A R T I C L E S
Yasuda et al.
Table 3. Diastereoselective Addition of 1d or 1e with 2a
Scheme 1. Plausible Path in the Reaction of 1a or 1b
of acetophenone 2a to the stirred mixture of 1a and SnCl2 for
20 min in acetonitrile gave a lower yield (41%) than that in
Table 1. The active species generated by transmetalation might
be gradually decomposed or deactivated by oligomerization.
Interestingly, no color change was found during the general
reaction procedure in which 1a was added to the premixed
suspension of 2a and SnCl2. In this case, the active species is
immediately consumed and smoothly gives the product.11 On
the basis of the stereochemical outcome, plausible reaction paths
are proposed in Scheme 1. At first, transmetalation of 1a with
SnCl2 proceeds to form an active nucleophile which is tenta-
tively considered to be a cinnamyltin(II) species.7,12,13 A cyclic
transition state is favorable because the chlorinated tin(II) center
is highly capable of accepting ligands.13a,b,14 In this transition
state A, a methyl group in ketone 2 (R ) Me) occupies an axial
position owing to its less steric hindrance than that of the aryl
group (Ar), giving highly diastereoselective formation of
homoallylic alcohols 3 (anti product). The alkylation with allylic
stannanes 1b or 1c which have a cis olefinic site also proceeds
through cyclic transition state B to give 6 (syn product). When
cinnamyl Grignard was used instead of 1a as an alkyl source
for transmetalation with SnCl2, no reaction with acetophenone
was observed.15,16 This result shows tin reagents are indispen-
sable to generate an active species for the alkylation in which
Bu3SnCl probably plays a key role to form a monomeric active
tin(II) species.13a
a Allylic stannane 1 was added to the mixture of 2 and SnCl2 in
acetonitrile. b The ketone 2 was added to the mixture of 1 and SnCl2 in
acetonitrile. c All reactions were carried out in dichloromethane with 1 (1.0
mmol), 2 (1.0 mmol), and BF3‚OEt2 (2.0 mmol).
of the olefinic moiety in the starting allylstannanes. For the BF3‚
OEt2 reactions with ketones in Tables 1 and 2, in fact, syn-type
products were predominantly obtained.
On the contrary, the reaction of crotylstannane showed
unexpected results as depicted in Table 3. Crotylstannane can
be prepared as the E-rich form 1d or Z-form 1e. The use of
E-rich crotylstannane 1d provided the corresponding homoallylic
alcohols 7a and 8a in syn selectivity (entry 1). The selectivity
was improved to 24/76 when the ketone was added to the
mixture of 1d and SnCl2 in acetonitrile (entry 2). Surprisingly,
the use of Z-crotylstannane 1e gave the same isomer 8a as a
major product in 70% selectivity (entry 3). Ketone 2b also gave
preferably syn product 8b using either 1d or 1e (entries 6 and
7). On the other hand, the reactions with 2a using either 1d or
1e at higher temperature after stirring at room temperature gave
anti product 7a as a major product (entries 4 and 5). These
results suggest that the intermediate nucleophiles are identical
in both cases using 1d and 1e and that the homoallylic alcohol
8a is a kinetic product and 7a is a thermodynamic product.
Interestingly, the BF3‚OEt2-accelerated alkylation showed op-
posite selectivity as compared with the reactions using SnCl2.
The reaction of 1d and/or 1e in the presence of BF3‚OEt2 gave
anti product 7a as the major product at -60 °C (entries 8 and
9) and syn alcohol 8a at higher temperature (entry 10).
The reaction course of crotylstannane with SnCl2 can be
explained as shown in Scheme 2. The transmetalation of either
1d or 1e gives a common intermediate 3-metallobut-1-enyl
species by rearrangement, which kinetically transforms to a
Z-crotyl species.19 The Z-crotyl species might then lead to 8Sn
preferably through cyclic transition state D. Under the reversible
reaction conditions, the product ratio was thermodynamically
controlled between 7Sn and 8Sn. Although the reaction mech-
anism using BF3‚OEt2 is not clear yet, it is quite interesting
that the Lewis acid-accelerated system has complementary
selectivity to the SnCl2-mediated system. The reported examples
Because there are many precedents of the reaction courses
of Lewis acid-accelerated allylation of aldehydes using allylic
stannanes,17 we assume that the reactions with ketones using
BF3‚OEt2 proceed through an acyclic transition state as proposed
by Yamamoto1b or Keck.18 In the reaction system, the stereo-
chemistry of the major products is independent of the geometry
(11) The mixture of 1a and SnCl2 in dichloromethane gave Bu3SnCl in 13%
even after 7 h. The SnCl2-mediated reaction of 1a with 2a in dichlo-
romethane gave no product as described above. Thus, the correlation
between transmetalation and yield was observed.
(12) Yasuda, M.; Matsukawa, Y.; Okamoto, K.; Sako, T.; Kitahara, N.; Baba,
A. Chem. Commun. 2000, 2149-2150.
(13) A referee suggested that the tin(II) species is unlikely to exist as a
monomeric structure. The result of stereoselectivity apparently shows the
cyclic transition state as illustrated in Scheme 1. The monomeric tin(II)
compounds coordinated by two appropriate ligands are reported, and the
tin center coordinates to Lewis acid. Therefore, in the transition state, the
actual species could be chlorinated allylic tin(II) coordinated by carbonyl
oxygen and acetonitrile and interacting with Bu3SnCl. (a) Drost, C.;
Hitchcock, P. B.; Lappert, M. F. Organometallics 1998, 17, 3838-3840.
(b) Barret, M. C.; Mahon, M. F.; Molloy, K. C.; Steed, J. W.; Wright, P.
Inorg. Chem. 2001, 40, 4384-4388.
(14) Zimmerman, H. E.; Traxler, M. D. J. Am. Chem. Soc. 1957, 79, 1920-
1923.
(15) To a solution of cinnamyl Grignard reagent (2 equiv) and SnCl2 (2 equiv)
in Et2O was added acetophenone (1 equiv) at 0 °C, and the mixture was
stirred for 3 h.
(16) The color change from gray to red-brown was observed when mixing the
cinnamyl Grignard reagent and SnCl2. This system probably gives an
inactive species, for example, bis(cinnamyl)tin(II), by transmetalation.
(17) Nishigaichi, Y.; Takuwa, A.; Naruta, Y.; Maruyama, K. Tetrahedron 1993,
49, 7395-7426.
(18) Keck, G. E.; Savin, K. A.; Cressman, E. N. K.; Abbott, D. E. J. Org. Chem.
1994, 59, 7889-7896.
(19) Miyake, H.; Yamamura, K. Chem. Lett. 1992, 1369-1372. In a cinnamyl
case, the Z-form is too unstable to form, and then the E-form, which is a
thermodynamic product, acts as the active species.
9
13444 J. AM. CHEM. SOC. VOL. 124, NO. 45, 2002