species is generated from aryl cyanides via the cleavage of
a carbon-carbon bond. Synthetic applications of this decy-
anative Mizoroki-Heck reaction, including orthogonal and
iterative functionalization of arenes, are also demonstrated.
Previously, we reported rhodium-catalyzed silylation6a,b
and decyanation6c reactions of aryl cyanides via the cleavage
of C-CN bonds. In these reactions, an arylrhodium species
generated from aryl cyanides and organosilicon reagents is
postulated to be a key intermediate. Indeed, we also reported
that tethered electrophiles can intercept the arylrhodium
intermediate to afford cyclized products.6b On the basis of
these results, we envisioned that an arylrhodium complex
generated in our catalytic system should react with external
alkenes to furnish an alkenylated product, especially in
consideration of the number of catalytic reactions involving
the addition of arylrhodium intermediates to alkenes.7 One
isolated example of a Mizoroki-Heck-type alkenylation of
nitrile (34% yield), in which a C-CN bond is cleaved via
oxidative addition to a nickel catalyst, was reported.8
First, we examined the reaction of methyl 4-cyanobenzoate
with several alkenes, including terminal alkenes, cyclic
alkenes, R,ꢀ-unsaturated carbonyl compounds, vinyl acetate,
vinyl ethers, and vinylsilanes, under the conditions we
identified as being suitable for the intramolecular arylation
reaction.6b As a result, the desired Mizoroki-Heck-type
product 2a was obtained in 46% yield, along with the
undesired silylated byproduct 3,6a,b when triethylvinylsilane
(1a) was used as an alkene component (entry 1 in Table 1).
ization of alkenes.9 Inspired by the promising result using
1a, we next optimized the reaction conditions to improve
the selectivity (Table 1). By increasing the amount of 1a (4
equiv to nitrile), the yield of alkenylated product was
improved (entry 2). The choice of the ligand also proved to
affect the selectivity significantly. Among the ligands
examined, P(4-FC6H4)3 proved to be the most effective.10
The effect of the substituents on the silicon atom of
vinylsilanes was also examined (entries 4-7). The use of
vinylsilanes bearing trimethyl-, tert-butyldimethyl-, and
triisopropylsilyl groups, as in 1b-d, lowered the yields of
the corresponding alkenylated products. On the other hand,
the use of triisopropoxylvinylsilane (1e) completely sup-
pressed the formation of 3 and selectively afforded the
alkenylated product 2e. Although 1e proved to be the best
alkene in terms of the yield and selectivity, we were unable
to isolate the pure product 2e due to the difficulty in
separating the byproducts, which were presumably formed
by the oligomerization of 1e. Thus, we decided to employ
1a as an alkene component for further investigation.11
Table 2 shows the scope of the rhodium-catalyzed alk-
enylation of nitriles using 1a. Functional groups, such as
esters (entries 1), ethers (entries 3, 5, 7, and 8), and amines
(entries 4 and 6), and a heteroaromatic ring (entry 12) were
tolerated. The selectivity between alkenylation and silylation
agreed with the trends of electronic and steric effects
observed for our previously reported silylation reaction.6c
Better selectivity for the alkenylation products was observed
with electron-rich substrates due to the relatively slow
silylation reaction (entries 3 and 4). It is important to note
that, although a small amount of silylated products were
formed with several substrates, all the alkenylated products
could successfully be isolated in pure form by standard silica
gel chromatography. In the case of sterically congested
nitriles, the desired alkenylated products were obtained
exclusively, presumably due to the sensitivity of the rate of
silylation to the steric effect6a,b (entries 7-10). Alkenyl
cyanides were also applicable to this catalysis to furnish
dienylsilanes 14 (entry 13).
Table 1. Reaction Optimizationa
A plausible mechanism is illustrated in Scheme 2. A
catalyst precursor A initially reacts with disilane to generate
catalytically active silylrhodium species B. Silicon-assisted
yields (%)b
(5) (a) Gooꢀen, L. J.; Paetzold, J.; Winkel, L. Synlett 2002, 1721. (b)
Myers, A. G.; Tanaka, D.; Mannion, M. R. J. Am. Chem. Soc. 2002, 124,
11250. (c) Tanaka, D.; Myers, A. G. Org. Lett. 2004, 6, 433. (d) Tanaka,
D.; Romeril, S. P.; Myers, A. G. J. Am. Chem. Soc. 2005, 127, 10323. (e)
Hu, P.; Kan, J.; Su, W.; Hong, M. Org. Lett. 2009, 11, 2341.
(6) (a) Tobisu, M.; Kita, Y.; Chatani, N. J. Am. Chem. Soc. 2006, 128,
8152. (b) Tobisu, M.; Kita, Y.; Ano, Y.; Chatani, N. J. Am. Chem. Soc.
2008, 130, 15982. (c) Tobisu, M.; Nakamura, R.; Kita, Y.; Chatani, N.
J. Am. Chem. Soc. 2009, 131, 3174.
entry
Si
ligand
2
3
1c
2
3
4
5
6
7
SiEt3 (a)
SiEt3 (a)
SiEt3 (a)
SiMe3 (b)
SitBuMe2 (c)
SiiPr3 (d)
Si(OiPr)3 (e)
P(4-CF3C6H4)3
P(4-CF3C6H4)3
P(4-FC6H4)3
P(4-FC6H4)3
P(4-FC6H4)3
P(4-FC6H4)3
P(4-FC6H4)3
46
60
65
25
36
11
68d
22
28
18
25
14
79
0
(7) For reviews on this topic, see: (a) Fagnou, K.; Lautens, M. Chem.
ReV. 2003, 103, 169. (b) Hayashi, T.; Yamasaki, K. Chem. ReV. 2003, 103,
2829. (c) Miura, T.; Murakami, M. Chem. Commun. 2007, 217.
(8) Nakao, Y.; Yada, A.; Satoh, J.; Ebata, S.; Oda, S.; Hiyama, T. Chem.
Lett. 2006, 35, 790.
a Reaction conditions: methyl 4-cyanobenzoate (0.5 mmol), vinylsilane
(2 mmol), hexamethyldisilane (1 mmol), [RhCl(cod)]2 (0.025 mmol), ligand
(0.05 mmol), ethylcyclohexane (0.5 mL), 130 °C, 15 h. b Isolated yields
based on nitrile. c Run using 1a (1 mmol). d NMR yield.
(9) Tobisu, M.; Hyodo, I.; Onoe, M.; Chatani, N. Chem. Commun. 2008,
6013.
(10) Other ligands examined (ligand (yield of 1a, yield of 2a)): PPh3
(53%, 27%), P(4-MeOC6H4)3 (37%, 31%), P(C6F5)3 (55%, 14%), P(2-
MeC6H4)3 (24%, 74%), PCy3 (22%, 0%), PBu3 (25%, 64%), P(OPh)3 (42%,
42%), bipyridine (25%, 10%), IPr (55%, 28%).
One possible reason for the inapplicability of other alkenes
may be the side reactions initiated by silylrhodation of
alkenes, such as dehydrogenative silylation and oligomer-
(11) For the results using other vinylsilanes, see the Supporting
Information.
Org. Lett., Vol. 12, No. 8, 2010
1865