Scheme 3
Scheme 5
appears to inhibit the reaction. The olefin geometry does not
appear to play a crucial role in this transformation as a cis-
and trans-alkene both furnish the same stereoisomer of
product in comparable yield and selectivity (compare entries
2 and 3). Notably, the reaction is also effective in forming
a quaternary stereocenter with a high level of stereoinduction
(entry 7).
Considering that Pd catalysts can facilitate both allylic
substitution reactions and alcohol oxidation reactions, the
stereochemical integrity of the stereocenter in the reaction
substrates will be of consequence to engaging the stereose-
lective Oshima-Utimoto reaction in asymmetric synthesis.
Chiral allylic alcohol 5 provided a suitable substrate to probe
for potential epimerization during the course of the Oshima
reaction. Fortunately, upon subjection of 5 to the standard
reaction conditions, the derived cyclic acetal could be isolated
in good yield, and each anomer was furnished as a single
diastereomer with the configuration as depicted in Scheme
3.
undergoes cis/trans isomerization more rapidly than the
Oshima reaction occurs.8 In a second experiment, 1 equiv
of ethyl vinyl ether was combined with 1 equiv of Pd(OAc)2
and was found to result in a structure consistent with
1
oxonium ion 7, as determined by H NMR (1D, COSY)
analysis (Scheme 4).9,10 The spectral data for compound 7
are similar to those reported for the analogous oxonium ion
derived from ethyl vinyl ether and fluoroantimonic acid in
SO2 solvent.11 Notably, the formyl hydrogen was found to
resonate at 9.6 ppm and appears as a triplet due to coupling
with HC (J ) 4.9 Hz).12 Consistent with this assignment,
the oxonium ion derived from deuterium-labeled substrate
(E)-6 exhibits a doublet for the formyl hydrogen resonance.
On the basis of the stereochemical outcome of the reaction
and the observations mentioned above, it appears plausible
that the reaction proceeds by addition of the allylic alcohol
to oxonium ion 7, as depicted in Scheme 5. Subsequent
stereoselective carbopalladation of the tethered alkene through
a transition structure that may resemble chairlike conformer
8 provides a rationale for the observed stereoinduction.
Finally, â-H elimination would provide the reaction product
and release a catalyst which requires reoxidation prior to re-
formation of ion 7. We posit that the lack of stereocontrol
at the anomeric carbon results from a kinetic profile wherein
nonselective addition of the alcohol to 7 is irreversible under
To learn about the mechanism of the Oshima-Utimoto
reaction, we examined this process with deuterated vinyl
ether (E)-6 (Scheme 4).7 Isolation of the reaction product
Scheme 4
(8) For Pd-catalyzed isomerization of vinyl ethers, see: McKeon, J. E.;
Fitton, P. Tetrahedron 1972, 28, 233.
(9) For IR characterization of a Pd-olefin complex derived from a vinyl
ether and palladium acetate, see: Wakatsuki, Y.; Nozakura, S. I.; Murahashi,
S. Bull. Chem. Soc. Jpn. 1972, 45, 3426.
(10) An oxonium ion related to 7 has been postulated previously: (a)
Henry, P. M. J. Am. Chem. Soc. 1972, 94, 7316. For a related iminium ion
which appears to be involved in indole arylation, see: (b) Lane, B. S.;
Brown, M. A.; Sames, D. J. Am. Chem. Soc. 2005, 127, ASAP.
(11) Olah, G. A.; DeMember, J. R.; Mo, Y. K.; Svoboda, J. J.; Schilling,
P.; Olah, J. A. J. Am. Chem. Soc. 1974, 96, 884.
and chromatographic separation of the acetal epimers
revealed that the deuterium atom was scrambled at the
methylene carbon of the furan ring. NMR analysis during
the course of the reaction showed that vinyl ether (E)-6
(12) The 1H NMR resonances for both HB and HC are broadened,
indicating a fluxional process. The resonances for unreacted ethyl vinyl
ether are unperturbed, suggesting the equilibrium transformation involving
6 includes a separate species. For a similar equilibrium process involving
pyridine addition to a Pd(II) vinyl ether complex, see ref 9.
(7) Keul, H.; Choi, H.-S.; Kuczkowski, R. L. J. Org. Chem. 1985, 50,
3365.
Org. Lett., Vol. 7, No. 15, 2005
3369