Electrochemical activation of diorganyl dialkoxysilanes for siloxane backbone extension

Article abstract of DOI:10.1039/b300620b

The reaction of diorganyl dialkoxysilanes PhRSi(OAlk)₂ (R = Ph, vinyl, OMe; Alk = Me, Et) with electrochemically reduced forms of oxygen provides reactive intermediates that insert into hexamethyldisiloxane or permethyl cyclosiloxanes, D_3<

Full text of DOI:10.1039/b300620b

View Article Online / Journal Homepage / Table of Contents for this issue
Electrochemical activation of diorganyl dialkoxysilanes for siloxane
backbone extension
Robert Keyrouz and Viatcheslav Jouikov*
Laboratory of Molecular and Macromolecular Electrochemistry (CNRS UMR 6510),
University of Rennes I, 35042, Rennes, France. E-mail: vjouikov@univ-rennes1.fr;
Fax: +33 2 23 23 62 92; Tel: +33 2 23 23 62 93
L e t t e r
Received (in Montpellier, France) 14th January 2003, Accepted 31st March 2003
First published as an Advance Article on the web 2nd May 2003
The conditions and yields of the electrolyses are summarized
in Table 1. Siloxane products supposedly result from cyclooli-
=
gomerization of electrogenerated transient silanone, Ph2Si O.
The insertion of the latter into Si–O bonds of the starting
Ph2Si(OMe)2 or of primarily formed siloxanes, one of the typi-
cal reactions of silanones,2 might account for the formation of
the higher siloxanes.
The reaction of diorganyl dialkoxysilanes PhRSi(OAlk)2
(R ¼ Ph, vinyl, OMe; Alk ¼ Me, Et) with electrochemically
reduced forms of oxygen provides reactive intermediates that
insert into hexamethyldisiloxane or permethyl cyclosiloxanes,
D3 and D4 , to give siloxane products with one additional
PhRSiO fragment; the process is thought to occur with the inter-
mediacy of diorganylsilanones.
Cathodic generation of diphenylsilanone from Ph2Si(OMe)2
was attempted using hexamethyldisiloxane (HMDS) as a sila-
none trap. Thus, the reduction of molecular oxygen in the
presence of Ph2Si(OMe)2 and HMDS in a 0.1 M solution of
Bu4NPF6 in DMF resulted in the formation of 1,1,1,5,5,5-
hexamethyl-3,3-diphenyltrisiloxane (3) as shown in Scheme 2.
When permethylcyclosiloxanes,13 Dn (n ¼ 3, 4), were used
Diorganyl silanones, silicon analogs of ketones, are thought to
be intermediates in many reactions of organic silicon com-
pounds.1,2 These species have never been observed in solution
at room temperature nor has their formation been unequivo-
cally proven, yet their transient formation was supposed in
thermal transformations of oxygen-containing silicon com-
pounds,3–6 in reactions of silenes with carbonyl compounds,7
in oxidation of silylenes by DMSO8 and epoxides9 and in the
oxidation of dichlorosilanes by O-donor reagents under rela-
tively mild conditions.2,10,11 We recently proposed an electro-
chemical method for the in situ generation of diorganyl
silanones from diorganyl dichlorosilanes.12 On the other hand,
although silicon chloro derivatives remain one of the most
common starting materials for various synthetic processes,
the use of easily available non-chlorinated organosilicon
reagents is among the utmost priorities in silicon chemistry.
Here, we report initial results of the use of diorganyl dialkoxy-
silanes as a source of an electrogenerated reagent, presumably
diorganyl silanone, allowing to selectively incorporate a
PhRSiO group into linear and cyclic siloxanes.
To check the hypothesis that diorganyl dialkoxysilanes
would react with nucleophilic forms of O2 , it was first shown
that they do not react with molecular oxygen itself under the
experimental conditions used: bubbling oxygen for 4 h through
a 10ꢀ2 mol Lꢀ1 solution of Ph2Si(OMe)2 in DMF containing
0.1 M Bu4NPF6 did not cause the formation of any new pro-
ducts. When the same solution was electrolyzed in a divided
three-electrode cell at the reduction potentials of O2 , where
Ph2Si(OMe)2 is electrochemically inactive, a mixture of silox-
ane products was formed (Scheme 1).
instead of HMDS, the corresponding cyclic products DnDPh
2
were obtained (Table 1). Some amount of linear insertion pro-
duct 3 was also found in the reaction mixture, probably due
to HMDS formed from Me3SiCl, which was usually added
to the solution, prior to the main electrolysis, in order to
remove traces of water. In principle, HMDS could also arise
from ring-chain equilibration14 of D3 or D4 , though such
processes usually need heating or a catalyst.
Similar results were obtained when using diphenyl diethoxy-
silane as the silanone precursor: in the experiment with no
trap, the product distribution was as in the case of
Ph2Si(OMe)2 (Table 1, run 1), whilst the addition of HMDS
gave rise to the linear insertion product 3. In general, the
product distribution with diorganyl dialkoxysilanes is much
narrower compared to that observed when diorganyl dichloro-
silanes are used as silanone precursors12 and less by-products
were formed.
As seen from Table 1, cyclic siloxanes diverge during the elec-
trolysis15 giving rise, under the usual conditions, to multiple
products so that the yield of each is generally not very high.
However, with HMDS, forming but a single product, the iso-
lated physical yield (taking into account the 60% conversion
The main product, hexaphenylcyclotrisiloxane (1), was
obtained with an 18% isolated yield [there was also 31% of
unreacted Ph2Si(OMe)2 left] along with octaphenyltetracyclo-
siloxane (2) and higher diphenylsiloxanes of linear structure
Scheme 1
Scheme 2
902
New J. Chem., 2003, 27, 902–904
DOI: 10.1039/b300620b
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
Table 1 Electrochemically induced extensiona of the siloxane backbone using diorganyl dialkoxysilanes PhRSi(OAlk)2
Run
PhRSi(OAlk)2
Ph2Si(OMe)2
Silanone trap
Q/F molꢀ1 b
Products (% yield)c
(Ph2SiO)3 (1, 18%), (Ph2SiO)4 (2, 2%), MDPh M (3, 5%)
d
2
1
2
3
4
5
6
7
8
9
No trap
HMDS
D3
1.9
1.7
2.1
2.2
2.0
2.0
1.8
1.9
2.0
3 (46%)
D3DPh (4, 24%), D4DPh (5, 5.3%), 3 (6.1%)
4 (21%), 5 (4.6%), 3 (2.8%)
5 (29%), 4 (13%)e
2
2
D4
D4
Ph2Si(OEt)2
No trap
HMDS
D3
1 (23%), 2 (5%), 3 (5.2%)
3 (43%)
D3DPh,MeO (6, 26%), D4DPh,MeO (7, 3.8%)
D3DPh,Vin (8, 18%)
PhSi(OMe)3
PhVinSi(OMe)2
D3
a
b
In DMF + 0.1 M Bu4NPF6 , T ¼ 45 ꢁC. Related to the total amount of dialkoxysilane taken, typical conversion of PhRSi(OAlk)2 was a ¼
60–75%. c Non-optimized isolated yields after column chromatography on SiO2 . d HMDS resulted from the hydrolysis of Me3SiCl. e T ¼ ꢀ10ꢁC.
of the starting Ph2Si(OEt)2) amounts to 71%. When using cyclic
traps for the silanones, there are two points worth addressing.
First, there is a remarkable formation of polysiloxanes in the
process compared to the experiments with HMDS as the trap
(Table 1), and second, products resulting from siloxanes other
than the starting siloxane (e.g., D3DPhR products from D4 and
D4DPhR products from D3) are obtained. Ring disproportiona-
tion (usually promoted by high concentrations and tempera-
ture14) might account for these phenomena. Since the
concentration of permethyl cyclosiloxane in the electrolyte is
not high enough to favor this process considerably, we ran
the electrolysis at a lower temperature.16 Now, the reduction
of O2 in the presence of Ph2Si(OMe)2 and D4 , carried out at
ꢀ10 ꢁC, shows a more correct product distribution (Table 1,
runs 4 and 5). The product yields were higher, the amount of
concentration is very low and the process occurs with a large
bulk excess of HMDS practically until the end of the electro-
lysis. This observation allowed the relative efficiency of diorga-
nyl dialkoxysilanes versus diorganyl dichlorosilanes for the
electrogeneration of diorganyl silanones to be evaluated. Since
silanones were never observed in solution, their rate of con-
sumption must be faster than their rate of formation. This
seems to be true, whatever is the substituent at Si, at least
for ordinary substituents like Me, vinyl (Vin), Et, Ph.12 As
no autocyclization products were obtained in the electrolyses,
one can assume this is also true 0in the competitive kinetic
0
scheme (Scheme 4), that is k1 ꢂ k1 and k2 ꢂ k2 .
From the ratio k1/k2 ¼ Y1/Y2 it follows that k1 ¼
k2(Y1/Y2) with k1 ¼ kOMe ¼ kPh ꢃ kOMeand k2 ¼ kCl
¼
2
Ph2
PhVi
kPhVi ꢃ kCl.
higher siloxanes was smaller and the ratio of D4DPh to
Therefore, assuming k1 and k2 to be limiting steps and taking
an excess of HMDS, one can estimate the relative rates of reac-
tion of two silanone precursors with reduced oxygen by the
ratio of the yields (Y1 and Y2) of silanone insertion products
in two competitive pathways [given that the mechanism of
2
D3DPh was 2:1 instead of 1:3 under the usual conditions. The
2
process is therefore much more selective at low temperature.
To check that the ‘‘silanone path’’ is the only one leading to
the PhRSiO insertion products and to rule out other possible
reaction pathways, a series of experiments was carried out.
First, after stirring Ph2Si(OMe)2 and D3 in a solution of 0.1
M Bu4NPF6 in DMF for 4 h (a typical duration of the electro-
lysis), no new products were observed. The same results were
observed in the electroreduction of O2 in the presence of either
HMDS or D3 and by saturating the solution of Ph2Si(OMe)2
and D3 in DMF + 0.1 M Bu4NPF6 with oxygen and stirring
it for 4 h without passing the current.
The presence of two alkoxy groups is essential for the pro-
cess since no products were obtained when using monoalk-
oxysilanes. Methoxytrimethylsilane only served as a silanone
trap, reacting with diphenylsilanone to give Me3SiOSi
(Ph)2OMe. On the other hand, when organyl trialkoxysilanes
were used, RSi(OAlk)3 , only two OAlk groups reacted to form
the intermediate silanone (Scheme 3) whereas the third one was
incorporated and conserved in the ring extension product.
Given that in the absence of an appropriate silanone trap
diorganylsilanones tend to undergo autocyclization,2 the rate
of this process is apparently slower than that of silanone inser-
tion into the Si–O bond of HMDS or that of permethylcyclo-
siloxanes, since practically no perphenylated siloxane products
were observed in the electrolyses with these silanone traps.
This fact can have a purely kinetic reason. Indeed, though
the starting ratio of Ph2Si(OMe)2 to HMDS was in favor of
the dimethoxy compound (10:4), the conditions of electro-
the process is similar in both cases, that CPh Si(OMe)2
CPhVinSiCl and both concentrations are much higher than
¼
2
2
CO 2ꢀ , and that the conversion is low, a ꢄ 5%17]. Keeping in
2
mind that the apparent rate constants have contributions kR
from the electronic and steric properties of the organyl groups
on Si on the one hand, and on the other from the leaving
groups, kOMe and kCl, these terms were estimated in two series
of experiments. In the reaction series with similar leaving
groups, Ph2SiCl2/PhVinSiCl2 and Ph2Si(OMe)2/PhVin-
Si(OMe)2 , the ratio of constants is the same, kPh /kPhVin
¼
2
0.31, indicating that the organyl groups at Si affect the reactiv-
ity of diorganyl dimethoxy and diorganyl dichlorosilanes in the
same way. Then the cross experiments with the couples
Ph2Si(OMe)2/PhVinSiCl2 and Ph2SiCl2/PhVinSi(OMe)2 gave
the ratios k1/k2 ¼ 1/5.8 and 1/1.794, respectively. Taking into
account that kPhVin ¼ 3.22 ꢃ kPh and thus correcting the data
2
for the replacement of Ph with Vin, one obtains the ratio
kOMe ¼ 0.556 ꢃ kCl. Thus, the diorganyl dialkoxysilanes are
about two times less reactive than diorganyl dichlorosilanes
and this is presumably the reason why the process in the first
case is more selective (cf.ref. 12). Studies aiming to reveal the
exact nature of the electrogenerated active intermediate in
the process and to track the transformations of the eliminated
alkoxy groups are in progress.
=
generation of Ph2Si O are such that its instantaneous
Scheme 3
Scheme 4
New J. Chem., 2003, 27, 902–904
903

View Article Online
(SiCH3), 52.88 (OCH3), 127.71 (C3 , C5 of Ph), 128.64 (C4),
135.6 (C2 , C6), 141.5 (C1); MS m/z: 374 (M+), 359
(M+ ꢀ CH3), 343, 207, 151, 89, 31.
Experimental
Typical experimental procedure for the electrolyses was as fol-
lows. A solution of 0.1 mmol of Me3SiCl and 2.5 ꢃ 10ꢀ3 mol
Bu4NPF6 in 25 ml of dry DMF was electrolyzed for 1.5 h to
reduce H+ formed in the hydrolysis of the chlorosilane by resi-
dual water. Diorganyl dialkoxysilane (5 mmol) was then added
(and, when necessary, 2 mmol of silanone trap) and the elec-
1
D4DPhMeO (7). NMR H: 0.08 (12H, s), 0.11 (12H, s), 3.50
(3H, s), 7.14–7.71 (5H, m); NMR 13C (CDCl3): 0.80 (SiCH3),
1.72 (SiCH3), 52.80 (OCH3), 127.1 (C3 , C5), 128.5 (C4),
134.2 (C2 , C6), 141.1 (C1); MS m/z: 433 (M+ ꢀ CH3), 371
(M+ ꢀ C6H5), 207, 181, 137, 73.
trolysis was carried out at a constant O2 flow rate (5 ml minꢀ1
)
in the galvanostatic mode with a current density j ¼ 5–7 mA
cmꢀ2. Graphite cathode and glassy carbon felt anode were
used. After passing about 2 F of electricity, Bu4NPF6 was
precipitated with hexanes, the solution concentrated and
separated on a column with SiO2 .
D3DPhVin (8). NMR 1H: (CDCl3) 0.14 (6H, s), 0.18 (12H, s),
5.82–6.32 (3H, m), 7.36–7.67 (5H, m); MS m/z: 370 (M+), 355
(M+ ꢀ CH3), 251, 207, 73.
Characterization of products
References
MS and NMR spectra were recorded using a Shimadzu QP-
5000 quadrupole spectrometer, ionization voltage 70 kV, and
a Bruker 200 MHz spectrometer (CDCl3 , TMS), respectively.
1
L. E. Gusel’nikov, N. S. Nametkin and V. M. Vdovin, Acc. Chem.
Res., 1975, 8, 18.
2
3
4
M. G. Voronkov, J. Organomet. Chem., 1998, 557, 143.
H. J. Schnockel, J. Mol. Struct., 1980, 65, 115.
J. M. T. Davidson, A. Fenton, G. Manuel and G. Bertrand, Orga-
nometallics, 1985, 4, 1324.
(Ph2SiO)3 (1). NMR 1H: 7.20–7.24 (6H, m), 7.42–7.46 (12H,
m), 7.96–7.98 (12H, m); NMR 13C (CDCl3): 125.91 (C3 , C5),
128.99 (C4), 137.11 (C2 , C6), 144.21 (C1); MS m/z: 594
(M+), 439, 219, 197, 144, 131, 77.
5
E. A. Chernyshev, T. L. Krasnova, N. A. Mudrova, A. V.
Golovkin and M. G. Kuznetsova, Zh. Obshch. Khim., 1987, 57,
1725.
(Ph2SiO)4 (2). NMR 1H: 7.19–7.24 (8H, m), 7.39–7.43 (16H,
m), 7.90–7.92 (16H, m); NMR 13C (CDCl3): 125.21 (C3 , C5),
129.21 (C4), 138.01 (C2 , C6), 145.41 (C1).
6
7
T. J. Barton and B. L. Groh, J. Am. Chem. Soc., 1985, 107, 7221.
L. E. Gusel’nikov, Z. A. Kerzina, Y. P. Polyakov and N. S.
Nametkin, Zh. Obsch. Khim., 1982, 52, 467.
H. S. D. Soysa, H. Okinoshima and W. P. Weber, J. Organomet.
Chem., 1977, 133, C17.
8
9
1
MDPh M (3). NMR H (CDCl3): 0.104 (18H, s), 7.30–7.66
W. F. Goure and T. J. Barton, J. Organomet. Chem., 1980,
199, 33.
2
(10H, m); NMR 13C (CDCl3) 2.47 (SiCH3), 126.5 (C3),
130(C4), 135.4 (C2), 142.2 (C1); MS m/z: 360 (M+), 345
(M+ ꢀ CH3), 330, 267, 197, 73.
10 M. G. Voronkov and S. V. Basenko, J. Organomet. Chem., 1995,
500, 325.
11 M. G. Voronkov, I. P. Tsyrendorzhieva, N. P. Ivanova and E. I.
Dubinskaya, Zh. Obshch. Khim., 1998, 68, 699.
12 D. S. Fattakhova, V. V. Jouikov and M. G. Voronkov, J. Orga-
nomet. Chem., 2000, 613, 170.
13 Here and thereafter in this paper, we used General Electric
Siloxane Nomenclature which depicts a Me2SiO unit as D and
reserves the symbol DX for structural units analogous to Me2SiO
but carrying substituents other than Me. Thus, cyclic products
derived from Dn and incorporating one or two Ph groups are
1
D3DPh (4). NMR H (CDCl3): 0.09 (12H, s), 0.11 (6H, s),
2
7.36–7.69 (10H, m); NMR 13C (CDCl3): 0.80 (SiCH3), 1.32
(SiCH3), 126.4 (C3 and C5 of Ph), 129.4 (C4 of Ph), 135.4
(C2 and C6 of Ph), 143.6 (C1 Ph); MS/EI m/z: 420 (M+),
405 (M+ ꢀ CH3), 207, 197, 135, 73.
1
D4DPh (5). NMR H (CDCl3): 0.086 (12H, s), 0.136 (12H,
2
depicted as DnDPh or DnDPh , respectively.
2
s), 7.30–7.66 (10H, m); NMR 13C (CDCl3): 0.80 (SiCH3),
1.32 (SiCH3), 126.4 (C3 , C5), 129.3 (C4), 134.6 (C2 , C6),
143.1 (C1); MS m/z: 479 (M+ ꢀ CH3), 417 (M+ ꢀ C6H5),
340, 207, 197, 137, 73.
14 P. V. Wright and J. A. Semlyen, Polymer, 1970, 11, 462.
15 V. Jouikov and R. Keyrouz, Silicon Chem., in the press.
16 The most resistive part of the electrolytic setup is the compart-
ment separating diaphragm so the Joule’s heat (W ¼ I2 ꢃ R) is
mostly produced there. Usually, with no special cooling, the bulk
electrolyte temperature was about 45 ꢁC.
1
D3DPhMeO (6). NMR H: 0.08 (12H, s), 0.12 (6H, s), 3.55
17 V. V. Jouikov and L. A. Grigorieva, Electrochim. Acta, 1996, 15,
2489.
(3H, s), 7.14–7.78 (5H, m); NMR 13C: 0.80 (SiCH3), 1.81
904
New J. Chem., 2003, 27, 902–904