Effects of silyl substitution of the olefin in olefin carbo-
metalation were investigated some time ago12 as well as in
recent years.13 It has been shown that a Group 14 metal
substituent, a stannyl substitutent in particular,14 not only
accelerates the carbometalation but controls the regioselec-
tivity so that the reaction generates a gem-dimetallic product
(e.g., the compound corresponding to 7 rather than 6). Thus,
the addition of allylzinc reagents to the silyl, germyl, and
stannyl CPAs (1d-f) proceeded far faster than that to the
carbon-subsituted CPAs (1b and 1c). For instance, the
addition of allylzinc bromide to the Group 14 metal deriva-
tives was complete in an hour at 0 °C at ambient pressure
and afforded the product with 80:20-95:5 selectivity favor-
ing the formation of 715 (Table 2, entries 1-3). Note that
regioselectivity has been ascribed to the predominant elec-
trostatic interaction between the Lewis acidic zinc atom and
the developing negative charge on the carbon connected to
the metal substituent (A in Scheme 3).7
Scheme 3. Regiochemistry of Allylzincation of CPAs Bearing
a Group 14 Metal Substituent
Table 2. Regio- and Enantioselective Allylzincation of
Substituted CPA (1d-f)
To our surprise, however, this intrinsic regioselectivity can
be completely reversed, favoring 6 with 94:6-98:2 selectivity
(entries 4-6), when the BOX-bearing allylzinc reagent 2
reacted with 1d-f. The enantioselectivity of the addition of
2 to 1d-f was excellent (97.8-99.8% ee), and the sense of
the selectivity was the same as that described in Table 1.
The observed regio- and enantioselectivities combined with
the transition structure (TS) obtained for the parent CPA (1a)
by the ab intio calculation8 indicate that the reaction took
place through TS B shown in Scheme 3. Because of the
presence of both the acetal and the Me3M moiety, the
electronically favored TS A′ is precluded by the steric effect
of the bulky BOX ligand.
To examine further the origin of the reversal of the
regioselectivity, we studied the reactions of allylzinc bromide
and 2 with the simple vinylstannane 8. In contrast to the
CPA cases, both reactions occurred in a manner dictated by
the electronic property of the vinylstannane and afforded the
same product 9 after hydrolysis. The results indicate that
suitable steric factors will overwhelm the electronics of the
reaction.
1
a Determined by H NMR. b Combined yield of regioisomers obtained
by isolation. c Determined by 1H NMR and HPLC analysis after conversion
to MTPA derivatives.
allylzinc bromide addition to 2-ethyl-CPA 1b was incomplete
even after 36 h at 25 °C (Table 1, entry 1). The observed
(9) The zinc reagent 2 (R ) C6H5) was prepared first. Thus, to a solution
of phenyl-substituted (R,R)-BOX (818 mg, 2.67 mmol) and 2,2′-bipyridyl
(ca. 1 mg) in dry THF (4.0 mL) was added a 1.49 M hexane solution of
BuLi (1.8 mL, 2.68 mmol) at 0 °C, and the mixture was stirred further for
20 min at 25 °C. To the mixture was added a 1.58 M THF solution of
allylzinc bromide (1.58 mL, 2.43 mmol) at 0 °C. After stirring for 20 min
at 25 °C, a solution of CPA 1c (480 mg, 2.20 mmol) in dry THF (2.67 mL)
was added at 0 °C. The reaction mixture (8.0 mL) was transferred into a
Teflon vessel under nitrogen, kept under high pressure (1 GPa) for 24 h at
25 °C, and then quenched with saturated NH4Cl (20 mL). Purification with
silica gel chromatography afforded the allylation product (R)-6c (R2 ) H)
(424 mg, 93% yield): IR (neat) 3073, 2956, 2856, 1639, 1602, 1498, 1471,
1446, 1351, 1292, 1157, 1132, 1070, 1043, 910, 700; 1H NMR (400 MHz,
CDCl3) δ 0.93 (s, 3 H), 1.07 (d, J ) 5.9 Hz, 1 H), 1.07 (s, 3 H), 1.37 (dd,
J ) 1.5, 5.9 Hz, 1 H), 2.34 (dd with shoulders, J ) 7.3, 14.7 Hz, 1 H),
2.76 (ddd, J ) 1.5, 6.4, 14.7 Hz, 1 H), 3.30 (d, J ) 11.0 Hz, 1 H), 3.35
(dd, J ) 0.97, 11.0 Hz, 1 H), 3.60 (distorted dd, J ) 0.97, 10.7 Hz, 1 H),
3.62 (distorted d, J ) 10.7 Hz, 1 H), 4.86-4.94 (m, 2 H), 5.61-5.73 (m,
1 H) 7.15-7.22 (m, 1 H) 7.27-7.31 (m, 4 H); 13C NMR (100 MHz, CDCl3)
δ 22.0, 22.1, 22.7, 30.6, 37.4, 38.6, 75.5, 76.4, 92.0, 116.3, 126.2, 127.9 (2
C), 129.2 (2 C), 135.4, 139.4; [R]20 ) 1.02 (c ) 3.9, benzene). Anal.
Calcd for C17H22O2: C, 79.03; H, 8.59. Found: C, 79.07; H, 8.83. The
R-configuration of this product was determined through an eight-step
conversion to (R)-2-methyl-2-phenylbutanoic acid.
D
(10) (a) le Noble, W. J.; Kelm, H. Angew. Chem., Int. Ed. Engl. 1980,
19, 841-946. (b) Matsumoto, K.; Sera, A.; Uchida, T. Synthesis 1985, 1-26.
(c) Matsumoto, K.; Sera, A. Synthesis 1985, 999-1027. (d) van Eldik, R.;
Asano, T.; le Noble, W. J. Chem. ReV. 1989, 89, 549-688. (e) Isaccs, N.
S. Tetrahedron 1991, 47, 8463-8497.
(11) Kubota, K.; Isaka, M.; Nakamura, M.; Nakamura, E. J. Am. Chem.
Soc. 1993, 115, 5867-5868.
In summary, we have demonstrated that an allylic zinc
reagent bearing a chiral BOX ligand adds to a substituted
Org. Lett., Vol. 2, No. 15, 2000
2195