with [Rh(CO)2(μ-Cl)]2
depends on the structure and the amount of added
amines and reaches its maximum value (30%) when
the [Rh(CO)2(μ-Cl)]2 : MyrtNH2 ratio equal 1 : 2.1.
The yield of tetraphenyldisiloxane in traditional
mixing reactor is by 2–5% higher than in microreactors
[15]. Increasing the reaction duration has practically no
effect on the siloxane yield.
of the formed silyl ether is carried out to determine its
enantiomeric composition.
Experimental research. The hydrosilylation of
acetophenone with diphenylsilane in the microreactor
in the presence of rhodium catalysts proceeds in
accordance with the above scheme. Carrying out the
reaction in aerobic conditions lead to the additional
1
appearance in the H–29Si 2D NMR spectrum of the
signals at δH 5.67 ppm and δSi –19.21 ppm, corres-
ponding to SiH group of 1,1,3,3-tetraphenyldisiloxane.
1,1,3,3-Tetraphenyldisiloxane is a typical product of
catalytic conversion of diphenylsilane at the hydro-
silylation in a non-inert atmosphere in a traditional
mixing reactor [15, 26], and its formation is a result of
the reaction competing with the hydrosilylation.
The catalytic system based on [Rh(cod)Cl]2
exhibits higher activity in hydrosilylation in the micro-
reactor compared with the traditional one. The aceto-
phenone conversion after 15 min in the presence of
[Rh], [O]
HPh2Si−O−SiPh2H
Ph2SiH2
H+, H2O
Polysiloxanes.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

2266
UVAROV et al.
[Rh(cod)Cl]2 without amines is 82% in the micro-
reactor and only 65% in the mixing reactor (see the
table). In the traditional mixing reactor the ketone
conversion reaches 84% only after 30 min, which is
close to the conversion in the microreactor in 15 min,
and 60 min is required to increase it to 95%.
selectivity, which reaches the value obtained in the
reaction without adding amine (78%).
Yield of 1-phenylethanol silyl ether in the micro-
reactor is little dependent on the amount of MyrtNH2
in the catalytic system (see the table), whereas in the
traditional reactor at increase in MyrtNH2 con-
centration the yield decreases.
The reaction chemoselectivity achieved in the
microreactor within 15 min is identical numerically
with the conversion of acetophenone (see the table),
the yield of 1-phenylethanol silyl ether is 67%. In the
traditional reactor the chemoselectivity in 15 min is
somewhat higher than in the microreactor, but the
yield of product is less than 11%. The increase in the
reaction duration in the mixing reactor to 60 min leads
to an increase in the 1-phenylethanol silyl ether yield
to 77%, while chemoselectivity remains virtually
unchanged.
The hydrosilylation of acetophenone in the pres-
ence of BornylNH2 proceeds more selectively than
with MyrtNH2. The chemoselectivity in the micro-
reactor at the [Rh(cod)Cl]2 : BornylNH2 ratio 1 : 5 is
78%. The use of amine both in higher and lower
concentration produces virtually no effect on
chemoselectivity, which is 73% and 71% for 2.1- and
10-fold BornylNH2 excess, respectively. The yield of
1-phenylethanol silyl ether in the microreactor
decreases only in going from 5-fold to 10-fold amount
of BornylNH2 (see the table). It is interesting to note
that in the traditional reactor the chemoselectivity after
60 min does not depend on the amount of BornylNH2
(~70%) and is comparable with the selectivity in the
microreactor after 15 min. The exception is the ratio of
[Rh(cod)Cl]2 : BornylNH2 equal to 1:5, at which the
selectivity is by 10% less than in the microreactor (see
the table). That is, the selectivity of the reaction is
almost independent of the content of BornylNH2 both
in the microreactor and in the traditional mixing
reactor.
Introducing in situ of an amine in the catalytic
system based on [Rh(cod)Cl]2 causes a decrease in the
reaction activity and selectivity compared to the
hydrosilylation without amines both in the micro-
reactor and in the traditional reactor.
The acetophenone conversion achieved in the
microreactor in 15 min in the presence of 2.1-fold and
5-fold amount of MyrtNH2 is 50 and 56% respectively
against 42 and 18% in the traditional reactor (see the
table). That is, the catalytic activity of [Rh(cod)Cl]2–
MyrtNH2 system in the microreactor is higher than in a
traditional reactor. Increasing the excess to 10-fold
leads to further decrease in the acetophenone con-
version, which is in the microreactor does not exceed
40%.
In the microreactor the catalytic system based on
[Rh(CO)2(μ-Cl)]2 in the absence of amines is more
active and selective than the system containing
[Rh(cod)Cl]2. The acetophenone conversion in the case
of [Rh(CO)2(μ-Cl)]2 is by 18% higher than in the
presence of [Rh(cod)Cl]2 (see the table), the selectivity
of the reaction with the carbonyl complexes is higher
than in the presence of [Rh(cod)Cl]2. In the traditional
reactor in the presence of carbonyl complex both
conversion (80–81%) and the reaction chemo-
selectivity (70–73%) are lower than in the micro-
reactor and practically do not depend on the reaction
duration (15–60 min), whereas at the hydrosilylation
with the cyclooctadiene complex an increase in
conversion with time is observed.
Bornylamine inhibites the reaction less than mirt-
anylamine does (see the table). The conversion of
acetophenone in the presence of bornylamine at a
[Rh(cod)Cl]2: BornylNH2 ratio equal to 1: 2.1 and 1: 5
is 65 and 61% respectively, and 10-fold excess of
BornylNH2 leads to an even greater decrease in the
ketone conversion (53%).
Selectivity of the 1-phenylethanol silyl ether
forma-tion in the presence of the [Rh(cod)Cl]2
complex and 2.1-fold amount of MyrtNH2 both in
microreactor, and in the mixing reactor after 15 min is
67%, that is, it is by 15% less than in the absence of
MyrtNH2. The increase in the amount of amine to 10-
fold shows a positive effect on the reaction
chemoselectivity in the microreactor, extending it to
90%. In the case of conventional reactor, a 10-fold
excess of MyrtNH2 also leads to an increase in
The mirtanyl- and bornylamines introduced in situ
in the catalytic carbonyl-containing system reduce the
activity of the system compared with the activity in the
absence of amine, as was in the case when [Rh(cod)Cl]2
was used as a pre-catalyst. However, the activity of the
[Rh(CO)2(μ-Cl)]2–amine system was lower than that of
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

EXPERIMENTAL STUDY AND SIMULATION OF KINETICS
2267
the systems based on [Rh(cod)Cl]2. This occurs at the
hydrosilylation both in the microreactor and in the
traditional reactor. This probably is connected with the
irreversible destruction of the catalytically active
carbonyl metallocomplex intermediate (or the true
catalyst) as a result of its reduction, which is observed
only in this catalytic system.
60 min in the traditional reactor. At the same time the
yield of 1-phenylethanol silyl ether in the presence of
[Rh(CO)2(μ-Cl)]2 and amine depends only slightly on
the structure and quantity of the nitrogen-containing
compounds (see the table), due both to low aceto-
phenone conversion and reaction selectively.
Asymmetric induction at the acetophenone hydro-
silylation with diphenylsilane in the presence of
rhodium complexes with chiral amines in the
microreactor does not exceed 8% . In all cases mainly
1-(R)-phenylethanol is formed, as is the case of
hydrosilylation in the traditional mixing reactor [15].
The maximum enantiomeric excess observed in the
presence of [Rh(cod)Cl]2 with BornylNH2 in the 1 : 10
ratio. A decrease in amount of BornylNH2 reduces the
enantiomeric excess to 5 and 2% ee at 5- and 2.1-fold
excess of amine relative to the complex respectively.
The enantiomeric excess achieved with the catalytic
system based on rhodium carbonyl complex is 5% ee
at carrying out the reaction in either the microreactor
or traditional mixing reactor (the ratio [Rh(CO)2(μ-Cl)]2 :
BornylNH2 = 1: 2.1). Note that the asymmetric
induction in the microreactor in 15 min can be
achieved only in the presence of BornylNH2, whereas
in the traditional mixing reactor both amines are
“working” [15]. The difficulty of achieving asym-
metric induction in the presence of MyrtNH2 may be
due to the presence a methylene hinge in its structure
allowing free rotation of the ligand chiral center
around the metal atom.
The maximum activity of the catalytic system
based on [Rh(CO)2(μ-Cl)]2 at carrying out the reaction
in the microreactor is achieved in the presence of
BornylNH2 at the complex : amine ratio 1 : 2.1 (60%
conversion), but the selectivity is extremely low
(15%). Replacing BornylNH2 by MyrtNH2 at the same
ratio of complex to amine leads to reduced activity of
the catalytic system, but increases selectivity to 21%
(see the table). At the same time in the traditional
reactor the system [Rh(CO)2(μ-Cl)]2–MyrtNH2 shows
higher activity. Conversion of acetophenone in the
microreactor at the [Rh(CO)2(μ-Cl)]2 : ligand ratio 1 : 5
is ~5 times less than at the ratio 1 : 2.1.
A 10-fold amount of BornylNH2 virtually deac-
tivates catalysts (after 15 min no formation of products
is observed), and in the case of MyrtNH2 the aceto-
phenone conversion did not exceed 8%. Thus, the catalytic
activity of the systems based on [Rh(CO)2(μ-Cl)]2 and
amines is not only lower than that of the systems based
on [Rh(cod)Cl]2, but shows greater sensitivity to the
content of amine. Apparently, amine increases stability
of the catalytic system, preventing the process of
reduction.
The selectivity of formation of 1-phenylethyl silyl
ether in the microreactor at the use of the catalytic
system based on [Rh(CO)2(μ-Cl)]2 and an amine is
lower than at the hydrosilylation with the [Rh(cod)Cl]2–
amine system. Minimum chemoselectivity corresponds
to the catalyst : amine ratio equal to 1 : 2.1 and the
maximum, to the ratio 1 : 10. In addition, at low
concentration of amine a higher selectivity is observed
in the presence of MyrtNH2, but with an increase in the
amine quantity the selectivity of hydrosilylation is
higher in the presence of the system containing
BornylNH2. The hydrosilylation in the traditional
reactor in the presence of a catalytic system containing
BornylNH2 proceeds more selectively than with the
system based on MyrtNH2 although the general trend
of increasing chemoselectivity with increasing amine
concentration in the mixture is retained. It is
noteworthy that in most cases the selectivity of the
reaction after 15 min in the microreactor at the use of
[Rh(CO)2(μ-Cl)]2 and amines is higher than after
Simulation of the process. The process of hydro-
silylation of acetophenone with diphenylsilane is a
homogeneous reaction. Its kinetics can be simulated in
the framework of the formal mathematic description.
In this regard, in the first stage of the simulation we
assumed the possible reaction schemes and their
formalization. Then the differential equations were
generated, the inverse kinetic problem was solved, and
the multicriteria optimization of process wasperformed
using the Kinetic software package [27] designed to
simulate the kinetics of chemical processes. The
transformation of diphenylsilane in 1,1,3,3-tetra-
phenyldisiloxane was not taken into account at the
simulation because its rate on the average is 3–
15 times lower than the rate of the formation of the
desired product. The increase in the rate of this process
was observed only at the almost complete absence of
the target product, that is, in the conditions, which are
not of interest for the hydrosilylation.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

2268
UVAROV et al.
of chemical reactions of various types suggests that in
H3C
H
O
this case the formation of 1-phenylethanol and enol
silyl ethers is the result of two parallel competitive
reactions.
SiHPh2
H3C
O
The results obtained by simulation allow us the
estimation of the difference in the rates of hydro-
silylation in the traditional mixing reactor and in the
microreactor. Owing to the effect of micro mixing, in
the microreactor in the absence of amines both the
catalytically active complexes show a fairly high rate
of the process in the first section of the kinetic curve
(Figs. 2–5). The rhodium(I) carbonyl complex by its
activity exceeds cyclooctadiene complex both in the
traditional reactor, and in the microreacor, the best
result is observed in the presence of [Rh(CO)2(μ-Cl)]2
at carrying out the reaction in the microreactor
(acetophenone conversion 100%, Figs. 3, 5).
+ Ph2SiH2
H2C
O
SiHPh2
The kinetic curves for the reaction of acetophenone
with diphenylsilane (Figs. 2–5) usually are charac-
terized by the presence of two steps with different rate.
The first section corresponds to the hydrosilylation
proceeding with a high rate, therewith in the reaction
medium there is a sufficient amount of the reactants
and catalytically active particles. The second section
corresponds to slowing down of the reaction due to the
decrease in the concentration of reactants in the
mixture and possible deactivation of the catalyst.
As already noted, the amine addition decreases the
catalytic properties of cyclooctadiene complex, even to
a greater extent than of the carbonyl complexes. A
catalytic system based on [Rh(CO)2(μ-Cl)]2 in this case
is not selective, as practically directs the reaction
entirely to the formation of enol silyl ether and ceases
to “work” towards the formation of the desired product
(see the table). This experimental fact agrees with the
assumption of parallel competing reactions. Applica-
tion of microreactor as compared to the traditional
reactor, for the hydrosilylation in the presence of both
complexes and amines, initially increases the reaction
rate. This is more expressed with the catalytic system
[Rh(cod)Cl]2–MyrtNH2 (Fig. 2). Increase in the
amount of the ligand added in situ in the case of this
system accelerates the hydrosilylation in the
The result of simulation of the process kinetics in
the presence of the complexes [Rh(cod)Cl]2 or
[Rh(CO)2(μ-Cl)]2 with various amounts of MyrtNH2 or
BornylNH2 are plotted in Figs. 2–5 where the kinetic
curves are obtained by simulation while the dots
correspond to experimental observations. In this case
the error of calculations is within the experimental
errors, which confirms the authenticity of the cal-
culations. The analysis of the kinetic curves of the
process taking into account the published data [28] and
comparing the simulation results with the regularities
(a)
(b)
20
60
100
140
180
Reaction time, min
Reaction time, min
Fig. 2. Kinetic curves of the changes in the concentration of acetophenone and accumulation of 1-phenylethanol silyl ester in the
(a) traditional reactor and (b) microreactor at the varied [Rh(cod)Cl2]2 : MyrtNH2 ratio (dots correspond to experimental data, lines
are obtained by simulation): (1, 6) 1:0, (2, 4) 1:10, (3, 5) 1:2.1; (1–3) silyl ether, (4–6) acetophenone.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

EXPERIMENTAL STUDY AND SIMULATION OF KINETICS
(a)
2269
(a)
20
60
100
140
180
20
60
100
140
180
Reaction time, min
Reaction time, min
(b)
(b)
20
60
100
140
180
20
60
100
140
180
Reaction time, min
Reaction time, min
Fig. 3. Kinetic curves of the changes in the concentration
of acetophenone and accumulation of 1-phenylethanol silyl
ether in the (a) traditional reactor and (b) microreactor at
the varied ratio of [Rh(cod)Cl2]2: BornylNH2 (dots
correspond to experimental data, lines are obtained by
simulation): (1, 6) 1:0, (2, 4) 1:10, (3, 5) 1:2.1; (1–3) silyl
ether, (4–6) acetophenone.
Fig. 4. Kinetic curves of the changes in the concentration
of acetophenone and accumulation of 1-phenylethanol silyl
ether in the (a) traditional reactor and (b) microreactor at
the varied ratio of [Rh(CO)2(μ-Cl)]2: MyrtNH2 (dots
correspond to experimental data, lines are obtained by
simulation): (1, 6) 1:0, (2, 4) 1:10, (3, 5) 1:2.1; (1–3) silyl
ether, (4–6) acetophenone.
microreactor (Fig. 2), but with other catalytic systems,
the reaction rate decreased with increasing con-
centration of amine both in the microreactor and in the
traditional reactor (Figs. 3–5).
of the absolute value of the so-called generalized
desirability index D.
According to multicriteria optimization, the best
complex for carrying the reaction in the presence of
amine both in the traditional reactor and in the micro-
reactor is [Rh(cod)Cl]2. The best results are achieved
in the microreactor in the presence of [Rh(cod)Cl]2 and
10-fold amount of MyrtNH2, D = 0.733 (Fig. 6), and
the same complex with BornylNH2 ranked second in
efficiency (the complex : amine ratio is 1 : 5, D =
0.719). On the contrary, in the traditional reactor at the
ratio of complex : amine = 1: 2.1 or 1: 10 the effect-
tiveness of this complex in the presence of BornylNH2
is higher (D = 0.646, and 0.658, respectively) than
with MyrtNH2 (D = 0.634–0.637).
The model obtained is quite adequate to allow us
the estimation of the optimum conditions of process,
including the area where the experimental data are
lacking. For this purpose, based on the principles of
multicriteria optimization and following the procedure
described in [31], we have formulated a generalized
criterion for finding the optimal mode of hydro-
silylation in the microreactor and the mixing reactor,
combining the maximum yield of target product and
the minimum yield of silyl enol ether, at a sufficiently
high value of acetophenone conversion. We also
included in the consideration the duration factor, the
structure and the ratio of metallocomplex and amine.
Optimal conditions were determined from the analysis
The optimal conditions of the reaction, obtained
from the analysis of simulation (an additional criterion
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

2270
UVAROV et al.
(a)
20
60
100
140
180
Reaction time, min
(b)
20
40
60
80
100
Fig. 6. Optimal regimes of hydrosilylation in the (1, 2, 10,
12) microreactor and (3–9) mixing reactors at the following
catalyst : amine ratio: (1, 5, 9, 10) 1:5, (2, 3, 6, 8, 11, 12)
1:10, (4, 7) 1:2.1; (1, 3, 4, 7, 8, 10) BornylNH2, (2, 5, 6, 9,
11, 12) MyrtNH2; (1–6) yield of 1-phenylethanol silyl
ether, (7–12) yield of silyl enol ether; D is generalized
maximum desirability index.
20
60
100
140
180
Reaction time, min
In the first case, the yield of silyl enol ether reaches
21%, in the second case 22%, while in the third case
18.0%.
Fig. 5. Kinetic curves of the changes in the concentration
of acetophenone and accumulation of 1-phenylethanol silyl
ether in the (a) traditional reactor and (b) microreactor at
the varied ratio of [Rh(CO)2(μ-Cl)]2: BornylNH2 (points
correspond to experimental data, lines are obtained by
simulation): (1, 6) 1:0, (2, 4) 1:10, (3, 5) 1:2.1; (1–3) silyl
ether, (4–6) acetophenone.
Thus, depending on the technological and
economic purposes, the application of multicriteria
optimization allows the choice between the different
optimal modes of hydrosilylation, advancing to the
first place one or another criterion.
From the results of multicriteria optimization it can
be concluded that the use of microreactors allows
reaching higher reaction rates of hydrosilylation and,
consequently, higher yields of the desired product,
maintaining cost-effective level of selectivity.
is the maximum reaction rate in the first section of the
kinetic curve) depend on the structure of the ligands
and the definite technological task. The best results in
the microreactor with MyrtNH2 (see the ratio of
[Rh(cod)Cl]2 : amine = 1 : 10) is achieved at the
duration 55 min (Fig. 6): yield of 1-phenylethyl silyl
ether is 68%, and of the byproduct, 7%. In turn, use of
bornylamine (ratio [Rh(cod)Cl]2: BornylNH2 = 1 : 5),
at duration 55 min provides a higher yield 1-
phenylethanol silyl ether (75%) with more than 2 times
greater amount of by-product (16%), which adversely
affects the value of the generalized criterion, lowering
it from the D = 0.733 to D = 0.719.
The mechanism of catalytic action of rhodium
complexes. To identify the structure of catalysts used
for the hydrosilylation, we investigated interaction of
rhodium complexes with amines in C6D6 using NMR
spectroscopy. The interaction of rhodium(I) complexes
with amines is very fast, and after 5 min in the 1H and
13C 2D NMR spectra were observed the characteristic
signals of NCH2 and NCH groups of coordinated
amines [δH 2.95 and 3.42 ppm, and δC 52.2 m (JRhC
=
In the traditional mixing reactor the yield of the
desired product with the best catalytic system is 54%
after 25 min at a ratio of [Rh(cod)Cl]2 : BornylNH2 equal
to 1 : 10, 52% after 25 min at a ratio of [Rh(cod)Cl]2 :
BornylNH2 equal to 1: 2.1, and 49% after 65 min at the
ratio of [Rh(cod)Cl]2 : MyrtNH2, equal to 1 : 5 (Fig. 6).
35.2 Hz) and 61.5 m (JRhC = 33.1 Hz) ppm for
MyrtNH2 and BornylNH2 respectively]. In addition,
the 13C NMR spectrum of the carbonyl complex
contains two doublets differing by intensity at δC 180–
185 ppm (JRhC =65.2 and 73.8 Hz, and JRhC = 65.8 and
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

EXPERIMENTAL STUDY AND SIMULATION OF KINETICS
2271
74.2 Hz for MyrtNH2 and BornylNH2 respectively),
corresponding to two asymmetric CO groups of the
new complex, cis-[Rh(CO)2(RNH2)Cl]. This allows us
a suggestion that the first macrokinetic stage is the
interaction of binuclear rhodium complexes with
amine leading to the mononuclear complexes contain-
ing the coordinated amine:
The modern schemes of the mechanism of Rh-
catalyzed hydrosilylation of ketones are based on the
mechanism proposed by Ogima [29]. This mechanism
involves oxidative addition of diphenylsilane to the Rh(I)
complex followed by the coordination of acetophenone
and migration of silyl group to the oxygen atom of the
coordinated acetophenone, resulting in the formation
of a Rh–C bond. The final stage is the reductive
elimination of the 1-phenylethanol silyl ether and
regeneration of the catalyst.
1/2[Rh(LL')(μ-Cl)] 2 + RNH2 → [Rh(LL')(RNH2)Cl],
L = L '= CO; LL' = cod; R = mirtanyl or bornyl.
Ph
CH−O−SiPh2H
[Rh]
H3C
Ph2SiH2
H3C
Ph
Rh−H
O
[H−Rh−SiPh2H]
SiPh2H
Ph
Ph
H3C−C
Rh
H3C−C
O
O
H−
SiPh2H
Our corrections of this scheme relate primarily to
the mechanism of formation of the enol silyl ether and
establishing existence of the stage of enantiomeric
differentiation. For example, Reyes et al. [28] suggest
that the key stage of formation of the enol silyl ether is
the β-elimination of proton in the siloxyalkyl
intermediate. As to the enantiodifferentiation, it was
shown in [30] that the ketone interacts with the
coordinated silicon hydride, while the coordination of
ketone to the rhodium atom does not occur.
The level of the stable activity of the complexes
responsible for the formation of the desired product
can be estimated from the selectivity. For example, in
the case of the catalytic system [Rh(CO)2(μ-Cl)]2–
amine (including variation of their ratio in the
traditional reactor), the selectivity falls to the range of
25 to 65%. In the microreactor the upper limit of the
range can be increased to 73%, but the lower limit falls
to 15%. The catalytic system based on [Rh(cod)Cl]2
shows a more narrow range of selectivity both in the
microreactor (60–90%), and in the traditional mixing
reactor (62–78%). Thus, in the latter case, the
selectivity of hydrosilylation in the traditional reactor
increases to 78%, while in the microreactor up to 90%.
Namely, it seems that in the microreactor the
stabilization by the amine of the intermediate
responsible for the formation of 1-phenylethanol silyl
ether upon increasing the concentration of the amine
occurs more effectively.
The formation 1-phenylethanol silyl ether proceeds
concurrently with the competing reaction generating
the corresponding enol as a byproduct, which is in
good agreement with the results of the kinetic-
mathematical simulation. The higher rate of the
process in the microreactor compared with the tradi-
tional reactor, the decrease in the rate of target reaction
upon increase in the concentration of amine in the
reaction medium, as well as the prevalence of the rate
of enol silyl ether formation over the target reaction in
the presence of [Rh(CO)2(μ-Cl)]2 and amines may be
partly due to the fact that micro-mixing facilitates the
ketone and amine entering into the coordination sphere
of rhodium atom and promotes the destruction of less
stable intermediates.
This all suggests that the coordination of ketone to
the rhodium atom does occur, in contrast to the
assumptions about the lack of such coordination [30].
Thus, a substantial positive effect at the use of
microstructural reactor (as compared with the
traditional mixing reactor) for the hydrosilylation in
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

2272
UVAROV et al.
the presence of rhodium catalyst is due to the
acceleration of all stages of the catalytic cycle and the
increased selectivity at the stage of transformation of
the siloxyalkyl intermediate.
the last portions was collected in a receiver cooled to
0°C. From the mixture was taken a sample for the
NMR investigation of the product conversion and the
reaction selectivity, and then the mixture was
hydrolyzed by adding 1 ml of methanol and, 1 h latter,
of 5 ml of 1N HCl. The products were extracted with
diethyl ether (2×20 ml), dried over Na2CO3, and
analyzed to estimate the enantiomeric composition as
in [15].
Thus, the experimental study of the reaction of the
acetophenone hydrosilylation with diphenylsilane in
the presence of rhodium complexes showed that this
process in most cases proceeds more efficiently in the
microreactor than in the traditional mixing reactor.
Maximum catalytic activity and selectivity are
observed in the presence of [Rh(CO)2(μ-Cl)]2 without
amines. Adding of amine to the complex in situ leads
to a decrease in activity of the catalytic system and
reaction selectivity, the best results were achieved with
the [Rh(cod)Cl]2–BornylNH2 system. Enantioselec-
tivity of hydrosilylation achieved in the microreactor is
lower than in the traditional reactor, and the
configuration of the formed isomer of 1-phenylethanol
remains unchanged.
In the experiments were used methanol, diethyl
ether, acetophenone, and THF of chemically pure
grade, (R)-(–)-cis-mitrylamine from Aldrich, (S)-(+)-α-
methoxy-α-trifluoromethylphenylacetyl chloride and
diphenylsilane from Acros, (R)-(+)-borhylamine, [Rh
(cod)Cl]2 and [Rh(CO)2(μ-Cl)]2 from Fluka, and silica
gel 60 from Merck.
ACKNOWLEDGMENTS
This work was performed at a financial support of
Russian Foundation for Basic Research (grant no. 09-
03-00341a), St. Petersburg Government (grant
MKH/15-05/005/), and the Program of DAAD
(Leonard-Euler-Stipendium).
The mathematical modeling of the process in a
microreactor and in traditional reactor at the assump-
tion of two parallel reaction is in good agreement with
experimental data and, thus, confirms the generally
accepted reaction mechanism of acetophenone
hydrosilylation with diphenylsilane. The multicriteria
optimization of the process allowed the choice of the
most effective catalytic system and optimum con-
ditions for its use depending on the specific
technological target.
REFERENCES
1. Ehrfeld, W., Hessel, V., and Löwe, H., Microreactors:
New Technology for Modern Chemistry, Weinheim:
Wiley–VCH, 2000.
2. Borovinskaya, E.S. and Reshetilovskii, V.P., Khim.
Prom–t’, 2008, no. 5, p. 1.
EXPERIMENTAL
3. Borovinskaya, E.S., Mammitzsch, L., Uvarov, V.M.,
Schael, F., and Reschetilowski, W., Chem. Eng.
Technol., 2009, vol. 32, no. 6, p. 919.
NMR spectra (1D, 2D, COSY, DEPT, HMBC,
HSQC) were registered on Bruker WМ-400 and DRX
500P instruments from solutions in CDCl3 at the
operating frequencies 400.13, 500.13 (1H), 125.76
(13C), and 99.35 (29Si) MHz. The measurements were
performed without additional reference compounds,
the frequency was taken relatively to the signal of the
deuterium-containing solvent.
4. Hessel, V., Löwe, H., Müller, A., and Kolb, G.,
Chemical Micro Process Engineering – Processing and
Plants, Weinheim: Wiley–VCH, 2005.
5. Haswell, S.J., Middleton, R.J., O’Sullivan, B., Skelton, V.,
Watts, P., and Styring, P., Chem. Commun., 2001, no. 5,
p. 391.
The acetophenone hydrosilylation with diphenyl-
silane was carried out in a mixing reactor as described
in [15], and in a microreactor according to the
following procedure. The diphenylsilane cooled to 0°C
(2.68 mmol) and a solution of 0.0215 mmol of a
complex, 0.0–0.215 mmol of amine and 2.15 mmol of
acetophenone in 0.25 ml of anhydrous THF cooled and
kept at 0°C for 10 min with stirring were fed to the
micro mixer at 0–5°C and then to the microreactor.
After 15 min, the reaction mixture except the first and
6. Fletcher, P.D.I., Haswell, S.J., Pombo-Villar, E.,
Warrington, B H., Watts, P., Wong, S.Y.F., and Zhang, X.,
Tetrahedron, 2002, vol. 58, no. 24, p. 4735.
7. Bellefon, C., Tanchoux, N., Caravieilhes, S., Gre-
nouillet, P., and Hessel, V., Angew. Chem., Int. Ed.,
2000, vol. 39, no. 19, p. 3442.
8. Jähnisch, K., Hessel, V., Löwe, H., and Baerns, M.,
Angew. Chem., Int. Ed., 2004, vol. 43, no. 4, p. 406.
9. Microreactors in Organic Synthesis and Catalysis,
Wirth, T., Ed., Weinheim: Wiley–VCH, 2008.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010

EXPERIMENTAL STUDY AND SIMULATION OF KINETICS
2273
10. Yu, Z.-X., Wender, P.A., and Houk, K.N., J. Am. Chem.
Soc., 2004, vol. 126, no. 30, p. 9154.
Applications of Asymmetric Synthesis, New York:
Wiley, 2001.
11. Yu, Z.-X., Cheong, P.H.-Y., Liu, P., Legault, C.Y.,
Wender, P.A., and Houk, K.N., J. Am. Chem. Soc.,
2008, vol. 130, no. 8, p. 2378.
20. Ojima, I. and Hirai, K., Asymmetric Synthesis, Moris-
son, J.D., Ed., New York: Academic Press, 1985, vol. 5.
21. Rollmann, L.D., Inorg. Chim. Acta, 1972, vol. 6, no. 1,
12. Helmchen, G., Krotz, A., Ganz, K.-T., and Hansen, D.,
p. 137.
Synlett., 1991, no. 4, p. 257.
22. Pavlov, V.A., Usp. Khim., 2001, vol. 70, no. 12, p. 1175.
13. Marciniec, B., Maciejewski, H., Pietraszuk, C., and
Pawluć, P., Hydrosilylation. A Comprehensive Review
on Recent Advances, Advances in Silicon Science,
Marciniec, B., Ed., Springer, 2009, vol. 1.
14. de Vekki, D.A., Uvarov, V.M., Reznikov, A.N., and
Skvortsov, N.K., Izv. Akad. Nauk, Ser. Khim., 2008,
vol. 57, no. 2, p. 594.
15. de Vekki, D.A., Uvarov, V.M., Skvortsov, N.K., and
Reshetilovskii, V.P., Zh. Obshch. Khim., 2010, vol. 80,
no. 1, p. 39.
16. Brunner, H., Störiko, R., and Nuber, B., Tetrahedron:
Asymmetry, 1998, vol. 9, no. 3, p. 407.
23. Brunner, H., Becker, R., and Riepl, G., Organo-
metallics, 1984, vol. 3, no. 9, p. 1354.
24. Brunner, H. and Kürzinger, A., J. Organomet. Chem.,
1988, vol. 346, no. 3, p. 413.
25. Brunner, H. and Obermann, U., Chem. Ber., 1989, vol. 122,
p. 499.
26. Brown-Wensley, K.A., Organometallics, 1987, vol. 6,
no. 7, p. 1590.
27. Borovinskaya, E.S., Candidate Sci. (Chem.) Disserta-
tion, St. Petersburg, 2008.
28. Reyes, C., Prock, A., and Giering, W.P., Organo-
metallics, 2002, vol. 21, no. 3, p. 546.
17. Chelucci, G., Gladiali, S., Sanna, M.G., and Brunner, H.,
Tetrahedron: Asymmetry, 2000, vol. 11, no. 16, p. 3419.
29. Ojima, I., Kogure, T., Kumagai, M., and Horiuchi, S.,
J. Organomet. Chem., 1976, vol. 122, no. 1, p. 83.
18. Odedra, A., Geyer, K., Gustafsson, T., Gilmour, R., and
Seeberger, P.H., Chem. Commun., 2008, no. 26, p. 3025.
30. Zheng, G.Z. and Chan, N.H., Organometallics, 1995,
vol. 14, no. 1, p. 70.
19. Lin, G-Q., Li, Y-M., and, Chan, A.S.C., Principles and
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 80 No. 11 2010