Next, we converted 10 into the 2,3-cis-3,4-trans-4,5-trans-
tetrahydrofuran 7a (Scheme 3). Treatment of 10 with 4-tert-
Table 1. Stereoselectivity of the Reductive Deoxygenation
Reaction
Scheme 3. Reductive Deoxygenation of Cyclic Hemiketal 9a
ratio
(I:II:III)
entry substrate
conditions
1
2
3
9a
17
9a
BF3·OEt2, NaBH3CN, -78 °C, 30 min 10:1:0
BF3·OEt2, Et3SiH, -78 to -20 °C, 3 h 25:1:0
BF3·OEt2, -78 to -20 °C, 2 h
1:0.1:4
then NaBH3CN, -78 °C, 30 min
butyldimethylsilyloxy-3-methoxyphenyllithium gave a 4:1
anomeric mixture of the cyclic hemiketal 9a in 70% yield
(86% based on recovered starting material). We expected
that treatment of 9a with Et3SiH in the presence of BF3‚
OEt211 would preferentially provide the 2,3-cis-3,4-trans-4,5-
trans-tetrahydrofuran 7a through the addition of hydride from
the inside face of the envelope conformer (vide infra).
However, the reaction conditions for reductive deoxygenation
(BF3‚OEt2, Et3SiH, -78 to -20 °C, 9 h) gave a 1.3:1
diastereomeric mixture of 2,5-diaryl-3,4-dimethyltetrahy-
drofurans in poor yield (<20%). To our surprise, careful
the reaction proceeded to give a 10:1 diastereomeric mixture
of 9a-I and 9a-II (99%) without epimerization of the C2-
aryl group (entry 1). In the case of the electron-withdrawing
Bz protecting group on the C2-aryl substituent, the reductive
deoxygenation reaction of 1715 (BF3‚OEt2, Et3SiH, -78 to
-20 °C, 3 h) also proceeded without epimerization of the
C2-aryl group to give 17-I in excellent diastereoselectivity
(17-I:17-II ) 25:1, 62%) (entry 2). However, epimerization
of 9a, afforded by treatment with BF3‚OEt2 (-78 to -20
°C, 2 h), followed by reduction with NaBH3CN provided a
1:0.1:4 mixture of 9a-I, 9a-II, and 9a-III (96%) (entry 3).
Treatment of 9a with BF3‚OEt2 (-78 to -20 °C, 2 h) in
the absence of a reducing agent resulted in an equilibrium
between 9a and 9b (1:4 ratio) proving the observed BF3‚
OEt2-promoted epimerization of 9a (Scheme 4). The epimer-
ization of 9a might occur through a quinonoid oxonium ion
intermediate facilitated by an inductive effect of the electron-
donating Bn group on the C2-aryl substituent.3a,b,13 Further
study is required to prove the putative mechanism. The
preference for 9b in equilibrium can be explained by
unfavorable steric interaction between cis substituents in 9a.
To determine the stereochemical outcome of hydride reduc-
tion, we isolated and independently subjected 9a and 9b to
reductive deoxygenation conditions (BF3‚OEt2, NaBH3CN,
-78 °C, 30 min). Under the reaction conditions, 9a and 9b
provided 7a (7a:7c ) 10:1, 99%) and 7b (single diastere-
omer, 92%), respectively.16 The stereochemical outcome can
be explained by Woerpel’s recent studies.17 Due to unfavor-
able steric interactions of the incoming hydride with axially
oriented 3,4-dimethyl groups in conformation A, the hydride
adds to the sterically more favorable conformation B from
1
analysis of H NMR spectral data revealed that the major
diastereomer had the 2,3-trans-3,4-trans-4,5-trans-configu-
ration 7b and the minor diastereomer had the desired 2,3-
cis-3,4-trans-4,5-trans-configuration 7a, indicating that epimer-
ization of the C2-aryl group occurred under the reaction
conditions.12
The observed epimerization of 9a was rationalized on the
basis that Lewis acid activation of the hemiketal 9a by BF3‚
OEt2 combined with an inductive effect of the electron-
donating Bn group on the C2-aryl substituent effectively
competed with slow reduction of the oxocarbenium ion
intermediate 8a by Et3SiH.13 On the basis of this rationale,
we expected that either fast reduction of 8a or a change of
the electron-donating Bn group on the aryl substituent to an
electron-withdrawing group would prevent the epimerization
of the C2-aryl group.14
Table 1 summarizes the stereoselectivity of the reductive
deoxygenation reaction of 9a. When 9a was treated with BF3‚
OEt2 combined with NaBH3CN (a strong reducing agent),
(11) Yoda, H.; Mizutani, M.; Takabe, K. Heterocycles 1998, 48, 679.
(12) It is known that 2,5-diaryl-3,4-trans-dimethyltetrahydrofurans have
1
unique chemical shifts for H2, H3, H4, and H5 in H NMR depending on
their relative stereochemistry. Thus, we determined the relative stereo-
(15) 17 was prepared from 10 by Bn-deprotection, Bz-protection, and
ArLi-addition; see the Supporting Information for details.
chemistry of tetrahydrofurans 7a and 7b by comparison of chemical shifts
1
in H NMR with literature values; see refs 3-9.
(16) It is important to note that the cis,trans-hemiketal 9a and trans,trans-
hemiketal 9b are both configurationally stable under the reaction condition.
(17) (a) Smith, D. M.; Tran, M. B.; Woerpel, K. A. J. Am. Chem. Soc.
2003, 125, 14149. (b) Bear, T. J.; Shaw, J. T.; Woerpel, K. A. J. Org.
Chem. 2002, 67, 2056. (c) Larsen, C. H.; Riggway, B. H.; Shaw, J. T.;
Woerpel, K. A. J. Am. Chem. Soc. 1999, 121, 12208. (d) Shaw, J. T.;
Woerpel, K. A. Tetrahedron 1999, 55, 8747. (e) Shaw, J. T.; Woerpel, K.
A. J. Org. Chem. 1997, 62, 6706.
(13) For an example of Lewis acid-mediated fragmentation/isomerization
of furofurans, see: Aldous, D. J.; Dalencon, A. J.; Steel, P. G. J. Org.
Chem. 2003, 68, 9159.
(14) Hanessian et al. reported a method for the stereocontrolled synthesis
of 2,5-diaryl-3,4-dimethyltetrahydrofuran lignans by modulating the nature
of a directing para substituent on one of the aryl groups; see refs 3a and
3b.
Org. Lett., Vol. 9, No. 20, 2007
3967