with TMSCN. The (S,S)-Ph-pybox ligand 2c afforded the
highest enantioselectivity, yielding the product in 67% ee
Investigation of the scope of the ARO was conducted
under the optimized conditions outlined above (Table 3).
Table 3. ARO of Meso Epoxides with TMSCN Catalyzed by
(pybox)YbCl3 Complexesa
(Table 2). Moreover, the product obtained with (S,S)-2c was
of opposite configuration to that obtained with the other (S,S)-
pybox ligands.17
Table 2. Effect of Ligand in the ARO of Cyclohexene Oxide
with TMSCN Catalyzed by YbCl3‚6H2O. Reactions Were Run
at Room Temperature for 4 h with 5 mol % of Catalyst
ligand
yield (%)a
ee (%)b
confignc
(S,S)-2a
(S,S)-2b
(S,S)-2c
(S,S)-2d
(S,S)-2e
95
93
96
96
93
47
57
-67
28
1S,2R
1S,2R
1R,2S
1S,2R
1S,2R
28
a ,bSee Table 1. c Absolute configuration of the predominant enantiomer
of 1 generated with (S,S)-ligand.
While it was possible to achieve good yield and enantiose-
lectivity for a variety of meso epoxides, the optimal ligand
and reaction temperature proved to be highly substrate-
dependent. For example, the ARO of cyclopentene oxide
carried out at room temperature proceeded in 33% ee with
ligand 2c but in 63% ee with the t-Bu-pybox ligand 2b. At
-10 °C, the enantioselectivity was improved to 92%. For
those substrates examined, epoxides fused to five-membered
rings (entries 2, 4, and 5) performed best with the t-Bu-pybox
ligand 2b, whereas the Ph-pybox ligand 2c proved superior
with cyclohexene oxide and cis-2-butene oxide. As was
observed in the ARO of cyclohexene oxide (see above), these
two ligands consistently afforded product of opposite abso-
lute configuration.
While the results summarized in Table 3 compare favor-
ably with the state-of-the-art for ARO of meso epoxides with
TMSCN,7 it is clear that the reaction is by no means ideal
from a practical standpoint. In particular, the reactions must
be carried out at reduced temperature in order to achieve
high enantioselectivity, and as a result several days are
required to attain optimal yields. To glean some insight into
this process as a first step toward the design of improved
catalysts, we have carried out a set of preliminary mechanistic
investigations that are summarized below.
The reaction conditions for the ARO of cyclohexene oxide
with 2c‚YbX3‚6H2O were optimized systematically with
respect to both enantioselectivity and yield. While all
ytterbium salts that were studied catalyzed formation of 1
in good yield, the complex derived from YbCl3 afforded
substantially higher enantioselectivity than those derived from
the corresponding bromide, triflate, or alkoxide salts. A
strong solvent dependence was observed, with halogenated
solvents providing best results and nonpolar solvents such
as toluene or TBME leading to low conversions and ee’s.
Chloroform proved to be the optimal solvent, providing ring-
opened product 1 in 92% yield and 72% ee at room
temperature. Decreased reaction temperatures led to im-
provement in ee’s, although at the expense of reaction rate.
Thus, at -40 °C, the ARO could be carried out with 10 mol
% of catalyst, affording 1 in 91% ee and 90% isolated yield
after 4 days.
(14) (a) Nishiyama, H.; Sakaguchi, H.; Nakamura, T.; Horihata, M.;
Kondo, M.; Itoh, K. Organometallics 1989, 8, 846. (b) Nishiyama, H.;
Sakaguchi, H.; Nakamura, T.; Horihata, M.; Kondo, M.; Itoh, K. Organo-
metallics 1991, 10, 500. (c) Nishiyama, H.; Yamaguchi, S.; Kondo, M.;
Itoh, K. Organometallics 1992, 11, 4306.
(15) Evans, D. A.; Kozlowski, M. C.; Murry, J. A.; Burgey, C. S.;
Campos, K. R.; Connell, B. T.; Staples, R. J. J. Am. Chem. Soc. 1999, 121,
1, 669. For a recent review of chiral bis(oxazoline)-metal catalyzed reactions,
see: Ghosh, A. K.; Mathivanan, P.; Cappiello, J. Tetrahedron: Asymmetry
1998, 9, 1.
(16) Aspinall, H. C.; Greeves, N.; Smith, P. S. Tetrahedron Lett. 1999,
40, 1763.
(17) A reversal of the absolute sense of selectivity from (S,S)-tert-butyl-
to (S,S)-phenyl-bis(oxazoline) ligands has been observed previously in
asymmetric bis(oxazoline)-Cu(II) catalyzed hetero-Diels-Alder and ene
reactions. Evans, D. A.; Johnson, J. S.; Burgey, C. S.; Campos, K. R.
Tetrahedron Lett. 1999, 40, 2879.
As noted above, a chloride transfer byproduct was obtained
in the initial stages of the ARO of cyclohexene oxide in
yields roughly equal to the amount of catalyst employed.
This, in combination with the observations made by Utimoto
with achiral ytterbium catalysts,10 is consistent with the
requisite formation of a cyanoytterbium species as the active
catalyst.18 While this might suggest a role of nucleophile-
delivery agent for the catalyst, the observation of significant
Org. Lett., Vol. 2, No. 7, 2000
1003