From inconsistent results to high speed hydrosilylation

Article abstract of DOI:10.1039/b617413b

After noting that the presence of dihydrogen, generated in situ from the partial hydrolysis of a silane with residual water, significantly enhances the rate of the rhodium-catalysed hydrosilylation of acetophenone, we developed a high speed hydrosilylation reaction under dihydrogen pressure. The Royal Society of Chemistry.

Full text of DOI:10.1039/b617413b

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
From inconsistent results to high speed hydrosilylation
Virginie Comte,* Ce´dric Balan, Pierre Le Gendre* and Claude Mo¨ıse
Received (in Cambridge, UK) 29th November 2006, Accepted 19th December 2006
First published as an Advance Article on the web 10th January 2007
DOI: 10.1039/b617413b
After noting that the presence of dihydrogen, generated in situ
from the partial hydrolysis of a silane with residual water,
significantly enhances the rate of the rhodium-catalysed
hydrosilylation of acetophenone, we developed a high speed
hydrosilylation reaction under dihydrogen pressure.
Rhodium complexes associated with chiral diphosphine ligands
Scheme 1 Hydrosilylation of acetophenone under dihydrogen pressure.
have been widely applied for the asymmetric hydrosilylation of
carbonyl compounds.1 Despite the large number of papers in this
area, only a few deal with catalysis in high enantiomeric excess
(ee).2,3 Examination of the literature revealed that high optical
yields are generally reached when the reaction is conducted at low
temperature. Indeed, efficient precatalysts have been shown to
promote the reduction in the temperature range 0 to 260 uC with
reasonable loadings and reaction times.4 This can be detrimental to
the development of an efficient asymmetric process, since the latter
requires a good catalyst activity. It can thus be assumed that most
of the diphosphine ligands tested to date have not yet lived up to
their full potential. In line with this conclusion, we decided to
focus first on catalytic activity and revisit this reaction using an
achiral diphosphine ligand. We first examined the reaction
of acetophenone with diphenylsilane in the presence of
[(o-dppbe)Rh(COD)]OTf.5 Serendipitously we have found that
water markedly enhances the rate of the reaction. The determina-
tion of its exact role allowed us to tune the reaction conditions in
order to obtain high speed hydrosilylation. The scope of this
reaction with several aromatic and aliphatic ketones, along with a
preliminary experiment using optimised conditions and a rhodium
complex associated with a chiral diphosphine (Deguphos),6 is
discussed in this contribution.
worth mentioning that the addition of diphenylsilane to the ‘‘wet’’
mixture was accompanied by vigorous bubbling. It is reasonable to
assume that this gas evolution results from the partial hydrolysis of
the silane, leading to the formation of H2 and siloxane
(Ph2SiH)2O.7 Therefore the higher conversion might be due to
water, siloxane or H2. The hydrosilylation reactions carried out in
the presence of a sub-stoichiometric amount of (Ph2SiH)2O
showed no improvement in conversion. On the contrary, runs
performed under an optimised pressure of 20 bar H2 led
consistently to phenylethanol in 98% yield after only 1 h
(Scheme 1).
This rate enhancement in the hydrosilylation of acetophenone
was further confirmed by comparing the kinetics of the reaction
under normal and optimised conditions (Fig. 1). The initial
turnover frequency (TOF) was determined to be 1700 h21 after
30 min at 20 bar H2, whereas the same catalyst under N2 gave only
a TOF of 40 h21. This value is comparable to those found for the
most active rhodium catalysts described so far in the literature.2,3,8
Interestingly, we found that the yield dropped upon using both
lower and higher dihydrogen pressures (the hydrosilylation of
acetophenone performed at 10 and 25 bar H2 afforded
phenylethanol in 90% and 84% yields after 1 h, respectively).
The hydrosilylation of acetophenone with diphenylsilane was
achieved in THF at 25 uC using 0.1 mol% [(o-dppbe)Rh-
(COD)]OTf. The reaction was stopped after 5 h and the
conversion determined by 1H NMR analysis of the crude product
obtained after desilylation. Using such conditions, hydrosilylation
readily proceeded, but the yields were found to be low and
irreproducible (5–20%). Envisaging that adventitious traces of
water could be responsible for the erratic yields, we carried out
experiments in which various quantities of water were added. To
our surprise, we found that small amounts of water had a
favourable effect on the activity of the catalyst. Indeed, while only
5% of phenylethanol was obtained under dry conditions (i.e. flame
dried vessel, dry THF), an optimum conversion of 56% was
achieved in the presence of 10 mol% water. At this point, it is
Laboratoire de Synthe`se et Electrosynthe`se Organome´talliques LSEO-
UMR 5188, Faculte´ des Sciences Mirande, Dijon, France.
E-mail: virginie.comte@u-bourgogne.fr;
pierre.le-gendre@u-bourgogne.fr; Fax: +33 3 8039 6098;
Tel: +33 3 8039 9048 Tel: +33 3 8039 6082
Fig. 1 Conversion into phenylethanol vs. time (X = reaction under 20
bar H2, m = reaction under N2 atmosphere).
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 713–715 | 713

View Article Online
Moreover, we noticed that the dihydrogen pressure remained
constant throughout the reaction, giving evidence that dihydrogen
is not consumed. This data is in accordance with the fact that the
rhodium complex [(o-dppbe)Rh(COD)]OTf does not catalyse the
hydrogenation of acetophenone under these conditions. Finally, it
is worth mentioning that a constant dihydrogen pressure was
required to maintain the high catalytic activity of the rhodium
complex. Indeed, a run partially conducted under dihydrogen
pressure for 15 min (20 bar H2), depressurized to 1 bar and finally
stopped after 1 h afforded phenylethanol in a moderate yield
(50%). Nevertheless, depressurization did not ‘‘shut down’’
definitively the catalytic activity of the rhodium centre. Thus, a
run conducted successively at 20 bar (15 min), 1 bar (30 min) and
20 bar (15 min) led to phenylethanol in good yield (90%). One
might claim that the rate enhancement was due to the pressure and
not to dihydrogen itself. However, this possibility is excluded by
the fact that the catalytic hydrosilylation of acetophenone under
20 bar N2 led to a very poor yield.
generally accepted Rh(III) intermediates can undergo a second
oxidative addition of another molecule of silane, facilitating the
final reductive elimination step of the product. Based on these
previously reported mechanistic studies, we propose that dihydro-
gen replaces here the second molecule of silane in the generation of
the oxidative Rh(V) adducts. These complexes would then exist in
solution, but only if the local concentration of dihydrogen was
sufficient. This hypothesis explains all of the experimental results
1
described above. Further experiments, notably H NMR spectro-
scopy under dihydrogen pressure, are in progress aiming to
provide support for this hypothesis.
Using our experimental protocol (i.e. under 20 bar H2), we
hydrosilylated a range of aromatic and aliphatic ketones, the
results of which are presented in Table 1. For comparative
purposes, hydrosilylation reactions under an inert atmosphere
were carried out in parallel and quenched after the same reaction
1
time. Yields were determined by H NMR spectroscopy of the
crude products after desilylation. The hydrosilylation of propio-
phenone, a-tetralone and pinacolone provided the corresponding
alcohols in almost quantitative yields after only 1 h. In contrast,
the same catalyst under N2 gave the products in much lower yields
(5, 7 and 4% yields, respectively). Although the hydrosilylation of
cyclohexanone and acetone led to moderate yields (46 and 32%,
respectively), these reactions were still, respectively, 11 and 16 times
faster compared to those performed under an inert atmosphere.
When the rhodium complex [(dppe)Rh(COD)]OTf was used as the
catalyst precursor, a significant rate enhancement under dihydro-
gen pressure was also observed (Table 1, results in brackets).14 It is
noteworthy that this structurally comparable complex shows
different efficiencies depending on the ketone used. While results
were comparable with acetophenone, a-tetralone and pinacolone,
a lower yield was obtained with propiophenone as the substrate.
On the other hand, the rhodium complex with dppe as the ligand
was found to be more efficient for the hydrosilylation of
cyclohexanone and acetone.
According to the literature, hydrosilylation is often accompanied
by the undesired dehydrogenative silylation of the enol of
acetophenone (silyl enol ether).1,9 In the subsequent hydrolysis
step, the silyl enol ether reverts to the original acetophenone.
However, this side reaction, which represents a major drawback,
would have no impact on the yield of the reaction if a subsequent
catalytic hydrogenation of the silyl enol ether occurred. This would
also rationalize why dihydrogen is, overall, not consumed, since
silylation of the enol is accompanied by the evolution of molecular
1
hydrogen. To check this hypothesis, we measured the H NMR
spectrum of the crude product before hydrolytic work-up. The
spectrum of the product obtained after a hydrosilylation reaction
run under ‘‘normal’’ conditions (i.e. under an inert atmosphere)
showed signals corresponding to the hydrosilylated product (silyl
ether) together with those of residual acetophenone. No trace of
silyl enol ether was detected. According to the NMR spectrum, the
hydrosilylation of acetophenone under 20 bar H2 also led
exclusively to the hydrosilylated product. Moreover, attempts to
hydrogenate the trimethylsilyl enol ether derived from acetophe-
none in the presence of a catalytic amount of [(o-dppbe)Rh-
(COD)]OTf remained unsuccessful. These results, combined with
those discussed above, strongly suggest that dihydrogen is not the
stoichiometric reducing agent.
Since the hydrosilylation of ketones can be highly accelerated by
using different ligands, we envisioned taking advantage of this
phenomenon for an asymmetric version of this reaction.
Preliminary experiments were performed with [((R,R)-
Deguphos)Rh(COD)]OTf as the chiral catalyst and acetophenone
as the substrate. A significant rate enhancement associated with
dihydrogen pressure was again observed. While only a 27% yield
was obtained after 5 h under an inert atmosphere, the reaction
went to completion after 1 h if run under 20 bar H2. Interestingly,
the reaction product (phenylethanol) was obtained with a higher ee
To elucidate the positive role of dihydrogen on the kinetics, we
examined the hydrosilylation of acetophenone at 20 bar H2 with
Ph2SiD2. Surprisingly, under these conditions, the silyl ether was
obtained with only 40% deuterium incorporation. Moreover, the
addition of H2 (20 bar) to a THF solution of Ph2SiD2 in the
presence of a catalytic amount of the rhodium complex resulted in
the partial conversion of Ph2SiD2 to Ph2SiD22nHn (n = 1 or 2,
product containing only 33% of deuterium after 1 h). These results
of H/D exchange strongly suggest that both H2 and Ph2SiD2
undergo an oxidative addition to the same metal centre to give a
cationic rhodium(V) adduct.10–12 Subsequent repeated sequences of
reductive elimination–oxidative addition would allow the conver-
sion of Ph2SiD2 to Ph2SiH2. Since this kind of isotopic exchange
occurs in the presence of the ketone, we suspect that Rh(V) species
would also be involved in the hydrosilylation process. This
hypothesis is supported by previous studies, suggesting that
mechanisms of hydrosilylation involving Rh(V) species are
reasonable.13 An important suggestion in these reports is that
Table 1 Catalytic hydrosilylation of various ketonesa
Yield (%)b
Entry
Substrate
Under 20 bar H2
Under N2
1
2
3
4
5
Acetophenone
Propiophenone
a-Tetralone
Pinacolone
Cyclohexanone
Acetone
98 (94)
95 (57)
92 (90)
99 (99)
46 (96)
32 (62)
4 (5)
5 (2)
2 (3)
4 (6)
4 (5)
2 (6)
6
a
Conditions:
[(o-dppbe)Rh(COD)]OTf
(0.1
mol%),
ketone
b
(2.2 mmol), Ph2SiH2 (2.4 mmol), THF (2 mL), rt, 1 h. The yields
in brackets correspond to reactions performed with [(dppe)Rh-
(COD)]OTf as the catalyst.
714 | Chem. Commun., 2007, 713–715
This journal is ß The Royal Society of Chemistry 2007

View Article Online
D. A. Evans, F. E. Michael, J. S. Tedrow and K. R. Campos, J. Am.
Chem. Soc., 2003, 125, 3534; B. Tao and G. C. Fu, Angew. Chem., Int.
Ed., 2002, 41, 3892; D. K. Heldmann and D. Seebach, Helv. Chim. Acta,
1999, 82, 1096; Y. Nishibayashi, K. Segawa, K. Ohe and S. Uemura,
Organometallics, 1995, 14, 5486; H. Nishiyama, M. Kondo,
T. Nakamura and K. Ito, Organometallics, 1991, 10, 500.
4 Lowering the temperature generally results in higher enantioselectivity
(see ref. 2 and ref. 3). Nevertheless, some exceptions have been reported:
C. G. Arena and R. Pattacini, J. Mol. Catal. A: Chem., 2004, 222, 47;
K. Kromm, P. L. Osburn and J. A. Gladysz, Organometallics, 2002, 21,
4275; D. Haag, J. Runsink and H. D. Scharf, Organometallics, 1998, 17,
398.
by carrying out the reaction under dihydrogen pressure (22% ee at
20 bar H2 vs. 11% ee under N2). Finally, the increased reactivity of
the catalyst under dihydrogen pressure allowed the hydrosilylation
to be run at a lower temperature, generally resulting in a higher
selectivity. Unfortunately, when the temperature was lowered to
240 uC, no influence on the enantioselectivity was seen. Despite
the modest level of enantioselectivity, this preliminary result
confirms that a sluggish catalyst can be converted into a highly
active species simply by conducting the hydrosilylation reaction
under a dihydrogen pressure.
5 o-dppbe = ortho-Bis(diphenylphosphino)benzene.
In summary, the activity of rhodium complexes associated with
common diphosphines such as dppbe or ddpe is several orders of
magnitude higher under dihydrogen pressure than under an inert
atmosphere. To the best of our knowledge, this phenomenon has
never been reported before. Further work aimed at elucidating the
mechanism of this acceleration and screening other chiral rhodium
catalysts is currently in progress.
6 Deguphos = 3,4-Bis(diphenylphosphino)-1-benzylpyrrolidine.
7 T. C. Bedard and J. Y. Corey, J. Organomet. Chem., 1992, 428, 315.
8 Highly active mono(phosphine)rhodium catalysts for the hydrosilylation
of ketones have been recently described: O. Niyomura, T. Iwasawa,
N. Sawada, M. Tokunaga, Y. Obora and Y. Tsuji, Organometallics,
2005, 21, 4275; O. Niyomura, M. Tokunaga, Y. Obora, T. Iwasawa and
Y. Tsuji, Angew. Chem., Int. Ed., 2003, 42, 1287.
9 C. Reyes, A. Prock and W. P. Giering, J. Organomet. Chem., 2003, 671,
13.
10 M. D. Curtis and P. S. Epstein, Adv. Organomet. Chem., 1981, 19, 213;
R. Goikhman and D. Milstein, Chem.–Eur. J., 2005, 11, 2983.
11 A H/D exchange from Et2SiH2 and D2 via a cationic Ir(V) complex has
been recently described: L. Turculet, J. D. Feldman and T. D. Tilley,
Organometallics, 2004, 23, 2488.
12 A mechanism analogous to the one described by Fryzuk, involving the
formation of dinuclear rhodium complexes, cannot be completely ruled
out. However, the low concentration of the rhodium complex in the
reaction mixture strongly supports a mechanism occurring at a rhodium
centre: M. D. Fryzuk, L. Rosenberg and S. J. Rettig, Organometallics,
1996, 15, 2871.
13 M. J. Fernandez, P. M. Bailev, P. O. Bentz, J. S. Ricci, T. F. Koetzle
and P. M. Maitlis, J. Am. Chem. Soc., 1984, 106, 5458; S. B. Duckett
and R. N. Perutz, Organometallics, 1992, 11, 90; H. Nagashima,
K. Tatebe, T. Ishibashi, A. Nakaoka, J. Sakakibara and K. Itoh,
Organometallics, 1995, 14, 2868.
Notes and references
1 O. Riant, N. Mostefa¨ı and J. Courmacel, Synthesis, 2004, 2943;
H. Brunner, in Transition Metals in Organic Synthesis, ed. M. Beller and
C. Bolm, Wiley-VCH, Weinheim, 1998, pp. 131–140; R. Noyori,
Asymmetric Catalysis in Organic Synthesis, Wiley, New York, 1994.
2 For the highly enantioselective rhodium-catalysed hydrosilylation of
ketones using diphosphine ligands, see: T. Imamoto, T. Itoh,
Y. Yamanoi, R. Narui and K. Yoshida, Tetrahedron: Asymmetry,
2006, 17, 560; T. Imamoto, N. Oohara and H. Takahashi, Synthesis,
2004, 1353; R. Kuwano, T. Uemura, M. Saitoh and Y. Ito,
Tetrahedron: Asymmetry, 2004, 15, 2263; Y. Yamanoi and
T. Imamoto, J. Org. Chem., 1999, 64, 2988; M. Sawamura,
R. Kuwano and Y. Ito, Angew. Chem., Int. Ed. Engl., 1994, 33, 111.
3 The most successful results have been obtained using other kinds of
chiral ligands. See, for example: V. Ce´sar, S. Bellemin-Laponnaz,
H. Wadepohl and L. H. Gade, Chem.–Eur. J., 2005, 11, 2862;
14 dppe = Bis(diphenylphosphino)ethane.
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 713–715 | 715