Scheme 1. Mitsunobu Cyclodehydration of 1,2-Diols:
Scheme 2. Phosphine-Dependent Stereodivergent Synthesis of
Commonly Accepted Mechanism
Triphenyloxirane
isolated the product with opposite absolute configuration and
exceptionally high ee (99%). In this case, an opposite
stereochemical pathway yielded inversion of the initial 1,2-
diol configuration. This example is an interesting case of
phosphine-dependent stereoselectivity in the Mitsunobu
cyclodehydration. To shed light on this subject, we undertook
a study to establish how the electronic factors in the starting
material affect the stereochemical outcome of the reaction.
Starting from optically pure (S)-mandelic acid and the
corresponding bromo benzene derivatives several 1,1-diaryl-
2-phenylethane-1,2-diols (Figure 1) were synthesized using
traditionally as the reaction pathway going through the less
sterically demanding betaine, the one bearing the oxyphos-
phonium ion in terminal position (path A). This trend, how-
ever, does not apply to the cyclodehydration of styrene 1,2-
diols. Evans and co-workers showed that the Mitsunobu dehy-
dration of (S)-1-phenyl-1,2-ehtanediol (PPh3/DEAD) provides
essentially racemic epoxide.12 It was postulated that the origin
of racemization was that the corresponding betaines, paths
A and B (Scheme 1), formed and collapsed at the same rate.
Recently, Weissman and co-workers reported that the stereo-
specificity of the Mitsunobu dehydration of 1-aryl-1,2-
ethanediols largely depends on the nature of the phosphine
used and the aryl groups in the substrate.16 Thus, electron-
rich phosphines (e.g., PCy3) and electron-withdrawing groups
in the aryl 1,2-diol moiety provide the corresponding epoxide
with retention of configuration (path A, Scheme 1).
Figure 1. Chiral 1,1-diaryl-2-phenylethane-1,2-diols used in this
study.
As part of a project devoted to the preparation of sterically
demanding chiral 1,2-amino alcohols and their use as a
catalysts in the enantioselective dialkylzinc addition to
carbonyl compounds,17-19 we explored the preparation of
optically active triphenylethylene oxide from the correspond-
ing 1,1,2-triphenylethanediol by means of the Mitsunobu
cyclodehydration reaction. With this purpose, we reacted
optically pure (S)-1,1,2-triphenylethandiol (1a) with triphenyl-
phosphine and diisopropylazodicarboxylate (DIAD) at room
temperature in CH2Cl2 (Scheme 2). We isolated the corre-
sponding oxirane 2a with retention of configuration and a
high optical purity (88% ee). Surprisingly, when the same
reaction was performed with n-Bu3P instead of PPh3 we
standard reaction procedures.20 The aryl groups introduced
were chosen to provide a diverse electronic environment at
the 1,2-diol quaternary center. Product diols were purified
by recrystallization until optically pure (99% ee) as deter-
mined by chiral HPLC.21 With these compounds in hand,
we examined the Mitsunobu cyclodehydration with DIAD
and three different phosphines, Ph3P, n-Bu3P and Oct3P,
which are widely used in synthesis (Table 1).
As with the parent compound 1a, cyclodehydration of diols
1b, 1d, 1e, and 1f also provided the corresponding epoxides
from excellent to moderate yields. Only oxirane derived from
1c could not be isolated or identified in the final reaction mix-
ture. This was probably due to the intrinsic instability of the
final product.22 The electron-releasing nature of the p-methoxy-
(15) Pautard-Cooper, A.; Evans, S. A., Jr. J. Org. Chem. 1989, 54, 2485-
2488.
(16) Weissman, S. A.; Rossen, K.; Reider, P. J. Org. Lett. 2001, 3, 2513-
2515.
(17) Sola`, L.; Reddy, K. S.; Vidal-Ferran, A.; Moyano, A.; Perica`s, M.
A.; Riera, A.; Alvarez-Larena, A.; Piniella, J. F. J. Org. Chem. 1998, 63,
7078-7082.
(18) Fontes, M.; Verdaguer, X.; Sola, L.; Pericas, M. A.; Riera, A. J.
Org. Chem. 2004, 69, 2532-2543.
(20) Millar, A.; Mulder, L. W.; Mennen, K. E.; Palmer, C. W. Org. Prep.
Proced. Int. 1991, 23, 173-180.
(21) Diol 1e could not be purified by crystallization, and it was used as
obtained in 93% enantiomeric excess.
(19) (a) Garc´ıa-Delgado, N.; Reddy, K. S.; Sola, L.; Riera, A.; Pericas,
M. A.; Verdaguer, X. J. Org. Chem. 2005, 70, 7426-7428. (b) Garc´ıa-
Delgado, N.; Fontes, M.; Perica`s, M. A.; Riera, A.; Verdaguer, X.
Tetrahedron: Asymmetry 2004, 15, 2085-2090.
(22) As suggested by a reviewer, the inability to find the expected epoxide
could also be explained by the migration of the electron-rich aryl group
and concomitant loss of triphenylphosphine oxide in the initial betaine to
yield a rearranged ketone.
636
Org. Lett., Vol. 9, No. 4, 2007