configuration of the resolution products was determined by
comparison with an authentic material, prepared from the
chiral 2-iodo-2-cyclohexen-1-ol7 and 3,4-dimethoxyphenyl-
boric acid by Suzuki coupling conditions.8 This comparison
revealed (+)-3 to be R and (-)-3 to be S (Scheme 2).
Scheme 2
Figure 1.
with chromium trioxide and 3,5-dimethylpyrazole complex12
in dichlomethane to give rise to cyclohexenone 15 by allylic
oxidation (Scheme 3).
Scheme 3
On the other hand, transformation of the same racemic
alcohol (()-3 into morphinanone 2 was examined to prepare
for future utilization of (-)-(S)-3 for construction of natural
(-)-morphine 1. To construct the quaternary 13-stereogenic
center first, (()-3 was treated4,9 with ethyl vinyl ether in the
presence of N-bromosuccinimide (NBS) to give bromoacetal
5 as a mixture of two epimers. In contrast to the 3,4-
dimethoxyphenyl analogue which gave the cyclization
product in excellent yield,4 5 furnished the cyclization
product 6 in moderate yield under the same radical initiated
conditions,10 probably because of steric hindrance by the
2-substituent on the aromatic ring. Since we could not
optimize the conditions to improve the cyclization, we carried
out the present examination using product 6 obtained in 48%
yield. Thus, 6 was first transformed into lactone 7 on peracid
treatment in the presence of a Lewis acid.11 Product 7 was
then transformed into xanthate 10 by a sequential three-step
reaction via diol 8 and monopivalate 9. Upon thermolysis
under various conditions, 10 afforded no cyclohexene 14 but
instead gave a complex mixture. As this was presumed to
be due to steric repulsion between the xanthate moiety and
the axially disposed alkyl functionality in a transition state
(Figure 1, 10a), its epimer 13, with no such steric hindrance,
seems to be more appropriate for syn-elimination. Although
the Mitsunobu reaction of 9 failed to give the inverted alcohol
12, oxidation of 9 followed by reduction of the generated
ketone 11 with sodium borohydride furnished the epimeric
alcohol 12, diastereoselectively. The observed diastereo-
selective reduction may be due to prior coordination of the
pivalate moiety with the borohydride reagent which delivered
a hydride exclusively from the â-face of the molecule (Figure
1, 11a). As expected, on thermolysis in o-dichlorobenzene,
the epimeric xanthate 13, obtained from 12, afforded
cyclohexene 14 in satisfactory yield presumably through
conformer 13a (Figure 1). Intermediate 14 was then treated
To construct the C9-C10 bridge, an allyl functionality
was introduced at the C14 center of 15 under Sakurai
conditions13 which proceeded in a highly diastereoselective
manner. Thus, treatment of 15 with allyltrimethylsilane in
the presence of titanium(IV) chloride gave a readily separable
mixture of two products in yields of 71 and 5% (Scheme 4).
Although the stereochemistry of the products could not be
determined at this stage, the major product was determined
at a later stage as the stereoisomer 16 containing the 14-âH
configuration. The diastereoselective reaction observed at this
pivotal point which determined the direction of the natural
14-âH and the unnatural 14-RH stereochemistry was note-
worthy. This may be elucidated by assuming the structure
(7) Johnson, C. R.; Sakaguchi, H. Synlett 1992, 813.
(8) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457.
(9) Ueno, Y.; Chino, K.; Watanabe, M.; Moriya, O.; Okawara, M. J.
Am. Chem. Soc. 1982, 104, 5564.
(10) Stork, G.; Mook, R., Jr. J. Am. Chem. Soc. 1983, 105, 3720.
(11) Grieco, P. A.; Oguri, T.; Yokoyama, Y. Tetrahedron Lett. 1978,
419.
(12) Salmond, W. G.; Barta, M. A.; Havens, J. L. J. Org. Chem. 1978,
43, 2057.
2786
Org. Lett., Vol. 2, No. 18, 2000